From news.umbc.edu!haven.umd.edu!uunet!aadt!mikee Fri Aug 6 18:58:39 EDT 1993 Article: 6182 of comp.unix.solaris Newsgroups: comp.unix.solaris,comp.emacs Path: news.umbc.edu!haven.umd.edu!uunet!aadt!mikee From: mikee@rambo.aadt.com (Mike Eggleston) Subject: Re: Emacs 19.17 on solaris 2.x In-Reply-To: boucher@chopin.sofkin.ca's message of Thu, 5 Aug 1993 12:34:09 GMT Message-ID: Sender: root@uunet!aadt (Operator) Organization: American Airlines Decision Technologies References: <231jhlINN67h@chief.isi.com> <1993Aug4.103333.20426@infodev.cam.ac.uk> <23oec2$a8@mercury.cs.waikato.ac.nz> Date: Thu, 5 Aug 1993 15:31:10 GMT Lines: 14 Xref: news.umbc.edu comp.unix.solaris:6182 comp.emacs:17777 Status: RO In article boucher@chopin.sofkin.ca (Robert Boucher) writes: >Runs well here. Only 'list-options makes emacs crash and burn, but it >will be fixed in the next release. I had the same problem with 'list-options and traced it down to the floating point routines. In my .emacs file I placed (setq float-output-format "%.5g") and everything then works fine. -- -- Mike Eggleston American Airlines Decision Technologies (817)931-2287 P. O. Box 619616, MD 4462 (817)967-9763 (FAX) Dallas/Fort Worth Airport, Texas 75261-9616, USA Only my opinions. From news.umbc.edu!haven.umd.edu!uunet!cs.utexas.edu!newsfeed.rice.edu!rice!harry-clay!bob Fri Aug 6 18:59:42 EDT 1993 Article: 6184 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!uunet!cs.utexas.edu!newsfeed.rice.edu!rice!harry-clay!bob From: bob@hanszen.rice.edu (Robert Morrison Dana) Subject: Performance fix from Sun Message-ID: Sender: news@rice.edu (News) Organization: Rice University X-Newsreader: TIN [version 1.1 PL9] Date: Thu, 5 Aug 1993 16:38:09 GMT Lines: 106 Status: RO I posted a few days ago complaining that system performance was bad on my Classic- it was a problem with process scheduling. If a computationally intensive process was running, the interactive response of the machine was terrible. There was a noticeable lag for commands as simple as "ls". Very soon after posting this and receiving some replies saying "get more memory" (an incorrect solution), I received a response through email from AN ENGINEER AT SUNSOFT. Here's the scoop- the dispatch table table (which is used by the process scheduler to determine running priorities and switching times) is not so good in Solaris 2.2. He mailed me a better "unsupported" dipatch table which I am including at the end of this post. This was exactly what I needed- my performance complaints are gone. He said that this new table will be included in Solaris 2.3 along with some other "scheduling goodies", and that there would be more extensive repairs of the process scheduling in 2.4. I hope tech support like this becomes the norm- maybe Sun will start actually posting stuff like this sometime... -Robert -- Robert Dana ONCS UNIX Student Systems Programmer Rice University, Houston, TX Instructions: Save this to a file, and load it with the command harry-clay# dispadmin -c TS -s filename This will alter the scheduler until the next reboot. To find out about making this change permanent, read the man page for ts_dptbl harry-clay# man -s 4 ts_dptbl Cut here ------- # RES=500 # ts_quantum ts_tqexp ts_slpret ts_maxwait ts_lwait PRIORITY LEVEL 100 0 1 0 10 # 0 100 0 2 0 11 # 1 100 1 3 0 12 # 2 100 1 4 0 13 # 3 100 2 5 0 14 # 4 100 2 6 0 15 # 5 100 3 7 0 16 # 6 100 3 8 0 17 # 7 100 4 9 0 18 # 8 100 4 9 0 19 # 9 80 5 11 0 20 # 10 80 5 12 0 21 # 11 80 6 13 0 22 # 12 80 6 14 0 23 # 13 80 7 15 0 24 # 14 80 7 16 0 25 # 15 80 8 17 0 26 # 16 80 8 18 0 27 # 17 80 9 19 0 28 # 18 80 9 19 0 29 # 19 60 10 21 0 30 # 20 60 11 22 0 31 # 21 60 12 23 0 32 # 22 60 13 24 0 33 # 23 60 14 25 0 34 # 24 60 15 26 0 35 # 25 60 16 27 0 36 # 26 60 17 28 0 37 # 27 60 18 29 0 38 # 28 60 19 29 0 39 # 29 40 20 31 0 40 # 30 40 21 32 0 41 # 31 40 22 33 0 42 # 32 40 23 34 0 43 # 33 40 24 35 0 44 # 34 40 25 36 0 45 # 35 40 26 37 0 46 # 36 40 27 38 0 47 # 37 40 28 39 0 48 # 38 40 29 39 0 49 # 39 20 30 41 0 50 # 40 20 31 42 0 50 # 41 20 32 43 0 51 # 42 20 33 44 0 51 # 43 20 34 45 0 52 # 44 20 35 46 0 52 # 45 20 36 47 0 53 # 46 20 37 48 0 53 # 47 20 38 49 0 54 # 48 20 39 49 0 54 # 49 10 40 51 0 55 # 50 10 41 52 0 55 # 51 10 42 53 0 56 # 52 10 43 54 0 56 # 53 10 44 55 0 57 # 54 10 45 56 0 57 # 55 10 46 57 0 58 # 56 10 47 58 0 58 # 57 10 48 59 0 59 # 58 10 49 59 0 59 # 59 From news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!sdd.hp.com!saimiri.primate.wisc.edu!nntp.msstate.edu!usenet Fri Aug 6 19:01:11 EDT 1993 Article: 6198 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!sdd.hp.com!saimiri.primate.wisc.edu!nntp.msstate.edu!usenet From: fwp@CC.MsState.Edu (Frank Peters) Newsgroups: comp.unix.solaris Subject: Re: Why "add net 244.0.0.0" at boot? Date: 5 Aug 1993 22:37:03 GMT Organization: Computing Center, Mississippi State University Lines: 43 Distribution: inet Message-ID: <23s22f$6pb@Tut.MsState.Edu> References: <1993Aug4.182455.17189@gsusgi1.gsu.edu> Reply-To: fwp@CC.MsState.Edu NNTP-Posting-Host: jester.cc.msstate.edu Status: RO In article <1993Aug4.182455.17189@gsusgi1.gsu.edu> matess@gsusgi2.gsu.edu (EStrickler) says: : When I boot my SparcLX running Solaris 2.2, : I get a message: : : setting netmask of le0 : setting default interface for multicast: add net 244.0.0.0: gateway name : : Why and what is net 244.0.0.0? My network people are confused with : why I have 244.0.0.0 in my routing tables. Solaris 2 adds support for multicasts (see RFC 1112 for a protocol description and RFC 1301 for a good example of its uses). Basically, multicast is sort of a directed broadcast. Every machine on a network receives packets sent to the broadcast address. Only the individual machine receives packets sent to its specific IP address (I'm ignoring promiscuous mode and sniffers and the like). Suppose machine A wants to send a packet of information to machines B, G and H but not to machines C, D, E and F. With conventional IP it can send three packets, one to each destination, or a broadcast that every machine on the network will see. With multicast the machines can agree on a multicast address to listen to (I think of it as "subscribing" to the address). Machine A can then send its packet to that multicast address and all of the subscribed machines will see it while none of the unsubscribed machines will. This has a variety of potential uses...think of a group of machines keeping a distributed database in sync or a network "radio broadcast" system (you "tune in" to a station by joining the multicast address it is sending to). Multicast uses the class D IP address range (yes, there is a range of addresses that aren't class A, B or C addresses) from 224.0.0.0 to 239.255.255.255. And address 224.0.0.1 is the "permanent group of all IP hosts" (from RFC 1112). All of which is a long winded way of saying, its normal, harmless and potentially useful if you ever have an application that takes advantage of it. -- Frank Peters - UNIX Systems Programmer - Mississippi State University Internet: fwp@CC.MsState.Edu - Phone: (601)325-7030 - FAX: (601)325-8921 From news.umbc.edu!haven.umd.edu!uunet!pipex!bnr.co.uk!uknet!edcastle!dcs.ed.ac.uk!gdmr Sun Aug 8 18:55:32 EDT 1993 Article: 2540 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!uunet!pipex!bnr.co.uk!uknet!edcastle!dcs.ed.ac.uk!gdmr From: gdmr@dcs.ed.ac.uk (George Ross) Newsgroups: comp.unix.solaris Subject: Re: Installed location of X11 (Was: rn/xrn ported to Solaris 2.1) Message-ID: Date: 22 Mar 93 09:43:20 GMT References: <1o5g6fINN6u0@tamsun.tamu.edu> <1993Mar17.152925.21662@news.duc.auburn.edu> <1993Mar18.154343.12003@ims.com> Sender: cnews@dcs.ed.ac.uk (UseNet News Admin) Organization: Department of Computer Science, Edinburgh University Lines: 30 Status: RO In article , aad@scr.siemens.com (Anthony A. Datri) writes: > >>setup. But X11R5 allows you the luxury of having all of your X stuff under one > >>managable directory tree: $(TOP). > > You have to jump through hoops to get a decent layout, though. ... Not really... Here's the relevant part of our site.def: #undef UsrLibDir #define UsrLibDir /usr/local/lib/X11R5 #undef IncRoot #define IncRoot /usr/local/share/X11R5/include #undef UsrIncDir #define UsrIncDir /usr/local/share/include #undef ManDirectoryRoot #define ManDirectoryRoot /usr/local/share/man #undef BinDir #define BinDir /usr/local/bin/X11R5 #undef LibDir #define LibDir /usr/local/lib/X11R5 -- George D M Ross, Department of Computer Science, University of Edinburgh Kings Buildings, Mayfield Road, Edinburgh, Scotland, EH9 3JZ Mail: gdmr@dcs.ed.ac.uk Voice: 031-650 5147 Fax: 031-667 7209 From news.umbc.edu!haven.umd.edu!uunet!mcsun!sun4nl!fwi.uva.nl!casper Sun Aug 8 18:55:41 EDT 1993 Article: 2541 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!uunet!mcsun!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Subject: Re: Installed location of X11 (Was: rn/xrn ported to Solaris 2.1) Message-ID: <1993Mar22.104422.7301@fwi.uva.nl> Sender: news@fwi.uva.nl Nntp-Posting-Host: adam.fwi.uva.nl Organization: FWI, University of Amsterdam References: <1o5g6fINN6u0@tamsun.tamu.edu> <1993Mar17.152925.21662@news.duc.auburn.edu> <1993Mar18.154343.12003@ims.com> Date: Mon, 22 Mar 1993 10:44:22 GMT Lines: 12 Status: RO aad@scr.siemens.com (Anthony A. Datri) writes: >You have to jump through hoops to get a decent layout, though. By default you >get usr/*/X11 directories under it, when what you really want is >DIR/{bin,lib,include}. My preference is to build it to live in /usr/*/X11, which >are links that indirect through another link which can be changed to point at >different builds. If you want DIR/{bin,lib,include}, all you have to #define is ProjectRoot. It'll give you ProjectRoot/{bin,lib,man,include} Casper From news.umbc.edu!haven.umd.edu!darwin.sura.net!gatech!swrinde!network.ucsd.edu!munnari.oz.au!uunet!pipex!bnr.co.uk!uknet!marble.uknet.ac.uk!mcsun!sun4nl!fwi.uva.nl!casper Sun Aug 8 18:56:13 EDT 1993 Article: 2543 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!gatech!swrinde!network.ucsd.edu!munnari.oz.au!uunet!pipex!bnr.co.uk!uknet!marble.uknet.ac.uk!mcsun!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: compiling X11R5 Message-ID: <1993Mar22.125835.10167@fwi.uva.nl> Date: 22 Mar 93 12:58:35 GMT References: <1993Mar19.162917.4294@hgl.signaal.nl> Sender: news@fwi.uva.nl Organization: FWI, University of Amsterdam Lines: 17 Nntp-Posting-Host: adam.fwi.uva.nl Status: RO pvdmey@hgl.signaal.nl (P.v.d.Meij) writes: >where can I get the 'patch' file for MIT X11R5-p23 to enable compilation on >solaris 2.1. >Or should I change the 'X-server' source and a number of 'Imakefile' s my self There's a patch on export.lcs.mit.edu called ``R5.SunOS5.patch.tar.Z'' in /contrib. The new version installed today contains support for the latest R5.Xsun.multi-screen patch (and the previous one) and pure X11. Support is also included for xview3pl2. The patch should allow you to build R5 w/o problems with gcc or SunPRO C. Casper From news.umbc.edu!haven.umd.edu!uunet!psinntp!ims.com!tonyf Sun Aug 8 18:56:22 EDT 1993 Article: 2545 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!uunet!psinntp!ims.com!tonyf From: tonyf@ims.com (Antonio Freixas) Subject: Re: Installed location of X11 (Was: rn/xrn port Message-ID: <1993Mar22.152301.21173@ims.com> Sender: usenet@ims.com (USENET News Poster) Reply-To: tonyf@ims.com Organization: Integrated Measurement Systems, Inc. References: <1993Mar18.154343.12003@ims.com> <1993Mar20.031336.10449@news.duc.auburn.edu> Date: Mon, 22 Mar 1993 15:23:01 GMT Lines: 40 Status: RO I write: >This brings up an interesting question. We ship Motif programs for the Sun. We >want these programs to run under either the OpenWindows2/3 or X11R4/5 servers >using any window manager, but particulary the OpenWindows and Motif window >managers. These requirements make things somewhat touchy. > >The best way to allow the user to customize his/her X11 location is to >dynamically link in the X11 libraries. That way, the application will use the >correct local locations of things such as the XKeysymDB file and the >app-defaults directory. elling@eng.auburn.edu (Richard Elling) replies: > Perhaps one could be able to set an environment variable or X resource to point > to the XKeysymDB? Of course, reasonable default location should be provided as > the default. > > In other words, allow administrators to be able relocate code as much as possible. > Many ISVs are getting better at relocation, but many have a *long* way to go. > Openwindows version 3.0 did a very good job of allowing relocation. I've never > installed it in /usr/openwin... Of course, X environment variables are not under my control -- they are under the jurisdiction of the X consortium and are accessed not by my code but by the X toolkit. There is an XKEYSYMDB environment variable (in R5 anyway), but I've heard that it is not officially supported. Here is the result of a grep on the mit/lib/X directory: > grep -i xkeysym * ... XStrKeysym.c: dbname = getenv("XKEYSYMDB"); I don't plan to depend on this, but it may be a useful temporary solution to an installation problem. -- tonyf@ims.com (503) 626-7117 x1349 From news.umbc.edu!haven.umd.edu!uunet!titan!lpi!greg Sun Aug 8 18:57:36 EDT 1993 Article: 2558 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!uunet!titan!lpi!greg From: greg@lpi.liant.com (Greg Bean) Subject: Compiling X11R5 on Solaris 2.1 Message-ID: <1993Mar22.205842.8938@lpi.liant.com> Keywords: X11R5 Sender: news@lpi.liant.com Reply-To: greg@lpi.liant.com Organization: Liant Software Corporation, Framingham MA Date: Mon, 22 Mar 1993 20:58:42 GMT Lines: 29 Status: RO I get the following error compiling X11R5 pl19 with the "R5.SunOS5.patch.tar.Z" patch applied: cc -c -O -DSTRINGS_ALIGNED -DNO_ISDIR -I../../. -I../../. -DSVR4 -DSYSV -Dsun -Dsparc -Dunix -DNO_REGCOMP -DINCLUDE_ALLOCA_H -DNO_SETPGRP AuRead.c ...skipping... socketpair ../libphigs.a(cp_ccom.o) socket ../../../.././lib/X/libX11.so setsockopt ../../../.././lib/X/libX11.so gethostbyname ../../../.././lib/Xmu/libXmu.so gethostname ../../../.././lib/Xmu/libXmu.so send phigsmon.o connect ../../../.././lib/X/libX11.so ld: fatal: Symbol referencing errors. No output written to phigsmon ...lots more of the same... any hints would be appreciated -greg -- --------------------------------------------------------------------- Greg Bean greg@lpi.liant.com Liant Software Corporation 959 Concord Street, Framingham, Mass. 01701-4613 - 508-872-8700 x229 --------------------------------------------------------------------- From news.umbc.edu!haven.umd.edu!uunet!elroy.jpl.nasa.gov!sdd.hp.com!hplabs!unix.sri.com!updike!ric Sun Aug 8 18:57:45 EDT 1993 Article: 2562 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!uunet!elroy.jpl.nasa.gov!sdd.hp.com!hplabs!unix.sri.com!updike!ric From: ric@updike.sri.com (Richard Steinberger) Newsgroups: comp.unix.solaris,comp.windows.x.motif Subject: Solaris 2.1 and Motif 1.2.2 build issues Date: 23 Mar 1993 00:06:29 GMT Organization: sri Lines: 46 Distribution: world Message-ID: <1olka5$2lb@unix.sri.com> NNTP-Posting-Host: updike.sri.com Originator: ric@updike Xref: news.umbc.edu comp.unix.solaris:2562 comp.windows.x.motif:15510 Status: RO After being out of town for a week, I have returned to the task of trying to build Motif 1.2.2 on a SUN Solaris 2.1 system. I have tried using the supplied sun.cf, sunLib.rules, and sunLib.tmpl files as well as the mit X11R5.p22 versions of these files (from the R5.SunOS5.patch.tar.Z distribution). The mit build attempt is more successful, as SYSV and SVR4 are defined. The steps I have followed are: 1) make Makefile "BOOTSTRAPCFLAGS=-I/usr/include/X11" \ IMAKE_DEFINES=-DUseInstalled 2) cd config 3 ./imake -DUseInstalled -I. su make install [actually, copy imake to /usr/bin/X11] exit cd .. make Makefiles [All of this is based on instructions in the README-1.2.1 file.] 4 make includes This gets as far as: cc -c -O -I. -I../.././X11 -I../.././X11 -I../.././lib -I../../. -DSVR4 -DSYSV -DFUNCPROTO=15 wmlutils.c rm -f libwml.a ar cq libwml.a wmlparse.o wml.o wmloutkey.o wmlouth.o wmloutmm.o wmloutp1.o wmlresolve.o wmlsynbld.o wmlutils.o cc -O -I. -I../.././X11 -I../.././X11 -I../.././lib -I../../. -DSVR4 -DSYSV -DFUNCPROTO=15 libwml.a -o wml Undefined first referenced symbol in file _mcount libwml.a(wmlparse.o) I would be interested in hearing from people who have successfully built motif 1.2.2 on Solaris 2.1 machines with SUN C (what I tried) or gcc. I need to know what files you modified from the 1.2.2 distribution [the specific changes you made], and what steps you used in the build process. Thanks in advance for any hints/suggestions or complete solutions. ric steinberger ric@updike.sri.com From news.umbc.edu!haven.umd.edu!uunet!mcsun!Germany.EU.net!urmel.informatik.rwth-aachen.de!uni-paderborn.de!news!futzi Sun Aug 8 18:58:11 EDT 1993 Article: 2572 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!uunet!mcsun!Germany.EU.net!urmel.informatik.rwth-aachen.de!uni-paderborn.de!news!futzi From: futzi@uni-paderborn.de (Michael Kutzner) Newsgroups: comp.unix.solaris Subject: Re: Compiling X11R5 on Solaris 2.1 Date: 23 Mar 93 08:29:44 Organization: Uni Paderborn FB 17 Informatik GERMANY Lines: 22 Distribution: inet Message-ID: References: <1993Mar22.205842.8938@lpi.liant.com> NNTP-Posting-Host: solaris-serv.uni-paderborn.de In-reply-to: greg@lpi.liant.com's message of Mon, 22 Mar 1993 20:58:42 GMT Status: RO >>>>> About Compiling X11R5 on Solaris 2.1 , Greg Bean said: Greg> I get the following error compiling X11R5 pl19 with the Greg> "R5.SunOS5.patch.tar.Z" patch applied: Greg> ...lots more of the same... You have to link with libnsl.. and libsocket... (-lnsl -lsocket). For this define ExtraLibraries = -lsocket -lnsl in your sun.cf. Greg> any hints would be appreciated Greetings, Michael *************************************************************************** ** --------------------------------------------------------------------- ** ** Michael Kutzner ** X-Betreuung ** ** ** ** ** --------------------------------------------------------------------- ** *************************************************************************** From news.umbc.edu!haven.umd.edu!darwin.sura.net!spool.mu.edu!olivea!koriel!male.EBay.Sun.COM!exodus.Eng.Sun.COM!lucknow!mukesh Mon Aug 16 09:00:00 EDT 1993 Article: 6303 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!spool.mu.edu!olivea!koriel!male.EBay.Sun.COM!exodus.Eng.Sun.COM!lucknow!mukesh From: mukesh@lucknow.Eng.Sun.COM (Mukesh Kacker) Newsgroups: comp.unix.solaris Subject: Re: Sparc10 MP with Solaris 2.2 frequent cr Date: 10 Aug 1993 00:21:12 GMT Organization: Sun Microsystems, Inc. Lines: 32 Distribution: world Message-ID: References: <2447g0$q4d@babbage.ece.uc.edu> Reply-To: mukesh@lucknow.Eng.Sun.COM NNTP-Posting-Host: lucknow Status: RO In article q4d@babbage.ece.uc.edu, montjoy@ra.ece.uc.edu (Robert Montjoy) writes: >Hi. > >In article adap@news.cs.andrews.edu (Edsel Adap) writes: >> Anybody have the same problem. Anybody have >>anyc clue how to fix it? > >Start by installing the following patches: > >100999-21 >101027-04 > >There are others but these seem to be the most important. > >You can get these patches from thor.ece.uc.edu in the directory >/pub/sun-faq/Solaris2.2-patches > > There are several other patches which are important to install. Sun customer support classifies some patches as "mandatory" based on their experience and you should get a hold of that list. The patches at this ftp site are somewhat behind the official Sun patches so I would recommend that if your site has access to Sun support, you get the list of latest mandatory patches and install atleast those. - Mukesh Kacker Sunsoft Inc. From news.umbc.edu!haven.umd.edu!darwin.sura.net!spool.mu.edu!olivea!koriel!west.West.Sun.COM!news2me.EBay.Sun.COM!cronkite.Central.Sun.COM!rmtc.Central.Sun.COM!zepf Mon Aug 16 09:02:37 EDT 1993 Article: 6324 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!spool.mu.edu!olivea!koriel!west.West.Sun.COM!news2me.EBay.Sun.COM!cronkite.Central.Sun.COM!rmtc.Central.Sun.COM!zepf From: zepf@rmtc.Central.Sun.COM (Tom Zepf) Subject: Re: Emacs 19.17 on solaris 2.x Message-ID: <1993Aug10.145315.21131@rmtc.Central.Sun.COM> Organization: Sun Microsystems, Inc. References: <1993Aug4.103333.20426@infodev.cam.ac.uk> <23oec2$a8@mercury.cs.waikato.ac.nz> <5885@calmasd.Prime.COM> Date: Tue, 10 Aug 93 14:53:15 GMT Lines: 55 Status: RO In <5885@calmasd.Prime.COM> gdh@calmasd.Prime.COM (Gerald Hall) writes: >In article wisner@well.sf.ca.us (Bill Wisner) writes: >>>Anyone get the impression emacs 19 is not *quite* ready for Solaris 2.2 >>>yet? No offense Richard et al - I'm not suggesting you're sleeping nights. >> >>Running flawlessly here. >Details please. I am running 2.2 with the SunPro (unbundled) (patched to [ deleted ] I find the following shell script is all I need to configure for Solaris. Note that I do not include any UCB stuff in my path, though I don't know if that makes any difference. As you can see, all I really do is define the CPP environment variable to point in the right place, and I use the SunPro compilers, not GCC. --------------------------- cut here ------------------------------------ #!/bin/sh -x CPP=/usr/ccs/lib/cpp export CPP #dot_h=src/s/sol2.h #egrep 'SYSTEM_MALLOC' $dot_h > /dev/null #case $? in #1) if [ ! -f ${dot_h}.orig ]; then # cp ${dot_h} ${dot_h}.orig # fi # chmod +w $dot_h # echo "#define SYSTEM_MALLOC 1" >> $dot_h # ;; # #0) ;; #*) echo "egrep egrep 'SYSTEM_MALLOC' src/s/sol2.h failed" # exit 1 # ;; #esac ./configure sparc-sun-solaris2.2 \ --with-x11 \ --x-includes=${OPENWINHOME}/include \ --x-libraries=${OPENWINHOME}/lib \ -with-gcc=no for file in config.status Makefile src/config.h; do echo $file if [ ! -f ${file}.orig ]; then cp ${file} ${file}.orig fi chmod +w $file sed -e 's/^CFLAGS=.*$/CFLAGS=-O/' < $file > /tmp/XXX cp /tmp/XXX $file done From news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!magnus.acs.ohio-state.edu!usenet.ins.cwru.edu!agate!doc.ic.ac.uk!pipex!uunet!mcsun!sun4nl!fwi.uva.nl!casper Mon Aug 16 09:03:06 EDT 1993 Article: 6326 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!magnus.acs.ohio-state.edu!usenet.ins.cwru.edu!agate!doc.ic.ac.uk!pipex!uunet!mcsun!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: Emacs 19.17 on solaris 2.x Date: 10 Aug 1993 15:35:32 GMT Organization: FWI, University of Amsterdam Lines: 35 Distribution: inet Message-ID: <248f84$6if@mail.fwi.uva.nl> References: <1993Aug4.103333.20426@infodev.cam.ac.uk> <23oec2$a8@mercury.cs.waikato.ac.nz> <5885@calmasd.Prime.COM> NNTP-Posting-Host: adam.fwi.uva.nl Status: RO gdh@calmasd.Prime.COM (Gerald Hall) writes: >In article wisner@well.sf.ca.us (Bill Wisner) writes: >>>Anyone get the impression emacs 19 is not *quite* ready for Solaris 2.2 >>>yet? No offense Richard et al - I'm not suggesting you're sleeping nights. >> >>Running flawlessly here. >Details please. I am running 2.2 with the SunPro (unbundled) (patched to >Patch-ID# 100961-02) C compiler and the BSD compatibility package installed. >I have made a "first pass" at several approaches (with and without the >solaris_2_1 'hints, and manually trying to exclude the ucb stuff) - to no >avail (starting from a new extraction from the .tar file in each case). >I have cleaned up the '==' problems in Configure and the spaces between >the '-' and the option chars in the Makefile. I get lots of "warnings" >from the compiler and then different fatal errors in each case. Trick you'll need for SunPRO C: env 'CC=cc -Xs' ./configure No LD_LIBRARY_PATH And don't use ``/usr/ucb/cc''. Usually, I compile with gcc but I tried it for this once with: env 'CC=cc -Xs' ./configure --run-in-place (making emacs only takes 10 minutes, so I thought ...) Builds and runs fine. Casper From news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!cs.utexas.edu!swrinde!elroy.jpl.nasa.gov!decwrl!uunet!mcsun!sun4nl!fwi.uva.nl!casper Mon Aug 16 09:04:34 EDT 1993 Article: 6337 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!cs.utexas.edu!swrinde!elroy.jpl.nasa.gov!decwrl!uunet!mcsun!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: How to make extra pts devices? Date: 10 Aug 1993 19:12:35 GMT Organization: FWI, University of Amsterdam Lines: 21 Distribution: inet Message-ID: <248rv3$age@mail.fwi.uva.nl> References: NNTP-Posting-Host: adam.fwi.uva.nl Status: RO matt@muddcs.claremont.edu (Matt Hughes) writes: >We are running Solaris 2.2 on a Sparc server 1000, and it seems for some reason >the default setup only creates 48 pts devices (pseudo tty slave devices). I >have changed the kernel variables max_users, npts (turns out is NOT the number >of pts devices - broke badly!), and npty (didn't break, also didn't help ...) >to no avail ... any ideas? In /etc/system: set pt_cnt = # touch /reconfigure # reboot npty has an effect on the number of old-style (BSD style) pseudo ttys. Casper From news.umbc.edu!haven.umd.edu!uunet!olivea!charnel!rat!koko.csustan.edu!nic.csu.net!eis.CalState.EDU!mark Mon Aug 16 09:04:47 EDT 1993 Article: 6338 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!uunet!olivea!charnel!rat!koko.csustan.edu!nic.csu.net!eis.CalState.EDU!mark Newsgroups: comp.unix.solaris Subject: Re: Pop3 server for solaris 2.2? Message-ID: From: mark@eis.calstate.edu (Mark Crother) Date: Tue, 10 Aug 1993 17:55:42 GMT References: <2C66EC6F.14710@news.service.uci.edu> Organization: California State University Office of the Chancellor Lines: 27 Status: RO In article <2C66EC6F.14710@news.service.uci.edu> iglesias@draco.acs.uci.edu (Mike Iglesias) writes: >Does anyone know of a source for a pop3 server that runs under Solaris 2.2? >The one that comes with the MH mail software does not compile correctly >under Solaris 2.2, and if there's another one available that works, I'd >rather use that than muck with the MH one to get it working. > > >Thanks, > >Mike Iglesias Internet: iglesias@draco.acs.uci.edu >University of California, Irvine BITNET: iglesias@uci >Office of Academic Computing uucp: ...!ucbvax!ucivax!iglesias >Distributed Computing Support phone: (714) 856-6926 Here is a possible source: Date: Tue, 13 Apr 1993 17:20:52 +0200 To: netinfo@garnet.berkeley.edu, Ilan@cinf.ucl.ac.be From: Maes@elec.ucl.ac.be (Pascal Maes) X-Sender: mp@hera.elec.ucl.ac.be Subject: popper 1.831 beta on Solaris 2.1 I have modified the sources of popper to work with solaris 2.1 The file popper-1.831beta-solaris.tar.Z is on ftp.auto.ucl.ac.be From news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!agate!doc.ic.ac.uk!uknet!nessie!news Mon Aug 16 09:10:57 EDT 1993 Article: 6373 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!agate!doc.ic.ac.uk!uknet!nessie!news From: sm@mags.ph.man.ac.uk (Steve Mowbray) Subject: Re: SparcWorks fails under 2.2. Message-ID: <1993Aug11.140047.2917@nessie.mcc.ac.uk> Sender: news@nessie.mcc.ac.uk (Usenet News System) Reply-To: sm@mags.ph.man.ac.uk Organization: University of Manchester, UK References: <2483ha$i8i@rugch4.chem.rug.nl> Date: Wed, 11 Aug 1993 14:00:47 GMT Lines: 17 Status: RO In article i8i@rugch4.chem.rug.nl, rudi@chem.rug.nl (Rudi van Drunen) writes: >After upgrading from solaris 2.1 to solaris 2.2 SparcWorks doesn't >work anymore on our system...... > >% sparcworks >ld.so.1: sparcworks: relocation error: symbol not found: XwcDrawString: >referenced in /usr/openwin/lib/libxview.so.3 Patch 100807-02. Use archie to find it. --- Steve Mowbray, Physics Department, University of Manchester, U.K. - Janet: sm@uk.ac.man.ph.g2 / Internet: sm@mags.ph.man.ac.uk - - "I have always known that cork is a funny thing" - The Carrot - From news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!pipex!sunic!ericom!exucom.exu.ericsson.se!news Mon Aug 16 09:17:26 EDT 1993 Article: 6449 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!pipex!sunic!ericom!exucom.exu.ericsson.se!news From: lmcguyb@LMC.Ericsson.SE (Guy Boudreault) Subject: Patch list for Solaris 2.2 Message-ID: <1993Aug13.174149.24240@exu.ericsson.se> Sender: news@exu.ericsson.se Nntp-Posting-Host: walrus.lmc.ericsson.se Reply-To: Guy.Boudreault@LMC.Ericsson.SE Organization: Ericsson Communication Inc. (LMC) Date: Fri, 13 Aug 1993 17:41:49 GMT X-Disclaimer: This article was posted by a user at Ericsson. Any opinions expressed are strictly those of the user and not necessarily those of Ericsson. Lines: 351 Status: RO Hi all, Here's some info about the patches for Solaris 2.2 directly from a SUN tech : Solaris 2.2 Patch Report as of 29Jul93 (Wide Document, use 'enscript -r ' to print this) ------- Totals: ------- Total Patches: 57 ------ Notes: ------ This report is generated to provide a summary list of available patches for Solaris 2.2 OS. There will be regular updates of this report to ensure visibility within Customer Service of latest 2.2 patches. Under "Mandatory Patches" are patches that are considered most important especially for large configuration MP systems. In some cases a newer version of a released patch may be listed in the "Pending Patches" section. These are listed for reference but are not counted in patch total. All pending patches listed are undergoing final testing and verification. Until they are formally released, pending patches are not generally available except to specific escalated cases. Except for pending patches, all listed patches should be available via local patch database servers or access services. Any general questions should be send to patch-manager@raid.corp. Any major customer issues related to these patches should be directed into normal support channels and escalated if necessary. ------------------ Mandatory Patches: ------------------ Patch-ID# 100985-04 Synopsis: SunOS 5.2: Jumbo patch for libc and other associated fixes BugId's fixed with this patch: 1113040 1121164 1121751 1130394 1114167 Patch-ID# 100992-02 Synopsis: SunOS 5.2: streams related panics involving local transport drivers BugId's fixed with this patch: 1121031 1110027 Patch-ID# 100999-21 Synopsis: SunOS 5.2: jumbo patch for fixes to kernel/unix & NFS driver BugId's fixed with this patch: 1119267 1121146 1123874 1124210 1126184 1129464 1129087 1113596 1123266 1130848 1127302 1122420 1117034 1115068 1131435 1133751 1123178 1126165 1123855 1132273 1131580 1123435 1122465 1132866 1123762 1134196 1134196 1127070 1103091 1137685 Patch-ID# 101014-02 Synopsis: SunOS 5.2: fixes for usr/lib/libsocket BugId's fixed with this patch: 1124510 1122704 1127186 Patch-ID# 101027-04 Synopsis: SunOS 5.2: fixes for /kernel/drv/tcp BugId's fixed with this patch: 1120065 1127396 1121957 1084913 1124247 1122464 Patch-ID# 101031-01 Synopsis: SunOS 5.2: The file descriptor limit is too low on inetd BugId's fixed with this patch: 1112111 Patch-ID# 101090-01 Synopsis: SunOS 5.2: fixes security hole in expreserve BugId's fixed with this patch: 1134063 Patch-ID# 101109-02 Synopsis: SunOS 5.2: fixes problems with ldtermclose() & ptmclose() BugId's fixed with this patch: 1130226 1133037 1129615 ---------------------------- Other Non-Mandatory Patches: ---------------------------- SunOS 5.2 Released Patch List: ------------------------------ Patch-ID# 100982-02 Synopsis: Sun0S 5.2: fixes for kernel/fs/fifofs BugId's fixed with this patch: 1126986 1116568 Patch-ID# 100983-02 Synopsis: SunOS 5.2: fixes and improvements for multi-thread application use BugId's fixed with this patch: 1114100 1117554 1125344 1134191 1134064 Patch-ID# 100991-01 Synopsis: SunOS 5.2: This provides a fix to a variety of panics induced by alm2 BugId's fixed with this patch: 1124866 1123603 1124711 1124801 Patch-ID# 100997-01 Synopsis: SunOS 5.2: Direct Xlib 2.0 dumps core BugId's fixed with this patch: 1124154 1124194 1125525 1125584 1126185 1126870 Patch-ID# 101016-01 Synopsis: SunOS 5.2: listen doesn't process all requests BugId's fixed with this patch: 1128009 Patch-ID# 101018-01 Synopsis: SunOS 5.2: patch to /kernel/drv/ip to fix default route problem BugId's fixed with this patch: 1119370 Patch-ID# 101020-01 Synopsis: SunOS 5.2: The xp_netid field is not initialized in makefd_xprt() upon creation. BugId's fixed with this patch: 1104088 Patch-ID# 101022-02 Synopsis: SunOS 5.2: fixes to rexd and NIS/NIS+ BugId's fixed with this patch: 1124497 1123372 1110837 1080235 1126829 1128078 1116175 Patch-ID# 101025-03 Synopsis: SunOS 5.2: Jumbo patch for lp fixes BugId's fixed with this patch: 1117241 1112940 1120892 1075755 1091638 1128229 1127154 1119455 1100070 1130512 1125900 1120403 1112528 1129389 1132618 1111420 1132877 1132445 1132530 1132534 1132877 1133272 1133717 1103421 1134692 1115798 1128488 1123879 1123879 1135911 1101969 1123325 Patch-ID# 101028-01 Synopsis: SunOS 5.2: system may hang due to memory leak in hsfs BugId's fixed with this patch: 1120625 Patch-ID# 101036-02 Synopsis: SunOS 5.2: The statfs() call returns incorrect f_bsize BugId's fixed with this patch: 1121173 Patch-ID# 101037-01 Synopsis: SunOS 5.2: in.ftpd does not correctly verify user's shell BugId's fixed with this patch: 1112877 Patch-ID# 101039-05 Synopsis: SunOS 5.2: Jumbo patch for login, ufs, and quota bugs BugId's fixed with this patch: 1123589 1120380 1123860 1098361 1121098 1119046 1124455 1120423 1129145 1134684 1121145 1134684 Patch-ID# 101046-03 Synopsis: SunOS 5.2: csh bugs with multibyte chars, command substitution, double quotes BugId's fixed with this patch: 1125782 1132160 1111547 Patch-ID# 101049-01 Synopsis: SunOS 5.2: access denied error when adding/deleting a client if user in sysadmin grp BugId's fixed with this patch: 1129229 Patch-ID# 101050-01 Synopsis: SunOS 5.2: sysmem routine causes hang with 2GB of phsical memory BugId's fixed with this patch: 1127771 Patch-ID# 101051-01 Synopsis: SunOS 5.2: ar core dumps if object filename over 56 characters BugId's fixed with this patch: 1130028 Patch-ID# 101052-02 Synopsis: SunOS 5.2: point patch for slow link-editing problem BugId's fixed with this patch: 1108453 Patch-ID# 101054-01 Synopsis: SunOS 5.2: tape device, FSF followed by a BSF does not work BugId's fixed with this patch: 1128481 Patch-ID# 101055-01 Synopsis: SunOS 5.2: Fix to /etc/system to support X25 and FDDI BugId's fixed with this patch: 1112510 Patch-ID# 101061-01 Synopsis: SunOS 5.2: In ksh, the syntax does not always work correctly BugId's fixed with this patch: 1126921 Patch-ID# 101062-03 Synopsis: SunOS 5.2: Various BCP runtime library fixes BugId's fixed with this patch: 1131365 1120416 1121675 1118250 1136583 1136129 Patch-ID# 101077-02 Synopsis: SunOS 5.2: sendmail: can hang; problems with metasymbol and NIS BugId's fixed with this patch: 1127546 1133654 Patch-ID# 101086-02 Synopsis: SunOS 5.2: fixes for kernel/drv/iommu & kernel/drv/sbi BugId's fixed with this patch: 1125743 1124673 Patch-ID# 101095-01 Synopsis: SunOS 5.2: fixes for the libaio library BugId's fixed with this patch: 1121430 1124799 1125120 1133572 Patch-ID# 101096-01 Synopsis: SunOS 5.2: rup output often shows localhost as well as the host's real name BugId's fixed with this patch: 1114934 Patch-ID# 101112-02 Synopsis: SunOS 5.2: find: "no action" error with parentheses; excessive messages BugId's fixed with this patch: 1106778 1089219 Patch-ID# 101115-01 Synopsis: SunOS 5.2: fixes kernel memory leak in sockmod driver BugId's fixed with this patch: 1131131 Patch-ID# 101122-02 Synopsis: SunOS 5.2: fixes for the packaging utilities BugId's fixed with this patch: 1131708 1126855 1129521 1136784 Patch-ID# 101143-01 Synopsis: SunOS 5.2: fd on SS10 does not return any result code to user. BugId's fixed with this patch: 1121433 Patch-ID# 101157-01 Synopsis: SunOS 5.2: rquotad not aware of added files BugId's fixed with this patch: 1136749 OpenWindows 3.2 Released Patch List: ------------------------------------ Patch-ID# 100989-02 Synopsis: OpenWindows 3.2: bringing up a pinned menu before menu_show() cause core dump BugId's fixed with this patch: 1126998, 1127086 Patch-ID# 101005-01 Synopsis: OpenWindows 3.2: window height calculation is incorrect for Asian locales BugId's fixed with this patch: 1119847 Patch-ID# 101006-01 Synopsis: OpenWindows 3.2: problem in Audiocontrol with window height and overlapping BugId's fixed with this patch: 1119847, 1122838 Patch-ID# 101012-01 Synopsis: OpenWindows 3.2: problem numbering the attached files in attachements window (ko,zh,zhTW) BugId's fixed with this patch: 1121681 Patch-ID# 101019-01 Synopsis: OpenWindows 3.2: Props does not search for locale-specific help. BugId's fixed with this patch: 1127961 Graphics Released Patch List: ----------------------------- Patch-ID# 100910-03 Synopsis: Solaris 2.2: XGL Runtime Library fixes (unstripped versions) BugId's fixed with this patch: 1116397,1127009,1124054,1130933,1131358,1131816,1136047,1136053 Patch-ID# 100996-01 Synopsis: Solaris 2.2, XIL 1.0: runtime multi-bug patch BugId's fixed with this patch: 1121493, 1121494, 1127152, 1127273, 1129390 Patch-ID# 101057-03 Synopsis: Solaris 2.2: XGL Runtime Library fixes BugId's fixed with this patch: 1127009,1124054,1130933,1131358,1131816,1136047,1136053 Solaris 2.2 Pending Patches (Not yet released): ----------------------------------------------- Patch-ID# T100999-22 (Preliminary Patch Only - Not Yet Released) Synopsis: SunOS 5.2: fixes to kernel/unix & NFS driver BugId's fixed with this patch: 1119267 1121146 1123874 1124210 1126184 1129464 1129087 1113596 1123266 1130848 1127302 1122420 1117034 1115068 1131435 1133751 1123178 1126165 1123855 1132273 1131580 1123435 1122465 1132866 1123762 1134196 1134196 1127070 1103091 1137685 1134886 1127446 1124135 Patch-ID# T101014-04 (Preliminary Patch Only - Not Yet Released) Synopsis: Solaris 2.2: fixes for usr/lib/libsocket BugId's fixed with this patch: 1124510 1122704 1127186 1131237 1095764 Patch-ID# T101025-05 (Preliminary Patch - Not Yet Released) Synopsis: SunOS 5.2: fixes for lp system BugId's fixed with this patch: 1117241 1112940 1120892 1075755 1091638 1128229 1127154 1119455 1100070 1130512 1125900 1120403 1112528 1129389 1132618 1111420 1132877 1132445 1132530 1132534 1132877 1133272 1133717 1103421 1134692 1115798 1128488 1123879 1123879 1135911 1101969 1123325 1137632 1122613 Patch-ID# T101039-06 (Preliminary Patch Only - Not Yet Released) Synopsis: SunOS 5.2: fixes for login & ufs quotas BugId's fixed with this patch: 1123589 1120380 1123860 1098361 1121098 1119046 1124455 1120423 1129145 1134684 1121145 1134684 1136965 Patch-ID# T101107-01 (Preliminary Patch Only - Not Yet Released) Synopsis: Solaris 2.2: Patch for libcursesw BugId's fixed with this patch: 1125680 Patch-ID# T101135-01 (Preliminary Patch Only - Not Yet Released) Synopsis: SunOS 5.2: xdm does not support xdmcp, nor local passwords BugId's fixed with this patch: 1102326 Patch-ID# T101141-01 (Preliminary Patch Only - Not Yet Released) Synopsis: SunOS 5.2: add -U option to fdformat BugId's fixed with this patch: 1135252 Patch-ID# T101148-01 (Preliminary Patch Only - Not Yet Released) Synopsis: SunOS 5.1: sunview command tool and ttysubwindows broken BugId's fixed with this patch: 1131321 Patch-ID# T101144-01 (Preliminary Patch Only - Not Yet Released) Synopsis: SunOS 5.2: fixes for usr/bin/rdist BugId's fixed with this patch: 1128709 Patch-ID# T101150-01 (Preliminary Patch Only - Not Yet Released) Synopsis: OWN 3.1: xview 3.1 jumbo patch BugId's fixed with this patch: 1129722 1122353 1122244 1068172 Patch-ID# T101153-01 (Preliminary Patch Only - Not Yet Released) Synopsis: SunOS 5.2: fixes problem where "at" and "batch" corrupt user's commands BugId's fixed with this patch: 1113572 Patch-ID# T101156-01 (Preliminary Patch Only - Not Yet Released) Synopsis: SunOS 5.2: fixes corruption problems with usr/bin/mail BugId's fixed with this patch: 1130545 Patch-ID# T101165-01 (Preliminary Patch Only - Not Yet Released) Synopsis: OWN 3.2: olwm imposes a limit on the number of submenus in a rootmenu BugId's fixed with this patch: 1127353 Patch-ID# T101168-01 (Preliminary Patch Only - Not Yet Released) Synopsis: SunOS 5.2: Patch to volume management to support WABI BugId's fixed with this patch: 1136914 --- //// (. .) -------------------------o00-(_)-00o------------------------------------- Guy Boudreault Ericsson Communications Inc. (LMC) Senior I.S. Specialist 8400 Decarie Montreal, Quebec Tel : (514) 738-8300 #2445 H4P 2N2 Fax : (514) 738-2070 e-mail : Guy.Boudreault@LMC.Ericsson.SE ------------------------------------------------------------------------- _/_/_/_/ _/_/_/ _/ _/_/_/_/ _/_/_/_/ _/_/_/_/ _/_/_/_/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/_/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/_/_/ _/_/ _/ _/ _/_/_/_/ _/_/_/_/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/_/ _/_/_/_/ _/ _/ _/ _/_/_/_/ _/_/_/_/ _/_/_/_/ _/_/_/_/ _/ _/ ------------------------------------------------------------------------- From news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!agate!dog.ee.lbl.gov!hellgate.utah.edu!cc.usu.edu!cc.usu.edu!scott Mon Aug 16 09:18:27 EDT 1993 Article: 6455 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!agate!dog.ee.lbl.gov!hellgate.utah.edu!cc.usu.edu!cc.usu.edu!scott Newsgroups: comp.unix.solaris Subject: Re: NIS+ root domain duplicate server setup - how? Message-ID: From: scott@goga.ee.usu.edu (Scott Budge) Date: 13 Aug 1993 20:19:06 GMT References: Organization: Utah State University, EE Dept., Logan, Utah Nntp-Posting-Host: goga.ee.usu.edu In-reply-to: matt@muddcs.claremont.edu's message of Fri, 13 Aug 1993 02:10:38 GMTLines: 23 Lines: 23 Status: RO In article matt@muddcs.claremont.edu (Matt Hughes) writes: Anyone have a clue how to set this up? I've been fiddling for 2 days now with no luck ... I can't seem to get the correct root directory object created on the (planned) replicate ... Thanks, Matt I had the same problem. The answer is to perform the following steps after setting up the replica for the domain directory: nismkdir -s org_dir [domainname] nismkdir -s groups_dir [domainname] /usr/lib/nis/nisping -C It started working fine after these steps. -- Scott Budge Dept. of Electrical Engineering Utah State University Logan, UT 84322-4120 Phone: (801)-750-3433 FAX: (801)-750-3054 From news.umbc.edu!haven.umd.edu!darwin.sura.net!opusc!usceast!tim Wed Aug 25 14:23:43 EDT 1993 Article: 6770 of comp.unix.solaris Newsgroups: comp.unix.solaris,comp.sys.sun.admin Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!opusc!usceast!tim From: tim@hssc.scarolina.edu (Tim White) Subject: Re: dumps exceedingly slow Message-ID: Sender: usenet@usceast.cs.scarolina.edu (USENET News System) Organization: USC Department of Computer Science References: <1993Aug20.180734.27697@news.arc.nasa.gov> <25d9bu$eh@fred.fwi.uva.nl> Date: 25 Aug 93 00:01:46 GMT Lines: 65 Xref: news.umbc.edu comp.unix.solaris:6770 comp.sys.sun.admin:19902 Status: RO janw@fwi.uva.nl (Jan Wortelboer) writes: >knight@george.arc.nasa.gov (Terri Knight -- IAS) writes: >>I am dumping a 4.1.2 system to a Sun 5 GB 8mm tape drive >>connected to a SPARCstation LX running solaris 2.2. The problem >>is that the dumps take an extremely long time to complete; >>much longer than when I was using a 2 GB 8mm tape drive. Go check out Buffer....Version 1.16 was just recently released. Here is the README( note: the factor of 5 speed up claim )...and also add the blocksize as in Jan's post... > This is a program designed to speed up writing tapes on remote tape > drives. Requirements are shared memory and locks which normally > means that these are supported in your kernel. > > Buffer has been tested under SunOS 4.0.*, SunOS 4.1.*, Solarix, HP-UX 7.0, > and Gould UTX 2.1A (sv universe). > > The program splits itself into two processes. The first process reads > (and reblocks) from stdin into a shared memory buffer. The second > writes from the shared memory buffer to stdout. Doing it this way > means that the writing side effectly sits in a tight write loop and > doesn't have to wait for input. Similarly for the input side. It is > this waiting that slows down other reblocking processes, like dd. > > I run an archive and need to write large chunks out to tape regularly > with an ethernet in the way. Using 'buffer' in a command like: > > tar cvf - stuff | rsh somebox "buffer > /dev/rst8" > > is a factor of 5 faster than the best alternative, gnu tar with its > remote tape option: > > tar cvf somebox:/dev/rst8 stuff > > We have been using buffer here at Imperial for a couple of years now > for writing tar tapes and the main system dumps. > > Thanks to Kevin Twidle for the -p and -B code. > > INSTALLATION: > Check that your kernel supports shared memory and semaphores. > A quick way to check is to build buffer and run it. > If it says "couldn't create shared memory segment" you probably > need to reconfigure and rebuild your kernel. > > To install edit the Makefile and tailor the variables to > your local systems. Then type make. > > DISCLAIMER: > This package is under the GNU GENERAL PUBLIC LICENSE! > In addtion under NO circumstances can I, or Imperial College, > be held liable for any event caused by the running or storing > of this program or its documentation. > > Lee McLoughlin. Phone: +44 71 589 5111 X 5085 > Dept of Computing, Imperial College, Fax: +44 71 581 8024 > 180 Queens Gate, London, SW7 2BZ, UK. Email: L.McLoughlin@doc.ic.ac.uk -- ============================================================================== Tim White University of South Carolina tim@otis.hssc.scarolina.edu Humanities and Social Sciences Laboratory (803)-777-7840 Columbia, S.C. 29208 From news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!yeshua.opl.com!zip.eecs.umich.edu!umn.edu!csus.edu!netcom.com!devs Wed Aug 25 14:25:07 EDT 1993 Article: 6775 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!yeshua.opl.com!zip.eecs.umich.edu!umn.edu!csus.edu!netcom.com!devs From: devs@netcom.com Subject: MOTIF Libraries and GUI Builders for SOLARIS Message-ID: Keywords: MOTIF, GUI Builder, Comparison Organization: NETCOM On-line Communication Services (408 241-9760 guest) Date: Wed, 25 Aug 1993 10:10:24 GMT Lines: 24 Status: RO I am trying to determine which third-party MOTIF library/GUI builder I should purchase for an application which I am porting from HP-UX to SOLARIS. The application has been written "by hand" (no builder-generated code). I actually have two questions: 1. Which third-party MOTIF library would you recommend from the standpoint of stability and reliability for SOLARIS? Which third- party GUI-builder would you recommend from the standpoint of functionalty and ease-of-use? I am aware of three companies providing MOTIF libraries and GUI-builders: MOTIF and UIM/X from QuestWindows MOTIF and UIM/X from Black and White Software MOTIF and IXI from Integrated Computer Solutions 2. If I enhance my application using a GUI builder (i.e. by adding dialog boxes), is it easy to integrate the automatically generated code with my existing application? Thanks for any information/recommendations you may have. I will summarize and post to the net. S. Desai From news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!butch!rapnet!miner Fri Aug 27 13:03:52 EDT 1993 Article: 6859 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!butch!rapnet!miner From: miner@Rapnet.Sanders.Lockheed.Com (Jonathan Miner x3065) Subject: Setting up a printer (for ASCII and Postscript) Message-ID: <1993Aug27.140108.5493@Rapnet.Sanders.Lockheed.Com> Organization: Sanders Associates Distribution: usa Date: Fri, 27 Aug 93 14:01:08 GMT Lines: 59 Status: RO Hi Folks! I have an Apple Laserwriter (Postscript) printer and have hooked it up to my Sparc 10, running Solaris 2.2. I would like to be able to send both ASCII (simple) and postscript files to the printer. Right now only the postscript files will print. Here is a terminal session: # lpfilter -f postprint -l Input types: simple Output types: postscript Printer types: any Printers: any Filter type: slow Command: /usr/lib/lp/postscript/postprint Options: PAGES * = -o* Options: COPIES * = -c* Options: LENGTH * = -l* Options: MODES group = -n2 Options: MODES group\=\([2-9]\) = -n\1 Options: MODES portrait = -pp Options: MODES landscape = -pl Options: MODES x\=\(\-*[\.0-9]*\) = -x\1 Options: MODES y\=\(\-*[\.0-9]*\) = -y\1 Options: MODES magnify\=\([\.0-9]*\) = -m\1 # lpstat -p -l printer lw is idle. enabled since Fri Aug 27 09:48:06 EDT 1993. available. Form mounted: Content types: postscript, simple Printer types: PS Description: Apple Laserwriter Connection: direct Interface: /usr/lib/lp/model/standard On fault: write to root once After fault: continue Users allowed: (all) Forms allowed: (none) Banner not required Character sets: Default pitch: Default page size: 80 wide 66 long Default port settings: # cat /usr/pub/ascii | /usr/lib/lp/postscript/postprint | lp -o nobanner request id is lw-44 (standard input) # cat /usr/pub/ascii | lp -o nobanner request id is lw-45 (standard input) The first lp command works, the second does not. What am I missing? Thanks in advance. -- | Jonathan Miner | | | miner@rapnet.sanders.lockheed.com | | | 603-885-3065 <> | | | BTW: I do not speak for Lockheed or anybody else. Just me! | From news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!howland.reston.ans.net!xlink.net!fauern!uni-regensburg.de!rphs1!c4080 Fri Aug 27 13:04:26 EDT 1993 Article: 6862 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!howland.reston.ans.net!xlink.net!fauern!uni-regensburg.de!rphs1!c4080 From: c4080@rphs1 (Wolfgang Ratzka t2085) Newsgroups: comp.unix.solaris Subject: Re: Setting up a printer (for ASCII and Postscript) Date: 27 Aug 1993 15:40:47 GMT Organization: University of Regensburg, Germany Lines: 42 Distribution: usa Message-ID: <25l9tvINNqm2@rrzs3.uni-regensburg.de> References: <1993Aug27.140108.5493@Rapnet.Sanders.Lockheed.Com> Reply-To: Ratzka@rphs1.physik.uni-regensburg.de NNTP-Posting-Host: rphs1.physik.uni-regensburg.de X-Newsreader: TIN [version 1.2 PL0] Status: RO Jonathan Miner x3065 (miner@Rapnet.Sanders.Lockheed.Com) wrote: : Hi Folks! : I have an Apple Laserwriter (Postscript) printer and have hooked it : up to my Sparc 10, running Solaris 2.2. I would like to be able to : send both ASCII (simple) and postscript files to the printer. Right : now only the postscript files will print. : Here is a terminal session: : : (...) : : # lpstat -p -l : printer lw is idle. enabled since Fri Aug 27 09:48:06 EDT 1993. : available. : Form mounted: : Content types: postscript, simple ^^^^^^^^^^^^^^^^^^ : Printer types: PS : Description: Apple Laserwriter : Connection: direct : Interface: /usr/lib/lp/model/standard : On fault: write to root once : After fault: continue : Users allowed: : (all) : Forms allowed: : (none) : Banner not required : Character sets: : : Default pitch: : Default page size: 80 wide 66 long : Default port settings: This tells the lp service that the printer can understand both `simple' and `postscrit' input without filtering, which is not true. -- _ | | ) |/\| \ Wolfgang Ratzka From news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!news.mentorg.com!sdl!not-for-mail Mon Aug 30 01:51:47 EDT 1993 Article: 6897 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!news.mentorg.com!sdl!not-for-mail From: tal@Warren.MENTORG.COM (Tom Limoncelli) Newsgroups: comp.unix.solaris,comp.sys.sun.admin Subject: Printing from Solaris2 Date: 28 Aug 1993 18:26:41 -0400 Organization: Mentor Graphics -- IC Group, Warren, NJ, USA Lines: 54 Distribution: inet Message-ID: <25om31$d2v@worf.Warren.MENTORG.COM> NNTP-Posting-Host: worf.warren.mentorg.com Distributions: na X-Newsreader: NN version 6.4.19 #2 Xref: news.umbc.edu comp.unix.solaris:6897 comp.sys.sun.admin:20121 Status: RO I have a postscript printer hooked up to a SunOS 4.1.3 server. I want to print to it from a Solaris 2.2 machine. When I try to send any data to this printer from the Solaris 2.2 machine, I get this error: worf[tal]% head -2 ps %!PS-Adobe-1.0 %%Creator: hog:tal (Tom Limoncelli) worf[tal]% lpr -Plw4 ps UX:lpr: ERROR: There is no filter to convert the file content. TO FIX: Use the lpstat -p -l command to find a printer that can handle the file type directly, or consult with your system administrator. Printing works fine from the SunOS server (serves an entire department). How can I get Solaris2 to realize that it's got postscript data and needs to do the (difficult... NOT) translation from postscript to postscript? Here are the commands I used to set up this printer. This is my best guess at what I'm supposed to do based on what I read in the TechNotes from Sun: setenv HOST wrnvps setenv PRINTER lw4 lpsystem -t bsd $HOST lpadmin -p $PRINTER -s $HOST lpadmin -p $PRINTER -T PS -I PS cd /etc/lp/fd lpfilter -f download -F dpost.fd lpfilter -f dpost -F dpost.fd lpfilter -f postdaisy -F postdaisy.fd lpfilter -f postdmd -F postdmd.fd lpfilter -f postio -F postio.fd lpfilter -f postior -F postior.fd lpfilter -f postmd -F postmd.fd lpfilter -f postplot -F postplot.fd lpfilter -f postprint -F postprint.fd lpfilter -f postreverse -F postreverse.fd lpfilter -f posttek -F posttek.fd reject $PRINTER disable $PRINTER accept $PRINTER enable $PRINTER Suggestions? Any help would be appreciated! Tom -- Tom Limoncelli -- tal@warren.mentorg.com (work) -- tal@plts.org (play) "Some people run 'biff' to alert them that | Disclaimer: I do not they have new email. I run '/bin/true'". | speak for Mentor Graphics. From news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!elroy.jpl.nasa.gov!decwrl!koriel!newscast.West.Sun.COM!seven-up.East.Sun.COM!newscan!newscan!gilh Wed Sep 1 14:56:10 EDT 1993 Article: 6970 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!elroy.jpl.nasa.gov!decwrl!koriel!newscast.West.Sun.COM!seven-up.East.Sun.COM!newscan!newscan!gilh From: gilh@canada.sun.com (Gil Hauer - OpCom Consultant) Subject: Re: where is bzero? In-Reply-To: rcc30@juts.ccc.amdahl.com's message of 1 Sep 93 10:20:28 GMT Message-ID: Sender: news@newscan.canada.sun.com Nntp-Posting-Host: misha Organization: Sun Microsystems, Inc. References: <11Mb02uW4fSv01@JUTS.ccc.amdahl.com> Date: Wed, 1 Sep 1993 14:01:12 GMT Lines: 30 Status: RO In article <11Mb02uW4fSv01@JUTS.ccc.amdahl.com> rcc30@juts.ccc.amdahl.com (Richard Croucher) writes: > Has anybody found a easy way around the strings.h -> string.h migration? > Does anybody know what library bzero, bcmp etc are in? Their functionality has been replaced by the mem* routines. The man page is memory(3c). For example you can use memset for bzero, memmove for bcopy etc. Migration to these routines is pretty easy. > I have Solaris 2.2 with the SunPro C compiler and ld is not picking up bzero from > anywhere. I know that its used within the kernel so it must be around somewhere. The fact that a function is part of the kernel is irrelevant in application-land. Applications link with libraries in order to resolve symbol names. When a kernel function needs to be called a system call is issued. > Thanks for you help. You're welcome. Gil -- Gil Hauer ,__@ OPCOM: Solaris 2.x Migration Support _-\_<, Gil.Hauer@Canada.Sun.Com (*)/'(*) =-=-=-=-=-=-= From news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!olivea!koriel!male.EBay.Sun.COM!jethro.Corp.Sun.COM!gap!jrt Thu Sep 2 13:45:54 EDT 1993 Article: 7008 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!olivea!koriel!male.EBay.Sun.COM!jethro.Corp.Sun.COM!gap!jrt From: jrt@gap.Eng.Sun.COM (Joel Tornatore) Newsgroups: comp.unix.solaris Subject: Re: account without a password Date: 2 Sep 1993 00:18:19 GMT Organization: Sun Microsystems, Inc. Lines: 12 Distribution: world Message-ID: <263e4b$gnu@jethro.Corp.Sun.COM> References: <25v14g$snl@fuhainf.fernuni-hagen.de> Reply-To: jrt@gap.Eng.Sun.COM NNTP-Posting-Host: gap.eng.sun.com Status: RO >does anyone know how to install an account without a pasword-prompting >under solaris 2.2 ? >erasing the password-field in /etc/passwd or shadowfile is not enough. If you are trying to make an account that never needs a password, you need to comment out this line in /etc/default/login: PASSREQ=YES Setting it to NO does not work. You need to comment it out with a # as the first character. joel From news.umbc.edu!haven.umd.edu!uunet!well!wisner Tue Sep 7 12:28:35 EDT 1993 Article: 7053 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!uunet!well!wisner From: wisner@well.sf.ca.us (Bill Wisner) Subject: list of mandatory Solaris 2.2 patches Organization: The Whole Earth 'Lectronic Link, Sausalito, CA Date: Fri, 3 Sep 1993 01:59:14 GMT Message-ID: Sender: news@well.sf.ca.us Nntp-Posting-Host: satori.well.sf.ca.us Lines: 37 Status: RO I've received many requests to post the list of mandatory patches for Solaris 2.2. All these patches can be FTPed from thor.ece.uc.edu or obtained from Sun support. Patch-ID# 100985-04 Synopsis: SunOS 5.2: Jumbo patch for libc and other associated fixes BugId's fixed with this patch: 1113040 1121164 1121751 1130394 1114167 Patch-ID# 100992-02 Synopsis: SunOS 5.2: streams related panics involving local transport drivers BugId's fixed with this patch: 1121031 1110027 Patch-ID# 100999-21 Synopsis: SunOS 5.2: jumbo patch for fixes to kernel/unix & NFS driver BugId's fixed with this patch: 1119267 1121146 1123874 1124210 1126184 1129464 1129087 1113596 1123266 1130848 1127302 1122420 1117034 1115068 1131435 1133751 1123178 1126165 1123855 1132273 1131580 1123435 1122465 1132866 1123762 1134196 1134196 1127070 1103091 1137685 Patch-ID# 101014-02 Synopsis: SunOS 5.2: fixes for usr/lib/libsocket BugId's fixed with this patch: 1124510 1122704 1127186 Patch-ID# 101027-04 Synopsis: SunOS 5.2: fixes for /kernel/drv/tcp BugId's fixed with this patch: 1120065 1127396 1121957 1084913 1124247 1122464 Patch-ID# 101031-01 Synopsis: SunOS 5.2: The file descriptor limit is too low on inetd BugId's fixed with this patch: 1112111 Patch-ID# 101090-01 Synopsis: SunOS 5.2: fixes security hole in expreserve BugId's fixed with this patch: 1134063 Patch-ID# 101109-02 Synopsis: SunOS 5.2: fixes problems with ldtermclose() & ptmclose() BugId's fixed with this patch: 1130226 1133037 1129615 -- Bill Wisner Gryphon Gang Mountain View CA 94039-0386 From news.umbc.edu!haven.umd.edu!darwin.sura.net!gatech!howland.reston.ans.net!vixen.cso.uiuc.edu!uwm.edu!rpi!ghost.dsi.unimi.it!37.1!dei.unipd.it!usenet Tue Sep 7 12:29:49 EDT 1993 Article: 7066 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!gatech!howland.reston.ans.net!vixen.cso.uiuc.edu!uwm.edu!rpi!ghost.dsi.unimi.it!37.1!dei.unipd.it!usenet Newsgroups: comp.unix.solaris Subject: Re: list of mandatory Solaris 2.2 patches Message-ID: From: jake@maya.dei.unipd.it (Luca Polo) Date: Fri, 3 Sep 1993 08:25:56 GMT Reply-To: jake@maya.dei.unipd.it Sender: usenet@dei.unipd.it References: Organization: DEI, Universita' di Padova Lines: 54 Status: RO In article 6Fu@well.sf.ca.us, wisner@well.sf.ca.us (Bill Wisner) writes: >I've received many requests to post the list of mandatory patches for >Solaris 2.2. All these patches can be FTPed from thor.ece.uc.edu or >obtained from Sun support. > Well, I have applied a number of patches on Solaris 2.2, but there are some minor differences in some of the patch-IDs (see below). BTW, what are the meaning of the numbers in the patch-IDs ? It may sound strange, but I never worried about this topic before. I also imagine that XXXX-YY means "patch # XXXX, release YY", but I may be wrong. >Patch-ID# 100999-21 >Synopsis: SunOS 5.2: jumbo patch for fixes to kernel/unix & NFS driver > I don't have 100999-21, but 100999-22 (same synopsis) >Patch-ID# 101014-02 >Synopsis: SunOS 5.2: fixes for usr/lib/libsocket > I have 101014-04 Synopsis: SunOS 5.2: libsocket problems with signals, too many open files, non-blocking I/O >Patch-ID# 101027-04 >Synopsis: SunOS 5.2: fixes for /kernel/drv/tcp > This is marked to be obsoleted (on my patch-list) by 101018-02 Synopsis: SunOS 5.2: Jumbo patch for ip and tcp fixes There's no difference in the rest of the list. For more information, here's the list of the patches I applied on my system: 100982-02 100985-04 100992-02 100999-22 101014-04 101018-02 101025-05 101031-01 101039-06 101090-01 101109-02 (BTW: this list has been recently published in this newsgroup, but I can't remember who posted it: Thank you very much, anyway). >Bill Wisner Gryphon Gang Mountain View CA 94039-0386 Luca Polo. --- ______Luca Polo___System administrator___Universita` di Padova___Italia________ (opinions expressed above are only mine, though I can't exclude that somewhere in the Universe someone (something ?) is crazy enough to share them...) ^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^ From news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!europa.eng.gtefsd.com!uunet!mcsun!sun4nl!fwi.uva.nl!casper Tue Sep 7 12:30:23 EDT 1993 Article: 7070 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!europa.eng.gtefsd.com!uunet!mcsun!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: list of mandatory Solaris 2.2 patches Date: 3 Sep 1993 11:10:59 GMT Organization: FWI, University of Amsterdam Lines: 19 Distribution: world Message-ID: <2678o3$es4@mail.fwi.uva.nl> References: NNTP-Posting-Host: adam.fwi.uva.nl Status: RO jake@maya.dei.unipd.it (Luca Polo) writes: >Well, I have applied a number of patches on Solaris 2.2, but there are some >minor differences in some of the patch-IDs (see below). >BTW, what are the meaning of the numbers in the patch-IDs ? It may sound strange, >but I never worried about this topic before. I also imagine that XXXX-YY means >"patch # XXXX, release YY", but I may be wrong. Quite right. Instead of issuing new patches all the time, Sun issues patches with a release number. If you request patch #xxxx from Sun, you'll get xxxx-yy with the highest currently available yy. The mandatory patchlist lists the patches current at the time of issue. New patches and new versions will be issued all the time. Use the version number of a patch to determine the lowest version number that you should install. Casper From news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!uunet.ca!mitra!larry Tue Sep 7 12:31:31 EDT 1993 Article: 7087 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!uunet.ca!mitra!larry From: larry@mitra.mitra.com (Larry Williamson) Subject: Re: Growing /var/adm/wtmpx file In-Reply-To: tpb@fozzy.fluent.com's message of Thu, 2 Sep 93 17:28:25 GMT Message-ID: Sender: larry@mitra.com (Larry Williamson) Organization: Mitra Imaging Inc. References: <1993Sep2.172825.17793@Fluent.COM> Date: Fri, 3 Sep 1993 20:58:34 GMT Lines: 42 Status: RO Almost certainly, sometime since you installed the OS, you have given your machine a new inet number. This number now no longer agrees with the corresponding numbers in /etc/saf/tcp/_pmtab. You should find lines similiar to these in the _pmtab file... lp::root:reserved:reserved:reserved:::p::/var/spool/lp/fifos/listenS5# lpd::root:reserved:reserved:reserved:\x00020203c0be4b270000000000000000::p::/var/spool/lp/fifos/listenBSD# 0::root:reserved:reserved:reserved:\x00020ACEc0be4b270000000000000000::c::/usr/lib/saf/nlps_server# The two long lines of hex numbers include your inet number encoded in hex. In my case the 'c0be4b27' is my inet number of 192.190.75.39 192 == c0 190 == be 75 == 4b 39 == 27 do a similiar conversion on your inet numbers, then *carefully* replace the corresponding digits in these two lines. Make certain that you don't delete too much, or add in too much, ("wc -c" should report the same before and after). Note: the conversion is simple if you use the command "dc" then set dc's output to hex.... dc 16 o 192 p 190 p 75 p 39 p ^D is what I did. Once you have fixed this file, you will have to reboot. I don't remember if you have to delete the ?tmpx files or not. -Larry From news.umbc.edu!haven.umd.edu!darwin.sura.net!gatech!kd4nc!vdbsan!wdawson Tue Sep 7 12:32:34 EDT 1993 Article: 7090 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!gatech!kd4nc!vdbsan!wdawson From: wdawson@vdbsan (Willard Dawson) Subject: Re: Mandatory patches for 2.2 Message-ID: <1993Sep3.210948.15100@vdbsan> Keywords: n Reply-To: wdawson@hobbes.atl.ga.us Organization: BellSouth Advanced Networks, Inc. References: <1993Aug27.094649.5643@uoft02.utoledo.edu> <1993Aug30.191633.10573@bnlux1.bnl.gov> Date: Fri, 3 Sep 1993 21:09:48 GMT Lines: 27 Status: RO cactus@clinton.com (L. Todd Masco) writes: >For a good time, call Sun (1-800-USA-4SUN in the US) and ask for all the mandatory > patches for Solaris 2.2. >I've never had a customer support person laugh at me quite so much... >-- >Todd Masco | "life without caution/ the only worth living >cactus@clinton.com | love for a man/ love for a woman >SysAdmin, Clinton Group, Inc. | love for the facts/ protectless" - A Rich Well, I did it, but not to the person who answers the phone. You have to establish a problem, and reach the kernel workgroup. Once you've done that, it's exceptionally easy to say "I'm running 2.1, but I'd like to install 2.2; I understand that you -really- have to install a set of mandatory patches to make it fly, and that 2.3 is the OS release to end all OS releases. Could you please have patch administration send me a tape with these mandatory patches?" Guess what... it works. -- Willard Dawson, BellSouth Advanced Networks, 1000 Holcomb Woods Pkwy Ste 412 Atlanta, GA 30076, Voice: +1 404 594 6737, UUCP: kd4nc!vdbsan!wdawson Internet -> X.400: /G=W/S=DAWSON/O=BSAN/ADMD=BELLSOUTH/C=US/@sprint.com Percent hack: wdawson%vdbsan@gatech.edu :: Standard disclaimer applies. From news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!howland.reston.ans.net!spool.mu.edu!nigel.msen.com!math.fu-berlin.de!uni-paderborn.de!news@uni-paderborn.de!futzi Wed Sep 8 11:25:15 EDT 1993 Article: 7163 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!howland.reston.ans.net!spool.mu.edu!nigel.msen.com!math.fu-berlin.de!uni-paderborn.de!news@uni-paderborn.de!futzi From: futzi@uni-paderborn.de (Michael Kutzner) Newsgroups: comp.unix.solaris Subject: Re: [Q] top compiled? Date: 07 Sep 1993 20:45:28 GMT Organization: Uni Paderborn FB 17 Informatik GERMANY Lines: 38 Distribution: world Message-ID: References: <26i2j6$4gm@mail.fwi.uva.nl> <26i43iINNgs1@rs18.hrz.th-darmstadt.de> <26i64l$320@link.ph.gmr.com> <26ifs1$h2c@vixen.cso.uiuc.edu> NNTP-Posting-Host: kreta-bb.uni-paderborn.de In-reply-to: ejk@ux2.cso.uiuc.edu's message of 7 Sep 1993 17:20:01 GMT Status: RO >>>>> About Re: [Q] top compiled?, Ed Kubaitis said: Ed> Top failed with the 'kvm_read for cpu: Error 0' after our 690MP was Ed> changed from a 4 x SPARC 2 to a 2 x 50 MHz SuperSPARC. The problem appears Ed> to be that on two processor models the processors are numbered 0 and *2*, Ed> not 0 and 1 as top expects. Attached is a hack that got top to work Ed> on our two processor machine. This was a fine hack, but if don`t set ncpus explicitely =3, it runs on one and two proc machines (I tried it). I don`t know, what getkval returns on 4 (or more) proc-machines, so ... But it returns 3 on a 2 PM and 1 on a 2 PM. So your modified hack (see below) runs fine. Greets, Michael *** machine/m_sunos5.c- Fri Jul 9 14:25:51 1993 --- machine/m_sunos5.c Sat Sep 4 10:02:33 1993 *************** *** 313,320 **** /* get the cp_time array */ for (j = 0; j < CPUSTATES; j++) cp_time[j] = 0L; ! ! for (i = 0; i < ncpus; i++) { (void) getkval (cpu_offset[i], &cpu, sizeof (struct cpu), "cpu"); for (j = 0; j < CPUSTATES; j++) --- 314,320 ---- /* get the cp_time array */ for (j = 0; j < CPUSTATES; j++) cp_time[j] = 0L; ! for (i = 0; i < ncpus; i+=2) { (void) getkval (cpu_offset[i], &cpu, sizeof (struct cpu), "cpu"); for (j = 0; j < CPUSTATES; j++) From news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!cs.utexas.edu!hermes.chpc.utexas.edu!news.utdallas.edu!news.oryx.com!xcpgzm Wed Sep 8 11:33:08 EDT 1993 Article: 7162 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!cs.utexas.edu!hermes.chpc.utexas.edu!news.utdallas.edu!news.oryx.com!xcpgzm From: xcpgzm@oryx.com (Phil Meyer) Newsgroups: comp.unix.solaris Subject: booting from 2.2 server Date: 7 Sep 1993 18:50:43 GMT Organization: Oryx Energy Lines: 65 Distribution: inet Message-ID: <26il63$fi3@lm1.oryx.com> Reply-To: phil@oryx.com NNTP-Posting-Host: archmage.oryx.com X-Newsreader: TIN [version 1.2 PL1] Status: RO We have bitten the bullet and upgraded our primary boot server to Solaris 2.2. It is a SS10 model 41, 128MB RAM, 13GB disk, SUN FDDI, Disk Suite. We installed all of the mandatory patches, including 100999-21. Output from showrev -p is appended. Here is what we have so far... 1. The 'at' command from all 4.1.2 and 4.1.3 diskless clients fail, due to a larger symptom of suid programs failing with (NFS error 13) permission denied. Programs with permissions like: -r-Sr-Sr-- or ---x--S--- will at jobs nqs fail if they reside on disks mounted from the Solaris 2.2 server 2. Performance of the SS10 has been degraded. We have poked and pried and done some tuning, but we have not found documentation describing which variables can be tuned, other than those few mentioned in the answerbook. Tuning maxusers and shmem stuff just isn't enough. I have included the lines we have added to /etc/system. Some pointers on other variables and possible values are welcome! Using nm to find them, is quite different from knowing what to set and what to set them to! 3. Disk Suite 1.0 and 2.0 cannot share state databases! We had hoped that we could simply drop our old 4.1.3 disk drive back in there and go (if we had to), but the meta partitions are at risk. I'm afraid of what would happen if I do another metadb on 4.1.3. Already, one partition containing three drives, shows only one drive from df and du, but NFS clients see the whole thing! I believe it may have been caused by trying to metainit -a using the old state databases. After that failed, I remade the state databases and tried metainit -a again. It hung. I rebooted and tried again. It now works, but I worry. 4. When the server gets loaded, diskless clients often get 'cannot get port for mount service' messages when they try to mount root. This causes the boot sequence to abort. Thanks for any pointers! ################ showrev -p Patch: 100985-04 Patch: 100992-02 Patch: 100999-21 Patch: 101014-04 Patch: 101018-01 Patch: 101027-04 Patch: 101046-03 Patch: 101090-01 Patch: 101109-02 Patch: 101055-01 ################ /etc/system set maxusers=64 set p_nbuf=4096 set ncsize=4096 set bufhwm=4096 set nhbuf=1024 set bf:bf_nbufs=4096 set kobj_map_space_len=0x100000 -- +-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-+ | Phil Meyer - The consummate UNIX analyst/SA/NA |+-=-=-=-=-=-=-=-=-=-=-=-=+ | (only cuz I eat too much) | +-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-+ From news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!protocol.com!kenw Thu Sep 9 16:44:02 EDT 1993 Article: 7203 of comp.unix.solaris Newsgroups: comp.sys.sun.wanted,comp.sys.sun.admin,comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!protocol.com!kenw From: kenw@protocol.com (Ken West) Subject: Solaris 2.2 Automounter kills performance, FIXED Message-ID: <1993Sep8.221617.21008@protocol.com> Summary: Fix found for Solaris 2.2 Automounter problem Keywords: solaris Sender: usenet@protocol.com Nntp-Posting-Host: doc Organization: Protocol Systems, Incorporated Date: Wed, 8 Sep 1993 22:16:17 GMT Lines: 32 Xref: news.umbc.edu comp.sys.sun.wanted:3637 comp.sys.sun.admin:20599 comp.unix.solaris:7203 Status: RO I'm happy to report that the problem with the Solaris 2.2 Automounter killing performance is fixed. I'm posting this because I got several responses from people with similar problems. Richard Elling of Auburn University provided the fix when he suggested that the problem was caused by the LD_LIBRARY_PATH environment variable. This path gets searched for each dynamically linked command. This wasn't a problem with SunOS 4.xx because it caches the library locations. Solaris 2.x has no such cache. Furthermore, several of the directories are now automounted where they weren't before. Overall, it killed performance on our system. So.. Get rid of LD_LIBRARY_PATH. By the way, check out the new LD_RUN_PATH which you'll probably need to set to get programs to compile correctly and find their libraries once LD_LIBRARY_PATH is gone. Again, thanks to Richard Elling for solving the nasty problem. ---------- Ken West Protocol Systems, Incorporated 8500 SW Creekside Place Beaverton, Oregon 97005-7107 (503) 526-8500 (503) 526-4200 FAX kenw@protocol.com ...uunet!protocol!kenw It's not safe out here! It's wonderous-- with things to satiate desires both subtle and gross... but it's not for the timid." --Q, "Q Who" From news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!vixen.cso.uiuc.edu!uxc.cso.uiuc.edu!paul Thu Sep 23 13:12:42 EDT 1993 Article: 7467 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!vixen.cso.uiuc.edu!uxc.cso.uiuc.edu!paul From: paul@uxc.cso.uiuc.edu (Paul Pomes - UofIllinois CCSO) Newsgroups: comp.unix.solaris Subject: Re: xv compilation on Solaris 2.2 Date: 16 Sep 1993 04:31:19 GMT Organization: University of Illinois at Urbana Lines: 60 Distribution: inet Message-ID: <278q6n$cjv@vixen.cso.uiuc.edu> References: <278lj6$2su@athena.cs.uga.edu> Reply-To: Paul-Pomes@uiuc.edu NNTP-Posting-Host: uxc.cso.uiuc.edu Status: RO dehaseth@phoenix.cs.uga.edu (James A de Haseth) writes: >I have tried to compile xv on a SPARC Classic running Solaris 2.2. >The problem is that some symbols are undefined, and I am used >to seeing these in the ucb libraries. What goes around, comes around. From: bfriesen@iphase.com (Bob Friesenhahn) Date: Mon, 26 Jul 1993 20:52:16 GMT Newsgroup: comp.unix.solaris Message-Id: Subject: Re: XV compiled for Solaris 2.1? In article , gordon@ug.cs.dal.ca (Scott King) writes: |> Has anyone had any luck compiling xv for Solaris 2.1. When I try to compile |> it, everything goes fine until I get to the last step when I get the |> following error: |> |> Undefined first referenced |> symbol in file |> sigblock xvmisc.o |> |> Does anyone know how I can get this to work? Make sure that these definitions occur in your xv.h file and define SVR4 for Solaris: /* signal macros */ #if defined(SVR4) && !defined(HPUX7) && !(defined(_AUX_SOURCE)) #define HOLD_SIG sighold(SIGALRM) /* block ALRM sig from occurring */ #define RELEASE_SIG sigrelse(SIGALRM) #define PAUSE_SIG sigpause(SIGALRM) /* sleep until ALRM signal */ #else #define HOLD_SIG sigblock(sigmask(SIGALRM)) #define RELEASE_SIG sigblock(0) #define PAUSE_SIG sigpause(0) #endif This is what I use in the Makefile: # Solaris Open Windows config: CCOPTS = -DSVR4 -DSOLARIS -O2 -funroll-loops -L/usr/openwin/lib -R/usr/openwin/lib -I/usr/openwin/include I cheat and claim that I do have the random function. I then use GNU's liberty.a library (supplied with GDB & libg++) as follows: LIBS = -lX11 $(LIBJPEG) $(LIBTIFF) -lm -liberty -lnsl BTFW that if you choose to use usleep(), you will have to modify the code since usleep takes a different size argument than the xv source assumes. Hope this helps, Bob -- Bob Friesenhahn, Interphase bfriesen@iphase.com -- The only thing that will be remembered about my enemies after they're dead is the nasty things I've said about them. --Camille Paglia From news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!math.fu-berlin.de!zib-berlin.de!news.dfn.de!scsing.switch.ch!epflnews!siisun!brossard Thu Sep 23 13:18:26 EDT 1993 Article: 7474 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!math.fu-berlin.de!zib-berlin.de!news.dfn.de!scsing.switch.ch!epflnews!siisun!brossard From: brossard@siisun.epfl.ch (Alain Brossard EPFL-SIC/SII) Newsgroups: comp.unix.solaris Subject: Re: Patch FTP site [Re: Mandatory patches for 2.2] Date: 16 Sep 1993 09:16:13 GMT Organization: Ecole Polytechnique Federale de Lausanne Lines: 56 Sender: brossard@siisun (Alain Brossard EPFL-SIC/SII) Distribution: world Message-ID: <279ast$p2u@info.epfl.ch> References: <1993Sep14.141827.31664@ucl.ac.uk> <2762ev$7m4@suntan.eng.usf.edu> <277nrs$amk@rome.software.com> Reply-To: brossard@sic.epfl.ch NNTP-Posting-Host: siisun.epfl.ch Keywords: sunsite.unc.edu Status: RO In article <277nrs$amk@rome.software.com>, mike@software.com (Michael D'Errico) writes: |> gohel@csee.usf.edu (Himanshu Gohel) writes: |> |> >The best site for Sun patches in the South-East is |> |> >sunsite.unc.edu |> |> >in /pub/sun-info/sun-fixes (please check exact directory name) |> >This has all Sun patches - Solaris 1.x and Solaris 2.x, and also |> >for PC-NFS, SunVision, Developer's Guide etc. |> |> Are these really the patches? Or some subset? I downloaded patch |> number 100985-04 (the libc jumbo patch) which is supposed to contain: The latest one is 100985-05, available from sunline.epfl.ch:pub/sun-patches/100985-05.tar.Z |> /usr/bin/jsh |> /usr/lib/rsh |> /usr/bin/sh |> /usr/include/dirent.h |> /usr/lib/libc.a |> /usr/lib/libc.so.1 |> /usr/lib/libp/libc.a |> /usr/lib/pics/libc_pic.a |> /sbin/sh |> /sbin/jsh |> |> When I unpacked the patch, it only had |> |> /usr/bin/jsh |> /usr/include/dirent.h |> /usr/lib/libc.a |> /usr/lib/libc.so.1 |> /usr/lib/libp/libc.a |> /usr/lib/pics/libc_pic.a |> /sbin/jsh |> sh or rsh are not in versions -03, -04 nor -05 of the patch, I've just checked. The fixes to sh, are probably contained in /usr/lib/libc.so.1. Meaning the problem was really with libc and was visible from sh (?). However this does NOT fix the static version of sh. Just checked the README, it does say that /usr/bin/sh and /sbin/sh are supposed to be included, I guess we should get in touch with Sun to find out what the problem is. Note also that if you use "LANG=fr", sh core dumps so easily it's not funny. I never managed to enter more than 3 lines before it core dumped! -- Alain Brossard, Ecole Polytechnique Federale de Lausanne, SIC/SII, MA-Ecublens, C.P. 121, CH-1015 Lausanne, Suisse, +41 21 693-2211 brossard@sic.epfl.ch From news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!howland.reston.ans.net!spool.mu.edu!olivea!pagesat!decwrl!tribune.usask.ca!sue!mercury.cs.uregina.ca!hameluck Thu Sep 23 13:22:07 EDT 1993 Article: 7504 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!howland.reston.ans.net!spool.mu.edu!olivea!pagesat!decwrl!tribune.usask.ca!sue!mercury.cs.uregina.ca!hameluck From: hameluck@mercury.cs.uregina.ca (Jeff Hameluck) Subject: Re: Patch FTP site [Re: Mandatory patches f Sender: news@sue.cc.uregina.ca Message-ID: <1993Sep16.214425.2183382@sue.cc.uregina.ca> Date: Thu, 16 Sep 1993 21:44:25 GMT References: <279ast$p2u@info.epfl.ch> Organization: University of Regina Lines: 16 Status: RO In article tjfs@tadtec.co.uk writes: >But there isn't a *real* Sun-approved Patch FTP site, is there? > >Tim > >--- >Tim Steele, Network Manager, tjfs@tadtec.co.uk. PGP key available via finger. > Yes but you need to have a service contract to have access to it. We have the SunSpectrum Silver Support package and we have access to the FTP patch site maintained by Sun. I don't know about any other support contracts though. Jeff Hameluck hameluck@rpl.regina.sk.ca Computer Operations Unit Regina, Saskatchewan, Canada Regina Public Library (306) 777-6057 From news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!noc.near.net!news.tufts.edu!rdorich Thu Sep 23 13:26:18 EDT 1993 Article: 7546 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!noc.near.net!news.tufts.edu!rdorich From: rdorich@jade.tufts.edu (Rob Dorich) Subject: setting up a printer Message-ID: Lines: 52 Sender: news@news.tufts.edu (USENET News System) Organization: Tufts University - Medford, MA Date: Fri, 17 Sep 1993 21:07:33 GMT Status: RO Hello netters, At this moment we have moved most of our stuff to Solaris 2.2, and would like to know how to add remote printers (for starters) completely manually. I have been able to successfully do this using admintool (wow! difficult), and the combo lpsystem, lpadmin, and request. Could some kind person tell me what files are involved/created when using lpsystem and admin? and request? I already tried to setup a remote printer by just adding what I thought was necessary, but had no luck. I think that one of the main problems was that I couldn't tell the lp service that a printer had been added. These are my notes on what I did or noticed... ---- Most of files are /etc/lp/* Systems: ss7:x:-:bsd:-:n:10:-:-: default... has default printer (hp) filter.table was changed: weird file. in /etc/lp/printers: directory per printer is created: /etc/lp/printers/*/comment Just a comment /etc/lp/printers/*/configuration: Content types: any Printer type: unknown Remote: ss7 (this is of main interest) In /var/spool/lp/tmp create a dir for each sys where there is a printer. In /var/spool/lp/requests/ create a dir for each sys where there is a printer. ---- I was hoping it would be something as easy as edting /etc/printcap :) Any help will be welcome, Thanks, Roberto -- Roberto Dorich o SKI rdorich@jade.tufts.edu Electrical Engineering /= EXTREME rdorich@athena.mit.edu Tufts University __/__, uucp: roberto@kastle.ext.tufts.edu ---------------------------------------------------------------------------- From news.umbc.edu!haven.umd.edu!darwin.sura.net!news.duc.auburn.edu!usenet Thu Sep 23 13:26:38 EDT 1993 Article: 7549 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!news.duc.auburn.edu!usenet From: elling@eng.auburn.edu (Richard Elling) Subject: Re: setting up a printer Message-ID: <1993Sep17.223557.11596@news.duc.auburn.edu> Sender: usenet@news.duc.auburn.edu (News Account) Nntp-Posting-Host: netman.eng.auburn.edu Reply-To: elling@eng.auburn.edu Organization: Auburn University Engineering References: Date: Fri, 17 Sep 1993 22:35:57 GMT Lines: 37 Status: RO In article 93Sep17170356@jade.tufts.edu, rdorich@jade.tufts.edu (Rob Dorich) writes: > Could some kind person tell me what files are >involved/created when using lpsystem and admin? and request? I >already tried to setup a remote printer by just adding what I thought >was necessary, but had no luck. I think that one of the main problems >was that I couldn't tell the lp service that a printer had been added. Simple. SERVER=... SERVER_TYPE={bsd|s5} PRINTER_NAME=... lpsystem -t $SERVER_TYPE $SERVER lpadmin -p $PRINTER_NAME -s $SERVER -I any accept $PRINTER_NAME enable $PRINTER_NAME This is *very* easy to roll into a script where you just create a list of SERVER/SERVER_TYPE/PRINTER_NAME triplets. >I was hoping it would be something as easy as edting /etc/printcap :) >Any help will be welcome, Ain't nothing as hard as editing /etc/printcap. What a piece of junk. You can blissfully forget every little typo, tab/space goofup, missing \, botched up fc and xc entry you ever made in a printcap file. With the lp system, there is no such grief. No editing files. Life is good. Of course, life will be even better when Sun gets the NIS+ printer stuff working... it's tauntingly close in SunOS 5.2... --- Richard Elling, Manager of Network Support, Engineering Administration Auburn University richard.elling@eng.auburn.edu kb4hb.ampr.org "The world is my lavatory." -- Albert Alligator From news.umbc.edu!haven.umd.edu!darwin.sura.net!news.duc.auburn.edu!usenet Thu Sep 23 13:27:07 EDT 1993 Article: 7550 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!news.duc.auburn.edu!usenet From: elling@eng.auburn.edu (Richard Elling) Subject: Re: SOLARIS2.2 : ifconf problems with full qual Message-ID: <1993Sep17.224800.11744@news.duc.auburn.edu> Sender: usenet@news.duc.auburn.edu (News Account) Nntp-Posting-Host: netman.eng.auburn.edu Reply-To: elling@eng.auburn.edu Organization: Auburn University Engineering References: Date: Fri, 17 Sep 1993 22:48:00 GMT Lines: 20 Status: RO In article 748291648@bruge.rouyn-ogc.hydro.qc.ca, ggagnon@mais.hydro.qc.ca (Guy Gagnon) writes: >I installed Solaris 2.2 on a server (IPX for tests). >At the first configuration of teh machine >When i use a hostname full qualified like >"toto.belzibuth.qc.ca" , the le0 (ethernet) >interface answer well... i can ping... Is this not in the FAQ yet? Then again I haven't seen the FAQ for a while... You need to fix the file /etc/init.d/rootusr. You will see a line that says: /sbin/ifconfig $1 $addr netmask + \ change it to: /sbin/ifconfig $1 "$addr" netmask + \ --- Richard Elling, Manager of Network Support, Engineering Administration Auburn University richard.elling@eng.auburn.edu kb4hb.ampr.org "The world is my lavatory." -- Albert Alligator From news.umbc.edu!haven.umd.edu!uunet!iphase.com!wildcat!bfriesen Thu Sep 23 13:27:28 EDT 1993 Article: 7552 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!uunet!iphase.com!wildcat!bfriesen From: bfriesen@iphase.com (Bob Friesenhahn) Subject: Re: xv compilation on Solaris 2.2 Message-ID: Sender: news@iphase.com Nntp-Posting-Host: wildcat_eng1.iphase.com Reply-To: bfriesen@iphase.com Organization: Interphase, Inc. References: <278q6n$cjv@vixen.cso.uiuc.edu> Date: Fri, 17 Sep 1993 23:40:49 GMT Lines: 43 Status: RO In article cjv@vixen.cso.uiuc.edu, paul@uxc.cso.uiuc.edu (Paul Pomes - UofIllinois CCSO) writes: > dehaseth@phoenix.cs.uga.edu (James A de Haseth) writes: > > >I have tried to compile xv on a SPARC Classic running Solaris 2.2. > >The problem is that some symbols are undefined, and I am used > >to seeing these in the ucb libraries. > > What goes around, comes around. > > From: bfriesen@iphase.com (Bob Friesenhahn) > Date: Mon, 26 Jul 1993 20:52:16 GMT > Newsgroup: comp.unix.solaris > Message-Id: > Subject: Re: XV compiled for Solaris 2.1? > > In article , gordon@ug.cs.dal.ca (Scott King) writes: > |> Has anyone had any luck compiling xv for Solaris 2.1. When I try to compile > |> it, everything goes fine until I get to the last step when I get the > |> following error: > |> > |> Undefined first referenced > |> symbol in file > |> sigblock xvmisc.o > |> > |> Does anyone know how I can get this to work? > > > This is what I use in the Makefile: > # Solaris Open Windows config: > CCOPTS = -DSVR4 -DSOLARIS -O2 -funroll-loops -L/usr/openwin/lib -R/usr/openwin/lib -I/usr/openwin/include I'd like to add that when I added the -mv8 option to the CCOPTS line when using gcc (on SPARCclassic or LX), xv performed measurably faster. Also, retrieve the xv patches from export.lcs.mit.edu in the contrib directory. The patch "xv-3.00.j1.unofficial.patch" leads to very significant speedups displaying JPEG pictures on 8 bit displays. Bob --- Bob Friesenhahn, Interphase bfriesen@iphase.com From news.umbc.edu!haven.umd.edu!nova.umd.edu!not-for-mail Thu Sep 23 13:31:36 EDT 1993 Article: 7597 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!nova.umd.edu!not-for-mail From: caruso@nova.umd.edu (John Caruso) Newsgroups: comp.unix.solaris Subject: Re: the mails in /var/mail can be read by groups! Date: 20 Sep 1993 12:27:30 -0400 Organization: University of Maryland University College Lines: 23 Distribution: inet Message-ID: <27klli$qff@nova.umd.edu> References: <27j6mvINNjcb@gap.caltech.edu> NNTP-Posting-Host: nova.umd.edu Status: RO In article <27j6mvINNjcb@gap.caltech.edu>, Oon-Gil Paik wrote: >In /var/mail directory, I manually delete my own mail file,then >send me a test mail,then this is what I get : >ls -l ogpaik >-rw-rw--- ogpaik ug ogpaik This is probably caused by incorrect protections on /usr/bin/mail. You need to set the protections to 2511 and make sure its group ownership is "mail". It may also be a problem with the flags in your local mailer definition; the default flags for Mlocal on Solaris are F=flsSDFMmnP. We've run into MANY problems with mail ownership on Solaris 2.2...for example, if you send mail locally to two users, neither of whom has a mailfile, you'll end up owning one of the resulting mailfiles (and if you send to 30 users without mailfiles, you'll own 29 of them--mail only sets the file ownership correctly for the first recipient). --------------------------------------------------------------------- John Caruso caruso@nova.umd.edu Unix/VMS System Administrator caruso@UMUC (Bitnet) University of Maryland University College Work: (301) 985-7447 --------------------------------------------------------------------- From news.umbc.edu!haven.umd.edu!uunet!news.univie.ac.at!fuw.edu.pl!cyfronet!infoserv!krzysiu Thu Sep 23 13:31:43 EDT 1993 Article: 7598 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!uunet!news.univie.ac.at!fuw.edu.pl!cyfronet!infoserv!krzysiu From: krzysiu@uci.agh.edu.pl (Krzysztof Rozycki PK) Subject: Re: Question about sendmail when not using NIS Message-ID: Sender: news@cyf-kr.edu.pl Nntp-Posting-Host: infoserv.ii.uj.edu.pl Organization: Academic Computer Centre, CYFRONET X-Newsreader: TIN [version 1.1 PL8] References: <1993Sep16.083255.16539@amoco.com> Date: Mon, 20 Sep 1993 15:11:47 GMT Lines: 19 Status: RO Scott Chasin (zsxc0i@hou.amoco.com) wrote: : > I am running Solaris 2.2 and am using DNS not NIS. I am : > having trouble getting sendmail to send and receive : > outside of my domain (or what is set by the Dm and Cm : > variables). : Is DNS working on your system? Is it resolving hosts? : If not, I had the same problem. I had to edit /etc/nsswitch.conf : and change the line "hosts: file" to "hosts: file dns" to alert : the OS to consult the hosts file first, then resolve with DNS. I'm running successfully such configration with sendmail.mx and changes in sendmail.cf throwing away all NIS dependencies from config. Krzysztof Rozycki rozycki@oeto.pk.edu.pl From news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!newsflash.concordia.ca!nstn.ns.ca!cs.dal.ca!ug.cs.dal.ca!gordon Thu Sep 23 13:33:08 EDT 1993 Article: 7609 of comp.unix.solaris Newsgroups: comp.unix.programmer,comp.windows.x,comp.sys.sun.misc,comp.unix.solaris,comp.sys.hp Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!newsflash.concordia.ca!nstn.ns.ca!cs.dal.ca!ug.cs.dal.ca!gordon From: gordon@ug.cs.dal.ca (Scott King) Subject: SPARCclassic locking up with remote applications. Message-ID: Sender: usenet@cs.dal.ca (USENET News) Nntp-Posting-Host: ug.cs.dal.ca Organization: Math, Stats & CS, Dalhousie University, Halifax, NS, Canada References: <1993Sep17.004514.9243@newsgate.sps.mot.com> Date: Mon, 20 Sep 1993 21:29:06 GMT Lines: 9 Xref: news.umbc.edu comp.unix.programmer:13264 comp.windows.x:71375 comp.sys.sun.misc:10774 comp.unix.solaris:7609 comp.sys.hp:37832 Status: RO I have a SPARCclassic running Solaris2.1 that is having a problem running remote X applications. I can get a remote application to display on the Classic but it locks up the classic after using the application for a few minutes. Killing the remote application unlocks the system. When I run X applications on the Classic there are no problems. It only happens when I run a remote application and display it on the Classic. Does anyone have any idea what the problem is? - Scott From news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!noc.near.net!newshost.williams.edu!newshost!95ssb Fri Sep 24 02:03:17 EDT 1993 Article: 7610 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!noc.near.net!newshost.williams.edu!newshost!95ssb From: 95ssb@khillari.cs.williams.edu (Stina Bridgeman) Newsgroups: comp.unix.solaris Subject: setting up sol2.2 printers (servers/clients) Date: 20 Sep 1993 21:46:41 GMT Organization: Williams College, Williamstown, MA Lines: 127 Distribution: world Message-ID: <95SSB.93Sep20174642@khillari.cs.williams.edu> NNTP-Posting-Host: khillari.cs.williams.edu Status: RO argh. has anyone successfully gotten a solaris 2.2 print server (with clients) working? we can print from our server now, but nothing else... i've tried adding the client both with printer manager and by hand (with the command lines), but to no avail. as far as i can tell, the /etc/lp/* stuff is ok (Systems, printer-name/configuration, etc). here's the output from working through the printer troubleshooting in chapter 7 of the "Setting Up User Accounts, Printers, and Mail" manual. thanks. stina bridgeman 95ssb@cs.williams.edu ---------------- (angus in the print server, khillari is the client, and bullogna is the printer (don't ask about the names... ) on angus: # lpstat -t scheduler is running system default destination: bullogna device for bullogna: /dev/term/a bullogna accepting requests since Tue Sep 14 15:47:57 EDT 1993 printer bullogna is idle. enabled since Tue Sep 14 15:47:57 EDT 1993. available. character set # tail /var/lp/logs/lpNet 09/19/93 22:33 p 158 Starting. 09/19/93 22:33 p 158 Starting lpNetParent. 09/19/93 22:33 p 158 Initialized & Polling. [these are from when angus got rebooted last night] # lpsystem -l System: + Type: s5 Connection timeout: never Retry failed connections: after 10 minutes Comment: Allow all connections System: khillari Type: s5 Connection timeout: never Retry failed connections: after 10 minutes Comment: none # sacadm -l PMTAG PMTYPE FLGS RCNT STATUS COMMAND tcp listen - 9999 ENABLED /usr/lib/saf/listen tcp # zsmon ttymon - 0 ENABLED /usr/lib/saf/ttymon # # pmadm -l PMTAG PMTYPE SVCTAG FLGS ID tcp listen lp - root - - p - /var/spool/lp/fifos/listenS5 # tcp listen lpd - root \x00020203000000000000000000000000 - p - /var/spool/lp/fifos/listenBSD # tcp listen 0 - root \x00020ACE000000000000000000000000 - c - /usr/lib/saf/nlps_server # zsmon ttymon ttya u root /dev/term/a I - /usr/bin/login - 9600 ldterm,ttcompat ttya login: - tvi925 y # zsmon ttymon ttyb u root /dev/term/b I - /usr/bin/login - 9600 ldterm,ttcompat ttyb login: - tvi925 y # on khillari: # lpstat -t scheduler is running system default destination: bullogna system for bullogna: angus bullogna accepting requests since Mon Sep 20 16:57:30 EDT 1993 printer bullogna faulted. enabled since Mon Sep 20 16:57:31 EDT 1993. available. system not responding [this here looks like the problem - any ideas?] # lpstat -p bullogna -l printer bullogna faulted. enabled since Mon Sep 20 16:57:31 EDT 1993. available. system not responding Content types: any Printer types: unknown Description: Users allowed: (all) Forms allowed: (none) Banner not required Character sets: (none) Default pitch: Default page size: # lpsystem -l System: + Type: s5 Connection timeout: never Retry failed connections: after 10 minutes Comment: Allow all connections System: angus Type: s5 Connection timeout: never Retry failed connections: after 10 minutes Comment: none # ls -l /var/spool/lp/requests/khillari total 14 -rw-rw---- 1 lp lp 53 Sep 9 17:36 1-0 -rw-rw---- 1 lp lp 53 Sep 14 15:20 19-0 -rw-rw---- 1 lp lp 53 Sep 15 10:14 34-0 -rw-rw---- 1 lp lp 54 Sep 20 10:32 45-0 -rw-rw---- 1 lp lp 55 Sep 20 10:36 56-0 -rw-rw---- 1 lp lp 46 Sep 20 16:51 74-0 -rw-rw---- 1 lp lp 54 Sep 14 14:54 9-0 # lpstat -o 74-0 UX:lpstat: ERROR: "74-0" already printed. TO FIX: You can't get any information about this request, since no record of it remains. # tail /var/lp/logs/lpNet 09/20/93 15:05 c 222 angus Could not connect to remote child. 09/20/93 15:15 c 222 angus Could not connect to remote child. 09/20/93 15:25 c 222 angus Could not connect to remote child. 09/20/93 15:35 c 222 angus Could not connect to remote child. 09/20/93 15:45 c 222 angus Could not connect to remote child. 09/20/93 16:05 c 222 angus Could not connect to remote child. 09/20/93 16:21 c 222 angus Could not connect to remote child. 09/20/93 16:31 c 222 angus Could not connect to remote child. 09/20/93 16:41 c 222 angus Could not connect to remote child. 09/20/93 16:51 c 222 angus Could not connect to remote child. From news.umbc.edu!haven.umd.edu!purdue!ames!agate!howland.reston.ans.net!newsserver.jvnc.net!newsserver.technet.sg!nuscc.nus.sg!iscyeoak Fri Sep 24 02:05:44 EDT 1993 Article: 7624 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!purdue!ames!agate!howland.reston.ans.net!newsserver.jvnc.net!newsserver.technet.sg!nuscc.nus.sg!iscyeoak From: iscyeoak@nuscc.nus.sg (YEO ANN KIAN) Newsgroups: comp.unix.solaris Subject: 4.3bsd-reno lpd/lpr for Solaris2.2 Date: 21 Sep 1993 08:46:43 GMT Organization: National University of Singapore Lines: 22 Distribution: inet Message-ID: <27mf1j$9t4@nuscc.nus.sg> NNTP-Posting-Host: nuscc.nus.sg X-Newsreader: TIN [version 1.2 PL0] Status: RO lpr-sol2.tar.gz or lpr-sol2.tar.Z is available in the following sites: sunok-wks.acs.ohio-state.edu:/pub/solaris2/src/lpr-sol2.tar.gz atlas.ce.washington.edu:/pub/lpr-sol2.tar.Z solomon.technet.sg:/pub/uploads/unix/lpr-sol2.tar.gz and another site in Germany which I have lost the address. Thank for the following people who have responded: Wilson Leong Frank G. Fiamingo David W. Barts -- Ann-Kian Yeo, yeoak@iscs.nus.sg, iscyeoak@nuscc.nus.sg From news.umbc.edu!haven.umd.edu!uunet!tcsi.tcs.com!agate!doc.ic.ac.uk!dundee.ac.uk!dundee.ac.uk!not-for-mail Fri Sep 24 02:06:13 EDT 1993 Article: 7625 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!uunet!tcsi.tcs.com!agate!doc.ic.ac.uk!dundee.ac.uk!dundee.ac.uk!not-for-mail From: pjclinch@dux.dundee.ac.uk (Pete Clinch) Newsgroups: comp.unix.solaris Subject: Re: Solaris 2.2 mandatory patch(es)???? Date: 21 Sep 1993 10:56:07 +0100 Organization: Information Technology Services , Dundee University, Scotland, UK Lines: 28 Distribution: inet Message-ID: <27mj3n$btl@dux.dundee.ac.uk> References: <1993Sep21.070727.17149@ericsson.se> NNTP-Posting-Host: dux.dundee.ac.uk X-Newsreader: TIN [version 1.2 PL1] Status: RO Stefan Frick (etxstfr@helix.ericsson.se) wrote: : Having just installed my first 2.2 system and when I popped into this : newsgroup I find a lot of articles which references the 'mandatory patch(es)'. : Please excuse if this is an FAQ, but what exactly is regarded is 'mandatory'? A letter from Sun a couple of weeks ago reads... "The UK Answer Centre strongly reccomends that te following 8 patches be applied to any Solaris 2.2 system." and the magic numbers are... 100985, 100992, 100999, 101014, 101027, 101031, 101090, 101109 Sun are saying you should have these, so I guess that makes them about as mandatory as you can get. Get in touch with Sun tech. support, and they'll send you a set, though they're a bit slow at the moment. Oddly enough, when they sent out a letter to all their support customers saying you needed the patches, there was a surge in demand for them... Pete. -- Peter Clinch University of Dundee phone 44 382 (voice) 60111 x 2050 (fax) 632970 Department of Medical Physics email pjclinch@dux.dundee.ac.uk Ninewells Hospital Dundee, DD1 9SY, Scotland, UK From news.umbc.edu!haven.umd.edu!darwin.sura.net!news.duc.auburn.edu!usenet Fri Sep 24 17:29:13 EDT 1993 Article: 7720 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!news.duc.auburn.edu!usenet From: elling@eng.auburn.edu (Richard Elling) Subject: Re: Porting to Solaris Message-ID: <1993Sep23.150849.7813@news.duc.auburn.edu> Sender: usenet@news.duc.auburn.edu (News Account) Nntp-Posting-Host: netman.eng.auburn.edu Reply-To: elling@eng.auburn.edu Organization: Auburn University Engineering References: <1993Sep20.154717.13940@ncar.ucar.edu> Date: Thu, 23 Sep 1993 15:08:49 GMT Lines: 79 Status: RO In article 13940@ncar.ucar.edu, gary@chinook.atd.ucar.edu (Gary Horton) writes: >Firstly, I'm following instructions in the document titled "Transition Planning >Guide for Application Developers". I'm building with SysV libraries and have >performed the linting process as suggested (svidii lint library). Next, I >started using the Pipeline tool to scan source code and flag problems, and >here is where I get bogged down. It seems in many cases the tool tells me >a given interface is not supported in SunOS 5.0, but doesn't tell me what >to do about it. Here is an example flag from Pipeline: Please note that the pipeline tool, while very useful, is dated. In particular, the UNIX OS API spec currently being drafted includes several of the functions that Sun moved out of the way in order to be more SVID, XPG, AS (Alphabet Soup :-) compliant. So several of the routines that pipeline 1.0 will flag as removed or obsolete, will reappear when Solaris 2 becomes UNIX OS API compliant. >*********************************************************** >PROBLEM SYMBOL: gethostbyname(3N) > >MESSAGE: > In SunOS 4.x.y/sysV: > This interface is not available in SystemV Release 3/svidII, > however it may be functional under 4.x.y/sysV. > > In SunOS 5.0: > This interface has same functionality under both SunOS 4.x.y > and SunOS 5.0. > > In SPARC ABI: > This interface is not available in SPARC ABI. Here is a prime example. One of the efforts behind SVR4 development was the concept of transport independent networking. Back when OSI was "imminent" :-) it made more sense to work hard to support it transparently. IMHO this is still a valid design goal even though OSI has faded: Novell Netware or Appletalk support may be handy for many developers. As a result, the Solaris 2.2 man pages recommend using netdir_getbyname() which is part of a generic transport name-to-address lookup. However, from what I can tell, netdir_getbyname() is not part of the UNIX OS API. What to do? >(2) Are the BSD socket calls (i.e., socket(), sendto(), bind(), etc.) still >supported under Solaris 2.x? If not, what is the wisest replacement - RPC? >TLI? something else? BSD sockets will be supported forever. The BSD-Reno versions appear in the UNIX OS API. With the growing popularity of Winsock, BSD socket style programming is cast in stone. >(3) There are many pipeline message where something is listed as unsupported and no >further guidance is given. For example (well, one possible example is the above >output, but it seems ambiguous so I'm uncertain...), here's more output from Pipeline: >*************************************************** > >PROBLEM SYMBOL: SIGUSR1, SIGUSR2 (signal.h) > >MESSAGE: > This symbol is not defined in System V.3. > >*************************************************** > >What is the replacement? Where do I look in general for ambiguities or deadends >like this? Is there some document somewhere that lists all library calls and >their Solaris counterparts? Would answerbook on Solaris be of some help here? Check the man pages. Also check the Solaris porting guides and FAQ. Signal is effectively replaced by sigaction() somewhere in POSIX or another AS. >(4) If I port an application directly to Solaris 2.x, i.e. compile on that >OS using whatever is needed to complete the port, will my app still be source-code >compatible with SunOS 4.1.x? It depends on how you write it :-). --- Richard Elling, Manager of Network Support, Engineering Administration Auburn University richard.elling@eng.auburn.edu kb4hb.ampr.org "The world is my lavatory." -- Albert Alligator From news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!spool.mu.edu!uwm.edu!linac!uchinews!iitmax!iitmax.acc.iit.edu!ldonahue Fri Sep 24 17:32:43 EDT 1993 Article: 7739 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!spool.mu.edu!uwm.edu!linac!uchinews!iitmax!iitmax.acc.iit.edu!ldonahue From: ldonahue@chicagokent.chicagokent.Kentlaw.EDU (Larry Donahue) Subject: A core dumping /usr/lib/sendmail Message-ID: Sender: news@iitmax.iit.edu (News) Organization: Chicago-Kent College of Law, 565 W. Adams St., Chicago, Il. 60661 Date: 23 Sep 1993 22:46:59 GMT Lines: 75 Status: RO Hello. Just wanted to run this problem by everyone and see if they are experiencing it too... I have Solaris v2.2 with all the manditory patches. My /usr/lib/sendmail core dumps every once in a while. Maybe once every two days, approximately. I've talked with Sun for weeks. At first, my sendmail was core dumping approximately 5 times/day. A guy at Sun helped correct the problem by giving me a new main.cf and we removing the old /etc/mail/sendmail.cf. I was told that you're really not supposed to modify that main.cf file, and core dumps were likely with too much modification. So, I had a new sendmail.cf. The problem is, it still core dumped every once in a while. Also, it had no BITNET capability. After much complaining, Sun gave me the following line to add in my sendmail.cf file for BITNET forwarding: R$*<@$*.bitnet>$* $#ddn $@$cunyvm.CUNY.EDU $:$1<@$2.bitnet>$3 The problem is, my sendmail IS STILL CORE DUMPING! Sun was good enough to send me a unstripped version of /usr/lib/sendmail.mx. The following is the result of dbx on sendmail: root@chicagokent[3667] cd /var/spool/mqueue root@chicagokent[3668] dbx /usr/lib/sendmail Reading symbolic information for /usr/lib/sendmail corefile read successfully Reading symbolic information for rtld /usr/lib/ld.so.1 Reading symbolic information for /usr/lib/libsocket.so.1 Reading symbolic information for /usr/lib/libnsl.so.1 Reading symbolic information for /usr/lib/libresolv.so.1 Reading symbolic information for /usr/lib/libelf.so.1 Reading symbolic information for /usr/lib/libc.so.1 Reading symbolic information for /usr/lib/libdl.so.1 Reading symbolic information for /usr/lib/libintl.so.1 Reading symbolic information for /usr/lib/libw.so.1 Reading symbolic information for /usr/lib/nswfiles.so Reading symbolic information for /usr/lib/switch.so Reading symbolic information for /usr/lib/nswdns.so program terminated by signal SEGV (segmentation violation) warning: can't find object file /usr/patches/sendmail/src/savemail.o warning: can't find object file /usr/patches/sendmail/src/envelope.o warning: can't find object file /usr/patches/sendmail/src/main.o warning: can't find object file /usr/patches/sendmail/src/srvrsmtp.o (dbx) where returntosender(0x1, 0x0, 0x1, 0x6f1c0, 0x0, 0xefffe98c) at 0x1de10 savemail(0x59c50, 0x49c8c, 0x5d960, 0x7efefeff, 0x81010100, 0xff00) at 0x1d7b8 dropenvelope(0x59c50, 0x0, 0x5b138, 0x6e, 0x21, 0x0) at 0x2e238 finis(0x4a408, 0x4a40c, 0xeffffdc5, 0x48d84, 0xeffff3d0, 0x48d84) at 0x17020 smtp(0x59c50, 0x9, 0x0, 0x59800, 0x5cb30, 0xeffff3d0) at 0x28320 main(0x73, 0x5b000, 0x73, 0x59000, 0x0, 0x5b2e0) at 0x16d48 (dbx) quit Notice the SEGV? At any rate, my first question/comment is: Is a SEGV likely with sendmail.cf hacking? I was under the firm impression that if anything SEGV's, it's bad and it's a bug. A SEGV usually represents a haywire pointer which should be controlled in a well written program. Sun insists that is normal. What do you think? Second, any ideas on what may cause this? Is it truely a bug, or is there something I might check on my system? Any comments/suggestions/ideas appreciated. Thanks! Larry Donahue. -- Laurence S. Donahue Student/Staff of Chicago-Kent College of Law ldonahue@chicagokent.Kentlaw.EDU (312)906-5313 Stop searching. Happiness is right next to you. Now, if they'd only take a bath ... From news.umbc.edu!haven.umd.edu!uunet!decwrl!pa.dec.com!aberdeen.ac.uk:george@bio-medical-physics.aberdeen.ac.uk! Fri Sep 24 17:37:28 EDT 1993 Article: 7762 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!uunet!decwrl!pa.dec.com!aberdeen.ac.uk:george@bio-medical-physics.aberdeen.ac.uk! From: george@bio-medical-physics.aberdeen.ac.uk ("g.cameron") Message-ID: <9309241225.AA08230@biomed.abdn.ac.uk> Subject: printing from BSD to Solaris Date: Fri, 24 Sep 93 13:25:00 BST X-Received: by usenet.pa.dec.com; id AA13317; Fri, 24 Sep 93 05:29:02 -0700 X-Received: by inet-gw-1.pa.dec.com; id AA14172; Fri, 24 Sep 93 05:28:59 -0700 X-Via: uk.ac.aberdeen; Fri, 24 Sep 1993 13:28:07 +0100 X-Received: from biomed.abdn.ac.uk (athena.biomed.abdn.ac.uk) by aberdeen.ac.uk; Fri, 24 Sep 93 13:23:59 BST X-Received: from apollo.biomed.abdn.ac.uk by biomed.abdn.ac.uk (4.1/SMI-4.1) id AA08230; Fri, 24 Sep 93 13:25:00 BST X-To: comp.unix.solaris.usenet, nijhof@nl.rug.th X-Content-Length: 1817 Lines: 59 Status: RO In article 5cp@rugch4.chem.rug.nl, nijhof@th.rug.nl (Jeroen Nijhof) writes: >How can I print from BSD-machine to a Sun running Solaris 2.2? >the BSD-machine at the moment is 'waiting for xxx to wake up' > >One thing seems fishy to me, `/usr/sbin/lpsystem -A' answers >00020203000000000000000000000000: I cannot recognize the TCP/IP >address in this. > >(The question is in the Solaris2-FAQ, but unfortunately it's answer is >`you're on your own'.) Hope I'm doing the right thing in posting this, as I'm very far from being an expert in this area. This is the shell script I put together from reading the manuals, and it did seem to do the job OK. You may need to use % pmadm -r -p tcp -s lp (and the same for lpd) if you already have it set up. Anyway - here's my script .. #!/bin/sh export PATH PATH=/bin:/usr/bin:/usr/ucb:/usr/sbin echo "Setting up print server ..." echo "" sacadm -a -p tcp -t listen -c "/usr/lib/saf/listen tcp" -v `nlsadmin -V` -n 9999 sacadm -l HEAD="`lpsystem -A | dd bs=4 count=1 2> /dev/null`" TAIL="`lpsystem -A | dd bs=8 skip=1 2> /dev/null`" ADDR="${HEAD}0203${TAIL}" MODADDR="${HEAD}0ACE${TAIL}" pmadm -a -p tcp -s lp -i root -m `nlsadmin -o /var/spool/lp/fifos/listenS5` -v `nlsadmin -V` pmadm -a -p tcp -s lpd -i root -m `nlsadmin -o /var/spool/lp/fifos/listenBSD -A "\x${ADDR}"` -v `nlsadmin -V` pmadm -a -p tcp -s 0 -i root -m `nlsadmin -c /usr/lib/saf/nlps_server -A "\x${MODADDR}"` -v `nlsadmin -V` echo "" echo "Checking log messages ..." echo "" cat /var/saf/tcp/log echo "" echo "Done." --------------------------------------------------------------------------- George Cameron g.cameron@biomed.abdn.ac.uk Dept. BioMedical Physics Aberdeen University Foresterhill Fax: +44 (0)224-685645 Aberdeen AB9 2ZD Telephone: +44 (0)224-681818 ext 53195 From news.umbc.edu!haven.umd.edu!uunet!olivea!nntp.msstate.edu!willis1.cis.uab.edu!cs.utk.edu!ornl!wingsun!w6i Sun Sep 26 03:00:28 EDT 1993 Article: 7795 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!uunet!olivea!nntp.msstate.edu!willis1.cis.uab.edu!cs.utk.edu!ornl!wingsun!w6i From: w6i@ornl.gov (WINGENBACH J C,45290, ORTC-3 MS-7618) Subject: netgroups & share ???? Message-ID: <1993Sep25.043359.8303@ornl.gov> Keywords: netgroup share Sender: usenet@ornl.gov (News poster) Reply-To: w6i@ornl.gov Organization: Martin Marietta Energy Systems, Inc. Date: Sat, 25 Sep 1993 04:33:59 GMT Lines: 35 Status: RO I have a bit of a problem. I have user directories on machine1 which need to be exported to machine2. Further, I need to set up a subdirectory in each user's account to be pcnfs mountable from any machine. How can I do this? Is there a better way to accomplish the exports? Basically what I am looking at now is: share -F nfs -o rw=machine2 -d "home" /export/home share -F nfs -o rw=netuser1grp -d "user1home" /export/home/user1/dos share -F nfs -o rw=netuser2grp -d "user2home" /export/home/user2/dos share -F nfs -o rw=netuser3grp -d "user3home" /export/home/user3/dos etc. I realize that this will not work since I can not export /export/home & a subdirectory of it. But I do need something like it. I also am having problems getting the netgroup setup. When I simply tried exporting the user's dos subdirectory to a netgroup, it was not recognized. What I had done was place the following in /etc/netgroup: netuser1grp (,user1,) any and all help is appreciated. :) Thanks in advance! --- John C. Wingenbach Martin Marietta Energy Systems Data Systems Research & Development w6i@ornl.gov Oak Ridge Technical Center III (615) 574-8345 1099 Commerce Park (615) 574-0792 (FAX) Oak Ridge, TN 37830 From news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!usenet.ins.cwru.edu!news.ecn.bgu.edu!anaxagoras.ils.nwu.edu!news.acns.nwu.edu!uicvm.uic.edu!u16072 Sun Sep 26 17:55:13 EDT 1993 Article: 7810 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!usenet.ins.cwru.edu!news.ecn.bgu.edu!anaxagoras.ils.nwu.edu!news.acns.nwu.edu!uicvm.uic.edu!u16072 Organization: University of Illinois at Chicago, ADN Computer Center Date: Sun, 26 Sep 1993 01:05:36 CDT From: Message-ID: <93269.010536U16072@uicvm.uic.edu> Newsgroups: comp.sys.sun.hardware,comp.unix.solaris Subject: dirent problem Lines: 11 Xref: news.umbc.edu comp.sys.sun.hardware:12507 comp.unix.solaris:7810 Status: RO I have encountered an interesting bug. When compiling the source code from the man pages (solaris 2.2) of readdir, instead of listing the files, the code lists the files minus the first two characters of the name. When I change the order of the last two entries everything works out correctly. I don't trust this solution, though. Has anyone else had the same problem? IOf so how did you fix it? Thanks in advance. -------------------------------------------------- Thaddeus Olczyk, University of Illinois at Chicago olczyk@uicws.phy.uic.edu From news.umbc.edu!haven.umd.edu!darwin.sura.net!spool.mu.edu!sdd.hp.com!col.hp.com!csn!news.sinet.slb.com!news.london.sinet.slb.com!news.gatwick.sgp.slb.com!news.gatwick.sgp.slb.com!willoch Tue Sep 28 15:43:53 EDT 1993 Article: 7871 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!spool.mu.edu!sdd.hp.com!col.hp.com!csn!news.sinet.slb.com!news.london.sinet.slb.com!news.gatwick.sgp.slb.com!news.gatwick.sgp.slb.com!willoch From: willoch@oslo.sgp.slb.com (thorbjorn willoch) Newsgroups: comp.unix.solaris Subject: Re: How/where to obtain patches for Solaris 2.2? Date: 28 Sep 1993 08:34:01 GMT Organization: Geco-Pracla Europa and Africa Marine Region Lines: 13 Distribution: inet Message-ID: References: <1993Sep28.010259.29836@novell.com> NNTP-Posting-Host: geos132.oslo.sgp.slb.com In-reply-to: jeremy@novell.com's message of Tue, 28 Sep 1993 01:02:59 GMT Status: RO Ftp to sunsite.unc.edu cd /pub/sun-info/sun-fixes There they are all of them. Thorbjørn Willoch | Schlumberger Geco-Prakla Phone: +47-67575548 | Jongsåsvn 4 Internet: willoch@oslo.sgp.slb.com | N-1300 Sandvika, NORWAY -- Thorbjørn Willoch | Schlumberger Geco-Prakla Phone: +47-67575548 | Jongsåsvn 4 Internet: willoch@oslo.sgp.slb.com | N-1300 Sandvika, NORWAY From news.umbc.edu!haven.umd.edu!darwin.sura.net!spool.mu.edu!olivea!koriel!newscast.West.Sun.COM!bardolph!jiml Wed Sep 29 12:09:39 EDT 1993 Article: 7893 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!spool.mu.edu!olivea!koriel!newscast.West.Sun.COM!bardolph!jiml From: jiml@bardolph.West.Sun.COM (James Litchfield - SunSoft Western Area OS Person) Newsgroups: comp.unix.solaris Subject: Re: Any way around remote passwd security? Date: 29 Sep 1993 01:55:46 GMT Organization: SunSoft, A Sun Microsystems, Inc. Business Lines: 76 Distribution: world Message-ID: <28apv2$ik4@newscast.West.Sun.COM> References: Reply-To: jiml@bardolph.West.Sun.COM NNTP-Posting-Host: bardolph.west.sun.com Status: RO If you're setting up accounts under the control of a naming service with a null passwd, the passwd program isn't smart enough to do things correctly if you've specified that passwords are required in the /etc/defaults/passwd file. This is a workaround I use: mv /usr/bin/passwd /usr/bin/passwd.orig (note: leave permissions as is: r-sr-sr-x owned by root and group of sys) insert this script as /usr/bin/passwd (note: 0755 permissions are sufficient for the script - no suid is necessary): ---- Cut Here ---- #! /bin/sh # # Fix for passwd's inability to deal with naming services. Based on an # an idea of Chuck McManis. Expects the original passwd program to live in # /usr/bin/passwd.orig (with its original permissions). Uses the "who" command # instead of the BSD whoami command since you're not guaranteed that the BSD source # compatibility software was installed. # # That is a space (040) between the double quotes in the cut command. # if [ $# -eq 0 ] then UID=`who am i | cut -f 1 -d " "` elif [ $# -eq 1 ] then UID=$1 else echo "passwd: one username at a time please" exit 1 fi DONE=0 NS=`egrep passwd: /etc/nsswitch.conf` for name in $NS do case "$name" in files) IPW=`egrep $UID /etc/passwd` if [ -n "$IPW" ] then /usr/bin/passwd.orig $UID DONE=1 break fi ;; nis) IYP=`ypmatch $UID passwd 2> /dev/null` if [ -n "$IYP" ] then /usr/bin/yppasswd $UID DONE=1 break fi ;; nisplus) DOM=`domainname` INS=`nismatch name=$UID passwd.org_dir.$DOM` if [ -n $INS ] then /usr/bin/nispasswd $UID DONE=1 break fi ;; esac done if [ $DONE -eq 0 ] then echo "Username $UID not found in local files or naming service" exit 2 fi exit 0 From news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!xlink.net!math.fu-berlin.de!uni-paderborn.de!urmel.informatik.rwth-aachen.de!baumeist Wed Sep 29 12:13:02 EDT 1993 Article: 7913 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!xlink.net!math.fu-berlin.de!uni-paderborn.de!urmel.informatik.rwth-aachen.de!baumeist From: baumeist@picasso.informatik.rwth-aachen.de (Markus Baumeister) Newsgroups: comp.unix.solaris Subject: Re: setenv DISPLAY for remote (again) Date: 29 Sep 1993 15:21:31 GMT Organization: Rechnerbetrieb Informatik - RWTH Aachen Lines: 26 Distribution: inet Message-ID: <28c95r$guh@urmel.informatik.rwth-aachen.de> References: NNTP-Posting-Host: goya.informatik.rwth-aachen.de Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Status: RO In willoch@oslo.sgp.slb.com (thorbjorn willoch) writes: >On my Solaris 2.2 box "who am i" works fine to give info about where I >logged in from EXCEPT if my username is eight characters long, then >that information is lost! Well, since many of our usernames are eight chars long, we had similiar problems as the original poster. Finally I found some code in an AnswerBook-script, which seems to solve the problem. (Perhaps Sun should do something about this bug) Try this in your .login file (no guarantue or whatever): # and finally InfoV proudly presents: automatic setting of # DISPLAY even under Solaris (MB) set tty=`/bin/tty | sed 's,/dev/,,g'` set rehost=`/bin/who | grep $tty'.*(.*)$' | sed -e 's,.*(,,g' -e 's,).*,,g'` /bin/who | grep -s $tty'.*(.*)$' >/dev/null if( $status == "0" ) setenv DISPLAY "${rehost}:0" unset tty unset rehost Hope this helps Markus Baumeister -- Markus Baumeister, EMail: baumeist@picasso.informatik.rwth-aachen.de From news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!agate!biosci!TUKKI.JYU.FI!eloranta Thu Sep 30 16:39:27 EDT 1993 Article: 7922 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!agate!biosci!TUKKI.JYU.FI!eloranta From: eloranta@TUKKI.JYU.FI (Jussi Eloranta) Newsgroups: comp.unix.solaris Subject: Re: SS1000 and Solaris 2.2 Message-ID: <199309291832.AA19123@tukki.jyu.fi> Date: 29 Sep 93 18:32:15 GMT References: <9309291615.AA05365@udgserv.cencar.udg.mx> Sender: daemon@net.bio.net Reply-To: "Solaris 2.0 & Solaris for Intel" Distribution: inet Organization: University of Jyvaskyla, Finland Lines: 20 Status: RO In article <9309291615.AA05365@udgserv.cencar.udg.mx> you write: >Hi there! > > I just get A Sparc Server 1000 and I install Solaris 2.2 on it. >The SS1000 has 2 processors, 128 Mb RAM, 2Gb HDD and 256 Mb Swap. > > The problem is that the machine just hang up for 2 o 3 minutes and then >works again. I started the Performance Meter and it goes to 100% every time this happens. >The machine is not using Nis+, NFS or anything else, it just has the operating system. > >Any patches? Yes, 100999-33 corrects many scheduling (cpu and I/O) problems. jussi -- ============================================================================ Jussi Eloranta Dept. of physical chemistry eloranta@jyu.fi University of Jyvdskyld, Finland From news.umbc.edu!haven.umd.edu!darwin.sura.net!spool.mu.edu!nigel.msen.com!well!wisner Thu Sep 30 16:39:54 EDT 1993 Article: 7925 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!spool.mu.edu!nigel.msen.com!well!wisner From: wisner@well.sf.ca.us (Bill Wisner) Subject: Re: pine for solaris 2.2 Message-ID: Sender: news@well.sf.ca.us Nntp-Posting-Host: satori.well.sf.ca.us Organization: The Whole Earth 'Lectronic Link, Sausalito, CA References: <28bmsb$sji@nuscc.nus.sg> Distribution: inet Date: Wed, 29 Sep 1993 20:19:06 GMT Lines: 5 Status: RO Pine 3.85 was recently released, and includes support for Solaris 2.2 when compiled with the BSD compatibility package. The source is on ftp.cac.washington.edu. -- Bill Wisner Gryphon Gang Mountain View CA 94039-0386 From news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!cs.utexas.edu!uunet!math.fu-berlin.de!uniol!majestix.informatik.uni-kiel.de!wib Thu Sep 30 16:41:31 EDT 1993 Article: 7937 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!cs.utexas.edu!uunet!math.fu-berlin.de!uniol!majestix.informatik.uni-kiel.de!wib From: wib@informatik.uni-kiel.de (Willi Burmeister) Newsgroups: comp.unix.solaris Subject: Re: Two sysadmin questions Date: 30 Sep 1993 06:54:52 GMT Organization: Dept. of Computer Science, University of Kiel, FRG Lines: 284 Distribution: inet Message-ID: <28dvrsINN8qj@majestix.informatik.uni-kiel.de> References: <1993Sep27.131019.25761@eng.cam.ac.uk> NNTP-Posting-Host: eden.cs.uni-kiel.de Status: RO In <1993Sep27.131019.25761@eng.cam.ac.uk> cdh@eng.cam.ac.uk (Chris Horne) writes: >We are in the process of moving our network to Solaris 2 running NIS+. >At the moment we have two unresolved questions (although I am sure that >many will follow). Both relate to passwords. >1) When we set up a new user using the User Account Manager, we would like to > specify the "Password: Cleared until first login" option. However, when we do this > the "test" user gets the following response on login > You don't have a password. Choose one. > passwd test > /usr/bin/passwd: test does not exist > Connection closed. Here the script we got from sun: -------------------------------------------------------------------------------- #!/bin/sh # Hack passwd command rapper for other name services. # # @(#)passwd.sh 1.3 03 Aug 1992 Chris.Gerhard@uk.sun.com # trap "exit 1" 1 2 3 4 5 6 7 8 10 12 13 14 15 16 17 22 28 29 30 31 if [ $# -gt 0 ];then pass=$* else pass=$LOGNAME fi result=`while read targ spam ; do if [ "$targ" = "passwd:" ]; then echo $spam break fi done < /etc/nsswitch.conf` for tat in $result do case $tat in files) if egrep "^$pass" /etc/passwd > /dev/null; then exec localpasswd $pass fi ;; nisplus) if nistest [name=$pass],passwd.org_dir.`domainname` then if nismatch name=$pass passwd.org_dir.`domainname` \ > /dev/null 2>&1 then : else echo "Enter old \c" ; keylogin fi exec nispasswd $pass fi ;; nis) if ypmatch $pass passwd > /dev/null ; then exec yppasswd $pass fi ;; *) echo Unknown name service $tat. Check nsswitch.conf exit 0 esac done echo $0:$pass user not found exit 1 -------------------------------------------------------------------------------- do a 'mv /usr/bin/passwd /usr/bin/localpasswd' first and save the script as /usr/bin/passwd. >2) Now once a user has a valid password which they want to change they use > nispasswd. (NB. we don't use password ageing - yet). > vernon% nispasswd > Changing password for cdh on NIS+ server. > Old login password: > New password: > Re-enter new password: > WARNING: Could not update aging information associated with the password. > Updated only the password > NIS+ password information changed for cdh > SERIOUS WARNING: Could not reencrypt NIS+ credentials for cdh; > login and keylogin passwords may differ. > Use NEW passwd for login and OLD passwd for keylogin. > Recommend using "nispasswd cdh" again to > revert back to the old passwd > Or, do "nisaddcred des" for cdh with the NEW password > Permission denied Your permissions on several NIS+ files are wrong. Here our passwd.org_dir: -------------------------------------------------------------------------------- taurec{wib} niscat -o passwd.org_dir Object Name : passwd Owner : taurec.hlt.e-technik.uni-kiel.de. Group : admin.hlt.e-technik.uni-kiel.de. Domain : org_dir.hlt.e-technik.uni-kiel.de. Access Rights : ----rmcdrmcd---- Time to Live : 12:0:0 Object Type : TABLE Table Type : passwd_tbl Number of Columns : 8 Character Separator : : Search Path : Columns : [0] Name : name Attributes : (SEARCHABLE, TEXTUAL DATA, CASE SENSITIVE) Access Rights : r--------------- [1] Name : passwd Attributes : (TEXTUAL DATA) Access Rights : -----m---------- [2] Name : uid Attributes : (SEARCHABLE, TEXTUAL DATA, CASE SENSITIVE) Access Rights : r--------------- [3] Name : gid Attributes : (TEXTUAL DATA) Access Rights : r--------------- [4] Name : gcos Attributes : (TEXTUAL DATA) Access Rights : r--------------- [5] Name : home Attributes : (TEXTUAL DATA) Access Rights : r--------------- [6] Name : shell Attributes : (TEXTUAL DATA) Access Rights : r----m---------- [7] Name : shadow Attributes : (TEXTUAL DATA) Access Rights : ---------------- -------------------------------------------------------------------------------- taurec is NIS+ server. NIS+ admins are all in the NIS+ group admin. All NIS+ client are members of the NIS+ group clients One line in passwd.org_dir looks like: -------------------------------------------------------------------------------- taurec{wib} niscat -o [name=wib]passwd.org_dir Object Name : passwd Owner : wib.hlt.e-technik.uni-kiel.de. Group : clients.hlt.e-technik.uni-kiel.de. Domain : org_dir.hlt.e-technik.uni-kiel.de. Access Rights : ----r---r------- Time to Live : 12:0:0 Object Type : ENTRY Entry data of type passwd_tbl Entry has 8 columns. [1] - [4 bytes] flags = 0x0 'wib' [2] - [14 bytes] flags = 0x0 'secret password' [3] - [4 bytes] flags = 0x0 '107' [4] - [3 bytes] flags = 0x0 '14' [5] - [17 bytes] flags = 0x0 'Willi Burmeister' [6] - [10 bytes] flags = 0x0 '/home/wib' [7] - [9 bytes] flags = 0x0 '/bin/ksh' [8] - [12 bytes] flags = 0x0 '8665:0:::::' -------------------------------------------------------------------------------- And cred.org_dir: -------------------------------------------------------------------------------- taurec{wib} niscat -o cred.org_dir Object Name : cred Owner : taurec.hlt.e-technik.uni-kiel.de. Group : admin.hlt.e-technik.uni-kiel.de. Domain : org_dir.hlt.e-technik.uni-kiel.de. Access Rights : r---rmcdrmcdr--- Time to Live : 12:0:0 Object Type : TABLE Table Type : cred_tbl Number of Columns : 5 Character Separator : : Search Path : Columns : [0] Name : cname Attributes : (SEARCHABLE, TEXTUAL DATA, CASE INSENSITIVE) Access Rights : ---------------- [1] Name : auth_type Attributes : (SEARCHABLE, TEXTUAL DATA, CASE SENSITIVE) Access Rights : ---------------- [2] Name : auth_name Attributes : (SEARCHABLE, TEXTUAL DATA, CASE INSENSITIVE) Access Rights : ---------------- [3] Name : public_data Attributes : (TEXTUAL DATA) Access Rights : -----m---------- [4] Name : private_data Attributes : (TEXTUAL DATA) Access Rights : -----m---------- -------------------------------------------------------------------------------- and one entry: -------------------------------------------------------------------------------- taurec{wib} niscat -o [cname=wib.hlt.e-technik.uni-kiel.de.]cred.org_dir Object Name : cred Owner : taurec.hlt.e-technik.uni-kiel.de. Group : clients.hlt.e-technik.uni-kiel.de. Domain : org_dir.hlt.e-technik.uni-kiel.de. Access Rights : ---------------- Time to Live : 12:0:0 Object Type : ENTRY Entry data of type cred_tbl Entry has 5 columns. [1] - [31 bytes] flags = 0x0 'wib.hlt.e-technik.uni-kiel.de.' [2] - [4 bytes] flags = 0x0 'DES' [3] - [35 bytes] flags = 0x0 'unix.107@hlt.e-technik.uni-kiel.de' [4] - [49 bytes] flags = 0x0 '000000000000000000000000000000000000000000000000' [5] - [65 bytes] flags = 0x0 '0000000000000000000000000000000000000000000000000000000000000000' Object Name : cred Owner : taurec.hlt.e-technik.uni-kiel.de. Group : clients.hlt.e-technik.uni-kiel.de. Domain : org_dir.hlt.e-technik.uni-kiel.de. Access Rights : ---------------- Time to Live : 12:0:0 Object Type : ENTRY Entry data of type cred_tbl Entry has 5 columns. [1] - [31 bytes] flags = 0x0 'wib.hlt.e-technik.uni-kiel.de.' [2] - [6 bytes] flags = 0x0 'LOCAL' [3] - [4 bytes] flags = 0x0 '107' [4] - [3 bytes] flags = 0x0 '14' [5] - [0 bytes] flags = 0x0 '(nil)' -------------------------------------------------------------------------------- This setup works for our NIS+ pool. But attention: this stupid admintool does not set this permissions correct. You have to set them by hand every time you modified one entry in the NIS+ passwd or cred table. We use this small script to clean up: -------------------------------------------------------------------------------- #!/bin/sh DOMAIN=`domainname` if [ $# -ne 1 ] then echo "usage `basename $0` " exit 1 fi nischgrp clients.$DOMAIN. [name=$1]passwd.org_dir nischmod g=r [name=$1]passwd.org_dir nischown $1.$DOMAIN [cname=$1.$DOMAIN.]cred.org_dir nischgrp clients.$DOMAIN. [cname=$1.$DOMAIN]cred.org_dir nischmod a=,n= [cname=$1.$DOMAIN.]cred.org_dir echo "NIS+ for $1 ok" -------------------------------------------------------------------------------- I hope this helps Willi -- -- Willi Burmeister, Department of Computer Science, Kiel University Tel: +49 431 5604 98 S-Mail: Preusserstrasse 1-9, D 24105 Kiel Fax: +49 431 5661 43 E-mail: wib@cs.uni-kiel.de From news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!caen!msuinfo!cuttlefish!vuppala Thu Sep 30 16:42:09 EDT 1993 Article: 7941 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!caen!msuinfo!cuttlefish!vuppala From: vuppala@cuttlefish.msu.edu (Vibhavasu Vuppala) Newsgroups: comp.unix.solaris Subject: Anonymous FTP new problem Date: 30 Sep 1993 07:39:18 GMT Organization: Michigan State University, CPS Department Lines: 222 Sender: vuppala@cuttlefish (Vibhavasu Vuppala) Distribution: world Message-ID: <28e2f6$tm7@msuinfo.cl.msu.edu> NNTP-Posting-Host: cuttlefish.cps.msu.edu Status: RO Hi, I setup anonymous ftp on 2.2 (with mandatory patches). I followed the method given in one of the sun-managers' summaries (see below). When ftp'ing as anonymous user ls command returns the following, ftp> ls 200 PORT command successful. 150 ASCII data connection for /bin/ls (35.8.60.47,33081) (0 bytes). 226 ASCII Transfer complete. ftp> The dir is not empty, it has files (pub,dev,etc,bin etc.). Strange thing is that ls command sometimes works (1 in 10) ! "chroot /bin/ls -l" works fine. Has anybody seen this before ? Any suggestions ? Thank you very much. - vasu I used this to setup ftp : >From sun-managers-relay@ra.mcs.anl.gov Fri May 7 02:43:40 1993 Received: from atlas.cps.msu.edu by atlantic.cps.msu.edu (4.1/rpj-5.0); id AA14837; Fri, 7 May 93 02:43:34 EDT Received: from cpswh.cps.msu.edu by atlas.cps.msu.edu (4.1/rpj-5.0); id AA05851; Fri, 7 May 93 02:43:31 EDT Received: from ra.mcs.anl.gov by cpswh.cps.msu.edu (4.1/4.1) id AA28301; Fri, 7 May 93 02:43:28 EDT Received: by ra.mcs.anl.gov id AA13855 (5.65c/IDA-1.4.4 for sun-managers-outbound); Thu, 6 May 1993 15:40:55 -0500 Sender: sun-managers-relay@ra.mcs.anl.gov Received: from delta.eecs.nwu.edu by ra.mcs.anl.gov with SMTP id AA13849 (5.65c/IDA-1.4.4 for ); Thu, 6 May 1993 15:40:49 -0500 Precedence: bulk Received: from moe.rice.edu by delta.eecs.nwu.edu with SMTP id AA03588 (5.65c/IDA-1.4.4 for ); Thu, 6 May 1993 15:40:33 -0500 Received: from toolik (toolik.Stanford.EDU) by moe.rice.edu (AA17755); Thu, 6 May 93 15:40:30 CDT Received: by toolik (5.0/inc-1.0) id AA29727; Thu, 6 May 93 13:42:26 PDT Date: Thu, 6 May 93 13:42:26 PDT From: scowles@toolik.stanford.edu (S. Cowles) Reply-To: scowles@toolik.stanford.edu (S. Cowles) Followup-To: junk Message-Id: <9305062042.AA29727@toolik> To: sun-managers@rice.edu Subject: SUMMARY: anonymous ftp on Solaris 2.1 X-Sun-Charset: US-ASCII Content-Length: 4361 X-Lines: 167 Status: RO Earlier this week, I asked what I was missing in my setup of anonymous ftp under Solaris 2.1. Output of the ls(1) command showed uid's and gid's with no resolution to user and group names. Oon-Gil Paik sent me a very nice script that covered all the basics. I have taken the liberty of modifying it and present it here. Sid S. Cowles voice: +1 415 812 1917 internet: uucp: scowles@camis.stanford.edu scowles@llnl.gov {backbone}!lll-winken!humpty!scowles ============================================================ #!/bin/sh # script for assisting in setting up anonymous ftp under Solaris 2.1. # original from ogpaik@caltech.edu (Oon-Gil Paik). # modifications and additional comments added by scowles@scheffer.stanford.edu. # sccsid: @(#)anonftpsetup 1.3 13:40:28 93/05/06 # # From ogpaik@caltech.edu Wed May 5 22:46 PDT 1993 # Date: Wed, 5 May 93 22:34:08 PDT # From: ogpaik@caltech.edu (Oon-Gil Paik) # To: scowles@toolik.stanford.edu # Subject: Re: Solaris 2.1 anonymous ftp setup # # Use the above script to setup anonymous ftp. # -OonGil # # Steps for setting up anonymous ftp under Solaris 2.1. # 1. read the ftpd(1M) section for background. # 2. create /etc/shells populated with your favourites. # 3. establish the ftp user account and home directory. # 4. run this script as root. # # Function of necessary files: # # Needed for 'dir' and 'ls' functionality. # /bin/ls # # The following are needed for basic operation. # /usr/lib/ld.so* # /usr/lib/libc.so.1 # /usr/lib/libdl.so.1 # /usr/lib/libintl.so.1 # /usr/lib/libw.so.1 # /etc/passwd # /etc/group # /etc/netconfig # /etc/nsswitch.conf # # The following are needed for 'ls' to resolve NIS names. # /usr/lib/nsw*.so # /usr/lib/libnsl.so.1 # /usr/lib/straddr.so # OPERATIONS: # a function for copying files: filcopy(){ f=$1 ff=`basename $f` d1=`dirname $f` expr "$d1" : "/.*" >/dev/null 2>&1 && fd=`expr "$d1" : "/\(.*\)"` # for cleansing: /bin/rm -f ${FTP}/${fd}/$ff ] && cp $f ${FTP}/${fd} # for quicker operation: # [ ! -f ${FTP}/${fd}/$ff ] && cp $f ${FTP}/${fd} echo "cp $f ${FTP}/${fd}" } # greeting banner. echo Fixing ${FTP} for SunOS 5.x # the home ftp directory FTP="/var/ftp" [ ! -d $FTP ] && { echo "ftp home directory not established." >&2; exit 1; } # necessary subdirectories. [ ! -d ${FTP}/bin ] && mkdir ${FTP}/bin [ ! -d ${FTP}/usr ] && mkdir ${FTP}/usr [ ! -d ${FTP}/usr/lib ] && mkdir ${FTP}/usr/lib [ ! -d ${FTP}/etc ] && mkdir ${FTP}/etc # necessary files. FILES=" /bin/ls /usr/lib/ld.so* /usr/lib/libc.so.1 /usr/lib/libdl.so.1 /usr/lib/libintl.so.1 /usr/lib/libw.so.1 /etc/passwd /etc/group /etc/netconfig /etc/nsswitch.conf /usr/lib/nsw*.so /usr/lib/libnsl.so.1 /usr/lib/straddr.so " # do the file copy shuffle. for FILE in $FILES do filcopy $FILE done # make device nodes. ticotsord and udp are necessary for # 'ls' to resolve NIS names. prefix="/devices/pseudo/mm:" for device in zero do [ -c ${FTP}/dev/$device ] && /bin/rm -f ${FTP}/dev/$device line=`ls -l ${prefix}${device} | sed -e 's/,//'` major=`echo $line | awk '{print $5}'` minor=`echo $line | awk '{print $6}'` mknod ${FTP}/dev/${device} c ${major} ${minor} echo "mknod ${FTP}/dev/${device} c ${major} ${minor}" done prefix="/devices/pseudo/clone:" for device in tcp udp ticotsord do [ -c ${FTP}/dev/$device ] && /bin/rm -f ${FTP}/dev/$device line=`ls -l ${prefix}${device} | sed -e 's/,//'` major=`echo $line | awk '{print $5}'` minor=`echo $line | awk '{print $6}'` mknod ${FTP}/dev/${device} c ${major} ${minor} echo "mknod ${FTP}/dev/${device} c ${major} ${minor}" done # set correct permissions. chmod 666 ${FTP}/dev/* chmod 555 ${FTP}/dev chmod 111 ${FTP}/bin/ls chmod 555 ${FTP}/bin chmod 444 ${FTP}/etc/* chmod 555 ${FTP}/etc chmod 555 ${FTP}/usr/lib/lib* chmod 555 ${FTP}/usr/lib chmod 555 ${FTP}/usr chown ftp ${FTP} chmod 555 ${FTP} exit 0 # OPTIONAL: This is for the wuarchive ftp server. [ ! -d etc/msgs ] && mkdir etc/msgs echo Setting up wuarchive FTP server cp -r /usr/local/etc/msgs ${FTP}/etc chmod -R a+r ${FTP}/etc/msgs chmod a+x ${FTP}/etc/msgs cp -r /usr/local/etc/msgs /etc chmod -R a+r /etc/msgs chmod a+x /etc/msgs if [ ! -f /usr/sbin/in.ftpd.orig ]; then mv /usr/sbin/in.ftpd /usr/sbin/in.ftpd.orig fi cp /export/local/etc/ftpd /usr/sbin/in.ftpd ============================================================ From news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!pipex!uunet!tarpit!gator!sawmill!sawmill!rjk Thu Sep 30 16:42:35 EDT 1993 Article: 7943 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!pipex!uunet!tarpit!gator!sawmill!sawmill!rjk From: rjk@sparcmill.sawmill.uucp (Richard J Kuhns) Subject: Re: /var/mail problem. In-Reply-To: keithc@tomahawk.welch.jhu.edu.'s message of 28 Sep 1993 16:50:57 GMT Message-ID: Sender: roots@sawmill.uucp (Admin (ksh)) Nntp-Posting-Host: sparcmill Organization: Grauel Enterprises References: <289q1hINNig7@jhunix.hcf.jhu.edu> Date: 29 Sep 93 11:52:23 Lines: 30 Status: RO In article <289q1hINNig7@jhunix.hcf.jhu.edu> keithc@tomahawk.welch.jhu.edu. (Keith Christopher) writes: ANyone have this problem : If a user's mailbox does not exist in /var/mail and I CC them using mail, their mailbox file will be created, however it will be owned by me !! e.g. where me is me and you is the user with no mailbox file. mail me CC:you References: <28eupiINN20q@fstgds15.tu-graz.ac.at> Reply-To: dminer@East.Sun.Com NNTP-Posting-Host: badboy.east.sun.com Status: RO In article 28eupiINN20q@fstgds15.tu-graz.ac.at, schinagl@fstghp04.tu-graz.ac.at (Hermann Schinagl) writes: ... >Not only admintool does the wrong permissions on credentials ! If you add an >account with nistbladm, and add the credentials by hand later with nisaddcred >you will see that the LOCAL credentials of a user are owned by the server, >but the DES credentials are owned by the owner of the entry ( == user ) > Patch 101159-03 is supposed to fix the problem with admintool. I don't think it's out yet, though. It's not clear to me exactly which part of nisaddcred you think is wrong, but this is how it should be. It is usually desirable for the DES credentials to be encrypted with the user's current password so that login can handle the keylogin automatically. Thus if the password is changed with nispasswd, it also needs to re-encrypt the secret key, and thus the user needs the rights to modify her own DES entry since nispasswd runs under that user's identity and directly updates the table. With a different ar chitecture, e.g. a password updating daemon running under some specific administrative identity, this would not be the case. The LOCAL creds, on the other hand, better be owned by the administrator (root on the server at present) so that users can't just go changing their uid or gid memberships insofar as Secure RPC is concerned. When you do a netname2user, the LOCAL cred is what's consulted for the answer. Dave (who put the bug in admintool in the first place and has learned a great deal about Secure RPC since..) --- --------------------------------------------------------------------------- Dave Miner dave.miner@east.sun.com SunSelect, PC Networking Engineering (508) 442-0463 --------------------------------------------------------------------------- From news.umbc.edu!haven.umd.edu!darwin.sura.net!gatech!swrinde!cs.utexas.edu!uunet!pipex!zaphod.crihan.fr!univ-lyon1.fr!scsing.switch.ch!bernina!feigin Sat Oct 2 21:56:40 EDT 1993 Article: 7980 of comp.unix.solaris Newsgroups: comp.sys.sun.admin,comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!gatech!swrinde!cs.utexas.edu!uunet!pipex!zaphod.crihan.fr!univ-lyon1.fr!scsing.switch.ch!bernina!feigin From: feigin@iis.ee.ethz.ch (Adam W. Feigin) Subject: Re: xntp for Solaris 2.2? Message-ID: Sender: news@bernina.ethz.ch (USENET News System) Nntp-Posting-Host: err-gw Organization: Swiss Federal Institute of Technology (ETH), Zurich, CH References: Date: Fri, 1 Oct 1993 07:52:52 GMT Lines: 20 Xref: news.umbc.edu comp.sys.sun.admin:21731 comp.unix.solaris:7980 Status: RO scott@goga.ee.usu.edu (Scott Budge) writes: >Does anyone know of a solaris 2.2 port of xntp? I would appreciate >any pointers. The latest xntp3 from louie.udel.edu:~ftp/pub/ntp has support for Solaris 2.x. We've been running it here for several months now. /AWF ------------------------------------------------------------------------------ Internet: awf@iis.ee.ethz.ch Adam W. Feigin UUCP:{backbones}!iis!feigin Network Systems Manager Mail: Integrated Systems Laboratory Institute for Integrated Systems ETH-Zentrum Swiss Federal Institute of Technology CH-8092 Zurich Zurich, Switzerland Switzerland Phone: +41 1 632 50 53 FAX: +41 1 252 09 94 "I'd like to donate my body to science after I die, but I'm afraid its taken far too much of it already" From news.umbc.edu!haven.umd.edu!darwin.sura.net!gatech!swrinde!cs.utexas.edu!uunet!mcsun!Germany.EU.net!pophh!nuki.NetUSE.de!mdc.netuse.de!olaf Sat Oct 2 21:57:26 EDT 1993 Article: 7981 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!gatech!swrinde!cs.utexas.edu!uunet!mcsun!Germany.EU.net!pophh!nuki.NetUSE.de!mdc.netuse.de!olaf From: olaf@mdc.netuse.de (Olaf Schlueter) Subject: Re: in.rarpd, NIS+, Solaris 2.1 Message-ID: Keywords: in.rarpd, NIS+, Solaris 2.1 Organization: Medical Diagnostic Computing GmbH References: Date: Fri, 1 Oct 1993 08:17:17 GMT Lines: 23 Status: RO dawson@willard.atl.ga.us (Willard Dawson) writes: >Try as I might, I've had absolutely no luck in getting a diskless client >to "boot net" from a Solaris 2.1 boot server, while NIS+ (running in YP- >compatibility mode "/usr/sbin/rpc.nisd -Y -r -S 0") is active. "boot net" didn't work for me either running NIS, until I discoverd that (probably) admintool had inserted the ethernet adresses with leading zero's into ethers.org_dir, i.e. 08:00:20... instead of 8:0:20. Should not break RARP but does. Obviously the bug is in in.rarpd, which tries to match the "without leading zeros" ethernet adress recieved in or generated from the RARP call against the "with leading zeros" adress in ethers.org_dir literally. Removing the leading zeros from the entries in ethers.org_dir made it working. -- Olaf Schlüter, System Administrator, Olaf.Schlueter@mdc.netuse.de Medical Diagnostic Computing GmbH, Düppelstr. 71, 24105 Kiel, Germany, Office Voice: 49-431-804220, Fax: 49-431-804230 From news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!sol.ctr.columbia.edu!news.kei.com!bloom-beacon.mit.edu!ai-lab!prep.ai.mit.edu!gnulists Sat Oct 2 22:07:24 EDT 1993 Article: 8017 of comp.unix.solaris Newsgroups: comp.unix.solaris,gnu.utils.bug From: Jean-Luc.Richier@imag.fr (Jean-Luc Richier) Subject: Re: gdb on sparc-solaris2.2 Message-ID: Organization: IMAG Institute, University of Grenoble, France References: <28hiqi$q50@geraldo.cc.utexas.edu> Date: Fri, 1 Oct 1993 19:14:05 GMT Apparently-To: gnu-utils-bug@uunet.uu.net Distribution: world Approved: info-gnu@prep.ai.mit.edu Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!sol.ctr.columbia.edu!news.kei.com!bloom-beacon.mit.edu!ai-lab!prep.ai.mit.edu!gnulists Followup-To: comp.unix.solaris,gnu.utils.bug Lines: 74 Xref: news.umbc.edu comp.unix.solaris:8017 gnu.utils.bug:7435 Status: RO In article <28hiqi$q50@geraldo.cc.utexas.edu> smcdow@mamba.asg.arlut.utexas.edu (Stuart McDow) writes: >I just built and installed gdb-4.10.pl1 on my SS10 running Sol 2.2 > >The build went without a hitch (praise be to configure). I used >gcc-2.4.5 for the build. However, it isn't running correctly. Here's a >simple example. The program was built using gcc-2.4.5. I know the >program runs correctly (it's small), and I've built and debugged the >same under SunOS4.1.{2,3} with no problems. Any clues would be >appreciated. Thanks. We had the same problem here (using gdb 4.9). The cause was the installation of the patch sunos 5.2 any 101052-02 SunOS 5.2: point patch to link-editor & related files This patch replaces the link editor, and seems to clobber the debug informations. If you uninstall this patch and go back to the link editor of the Solaris 2.2, everything is fine. PS: There is a new version of the patch, 101052-03, but we have not tested it. README of the patch: Patch-ID# 101052-02 Keywords: performance enhancements ld link large relocatable archive Synopsis: SunOS 5.2: point patch for slow link-editing problem Date: Jul/23/93 Solaris Release: 2.2 SunOS release: 5.2 Unbundled Product: Unbundled Release: Topic: Solaris 2.2: point patch to usr/ccs/bin/ld BugId's fixed with this patch: 1108453 Changes incorporated in this version: fixes packaging issue in -01 Relevant Architectures: sparc Other patches which conflict with this patch: This patch obsoleted by: Files included with this patch: usr/ccs/bin/ld usr/lib/libld.so.1 usr/lib/liblddbg.so.2 usr/lib/libelf.so.1 usr/lib/libelf.so usr/lib/libelf.a Problem Description: This is a POINT PATCH which may NOT be necessary for all customers, and as such is not scheduled to be integrated fix in a future SunOS release. This patch fixes the following problem: Link-editing is too slow when using vast quantities of relocatable objects and/or archive input files. -- Jean-Luc RICHIER (Jean-Luc.Richier@Imag.Fr richier@imag.fr) LGI, IMAG-CAMPUS, BP 53, F-38041 GRENOBLE Cedex 9 Tel : (33) 76 51 48 32 Fax : (33) 76 44 66 75 From news.umbc.edu!haven.umd.edu!ames!elroy.jpl.nasa.gov!usc!venice!wp-sp.nba.trw.com!deneva.sdd.trw.com!news Mon Oct 4 15:36:24 EDT 1993 Article: 8049 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!ames!elroy.jpl.nasa.gov!usc!venice!wp-sp.nba.trw.com!deneva.sdd.trw.com!news From: jumper@spf.trw.com (Greg Jumper) Newsgroups: comp.unix.solaris Subject: Re: Solaris 2.2/xdm question Date: 3 Oct 1993 19:57:54 GMT Organization: TRW Signal Processing Facility, Redondo Beach, CA Lines: 27 Distribution: world Message-ID: <28nas2$390@deneva.sdd.trw.com> References: <28m756INNe5u@fstgds15.tu-graz.ac.at> Reply-To: jumper@spf.trw.com NNTP-Posting-Host: lens.spf.trw.com Status: RO Regarding the latest version of the "xdm" patch, Hermann Schinagl wrote: But, where do I find this patch via ftp ? I have searched the relevant archives such as sunsite, thor, uni-greiburg ..., but they only have 101135-01. Because I have more than one sun to patch, I won't apply -01 first and repeat the whole procedure again with -02... By far the best way to get the latest patches is to have a software support contract with Sun and get your own account with the "Online Support Service". Telnet to "sunsolve1.sun.com" and follow the instructions; you will need your support contract number, and I believe you must be one of the registered technical contacts for the contract as well. After your credentials are verified and your account is created, you can ftp to the same site for all the latest patches. If, for whatever reason, you can't create your own account on "sunsolve1", then you could try contacting the appropriate sales person or system engineer at your local Sun office. Presumably, the latest patches eventually show up at "sunsite", etc. -- Greg Jumper TRW Signal Processing Facility jumper@spf.trw.com Status: RO Status: RO Status: RO Status: RO From news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!magnus.acs.ohio-state.edu!usenet.ins.cwru.edu!agate!ames!koriel!male.EBay.Sun.COM!news2me.EBay.Sun.COM!cronkite.Central.Sun.COM!sixgun.East.Sun.COM!sungy!uk-usenet.uk.sun.com!news Mon Oct 4 15:38:45 EDT 1993 Article: 8062 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!magnus.acs.ohio-state.edu!usenet.ins.cwru.edu!agate!ames!koriel!male.EBay.Sun.COM!news2me.EBay.Sun.COM!cronkite.Central.Sun.COM!sixgun.East.Sun.COM!sungy!uk-usenet.uk.sun.com!news From: ta@trinity.uk.sun.com (Tim Addison - Sun UK - SE Cambridge) Newsgroups: comp.unix.solaris Subject: Re: /etc/system variables Date: 4 Oct 1993 13:05:12 GMT Organization: Sun Microsystems (UK) Lines: 45 Distribution: world Message-ID: <28p728$o92@uk-usenet.uk.sun.com> References: <1993Sep23.082918.60@cine88.cineca.it> Reply-To: ta@trinity.uk.sun.com NNTP-Posting-Host: trinity.uk.sun.com In article 60@cine88.cineca.it, marcello@deis27.cineca.it (Marcello Dalpasso) writes: > I don't know if this s a FAQ, but: > > Can anyone post a "complete" list of available settings (variables and/or options) > in /etc/system for Solaris 2.1 ? > > Thanks in advance. Marcello. >From my archives this is a reply from SMCC technical marketing to a similar question. Regards, Tim Addison ------------------------------------------ There is no definitive list because /etc/system lets you patch any kernel symbol. The kernel sourcecode is about as definitive as you will get :-) The *documented* list of /etc/system parameters is in the manual section "Administering SunOS 5.2 Security Performance and Accounting - Appendix A - Tuning Kernel Parameters". Some benchmarks require that tuning parameters are disclosed, so the SPEC LADDIS report "documents" some tweaks for Solaris 2.3alpha that helps NFS performance on an SS1000 using FDDI. set maxusers=32 ( some of the dependent parameters are set explicitly below) set ufs_ninode=32768 set ncsize=32768 set nsegkp=1024 set nhbuf=256 set p_nbuf=2048 set bufhwm=20480 set bf:bf_nbufs=2048 set bf:bf_nrmds1=256 ------------ From news.umbc.edu!haven.umd.edu!purdue!news.bu.edu!olivea!spool.mu.edu!howland.reston.ans.net!usc!elroy.jpl.nasa.gov!nntp-server.caltech.edu!dori Mon Oct 4 15:40:56 EDT 1993 Article: 8072 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!purdue!news.bu.edu!olivea!spool.mu.edu!howland.reston.ans.net!usc!elroy.jpl.nasa.gov!nntp-server.caltech.edu!dori From: dori@cco.caltech.edu (Dori S. Levanoni) Newsgroups: comp.unix.solaris Subject: Re: SS1000 and Solaris 2.2 Date: 4 Oct 1993 18:49:08 GMT Organization: California Institute of Technology, Pasadena Lines: 38 Distribution: inet Message-ID: <28pr74INNt8q@gap.caltech.edu> References: <9309291615.AA05365@udgserv.cencar.udg.mx> NNTP-Posting-Host: punisher-gw.caltech.edu Status: RO marco@UDGSERV.CENCAR.UDG.MX ("Marco A. Gaspar R.") writes: >Hi there! > I just get A Sparc Server 1000 and I install Solaris 2.2 on it. >The SS1000 has 2 processors, 128 Mb RAM, 2Gb HDD and 256 Mb Swap. > The problem is that the machine just hang up for 2 o 3 minutes and then >works again. I started the Performance Meter and it goes to 100% every time this happens. >The machine is not using Nis+, NFS or anything else, it just has the operating system. >Any patches? Hmmm...we've got the identical setup, and had lots of problems with it, until we started applying the patches. There seem to be roughly 11 mandatory patches, some of which fix kernel memory bugs, nfs bugs, etc. You might want to ask for these patches. The ones we've got applied are: 100982-02 100985-04 100992-02 100999-33 101014-04 101018-02 101025-05 101031-01 101039-06 101065-01 101090-01 101109-02 Hope this helps... Dori -- *---------------------------------------------------------------------------- Dori Levanoni PADI AOW #88302104 | Did you ever wonder what hood ornaments Internet : dori@cco.caltech.edu | on cars were for? They're to help the Compuserve : 76646,3433 | drivers aim better... From news.umbc.edu!haven.umd.edu!purdue!lerc.nasa.gov!usenet.ins.cwru.edu!howland.reston.ans.net!xlink.net!fauern!news.th-darmstadt.de!news.th-darmstadt.de!news Tue Oct 5 13:25:24 EDT 1993 Article: 8084 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!purdue!lerc.nasa.gov!usenet.ins.cwru.edu!howland.reston.ans.net!xlink.net!fauern!news.th-darmstadt.de!news.th-darmstadt.de!news From: zim@dvs1.informatik.th-darmstadt.de (Juergen Zimmermann) Newsgroups: comp.unix.solaris Subject: mxgdb for Solaris 2.2 Date: 4 Oct 1993 21:12:15 GMT Organization: Tech. University Darmstadt, Germany Lines: 17 Distribution: world Message-ID: <28q3jfINNc3b@rs18.hrz.th-darmstadt.de> Reply-To: zim@dvs1.informatik.th-darmstadt.de NNTP-Posting-Host: athen.dvs1.informatik.th-darmstadt.de Status: RO Hi, I ported the Motif based interface to gdb `mxgdb 1.1.6' to Solaris 2.2 and created a patch file which is located at ftp.dvs1.informatik.th-darmstadt.de. -- Juergen ______________________________________________________________________ Juergen Zimmermann phone: +49 6151 16 6841 Technical University fax: +49 6151 16 5554 Dept. of Computer Science email: zim@dvs1.informatik.th-darmstadt.de Frankfurter Str. 69a finger: zim@paris.dvs1.informatik.th-darmstadt.de D-64293 Darmstadt Germany From news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!magnus.acs.ohio-state.edu!csn!csn!evolving.com!dct Wed Oct 6 13:42:26 EDT 1993 Article: 8133 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!magnus.acs.ohio-state.edu!csn!csn!evolving.com!dct From: dct@evolving.com (Dave Twinam) Subject: RE: How/where to obtain patches for Solaris 2.2 Message-ID: <1993Oct5.133612.19353@evolving.com> Date: Tue, 5 Oct 1993 13:36:12 GMT Organization: Evolving Systems, Inc. X-Newsreader: TIN [version 1.2 PL1] Lines: 409 Status: RO Jeremy, I received the following from a Sun engineer. It works quite well !! Hope this helps out. Dave SUNSOLVE ONLINE (8.6.93) SunSolve Online(TM) is an extensive database of the latest support solutions, including common questions and answers, bug and patch descriptions, and technical notes. Updated nightly, SunSolve Online contains the most current information on problems and solutions. It may be accessed using SprintNet(TM) or over the Internet. FEATURES BENEFITS ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ FTP Patches over the Internet * Easy way of retrieving patches of any size ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Updated nightly * Most current information ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Internet access * Easy access for customers connected to the Internet ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ SprintNet access * Modem access for customers without a connection to the Internet ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Access to extensive collections of articles developed by Sun Experts EarlyNotifier * Proactive notification of newly discovered key problems or bugs Bug Reports Summaries * Summaries of all bugs logged Patch Descriptions * Information on patches that can be requested automatically Symptoms/Resolutions * Answers to questions asked by customers world-wide Information Documents * More in-depth articles and other information Solaris 2.X Q&A * Specific information to help migrate to SVR4/Solaris 2.x ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ If you are a named contact on a maintenance contract, and you have access to a modem or to the Internet, you can begin using SunSolve Online immediately by following the instructions in the SunSolve Online Access Reference Card which is attached below. SUNSOLVE FAX (8.6.93) SunSolve Fax, SunService's fax-on-demand system, provides a subset of the SunSolve data base for those who don't have access to the Internet, or a modem, but still want the latest information. SunSolve FAX features the most current and most frequently requested articles and information. All you need is a touch tone telephone and a Fax machine. To access SunSolve Fax, dial 800-USA-4SUN (800-872-4786) and select the SunSolve Fax option from the menu. You will then be prompted to select one of the following menu options to have that information sent to a specified fax machine telephone number: PROMPT DESCRIPTION Instructions SunSolve Fax instructions EarlyNotifier Proactive notification of key problems or bugs just discovered Quick reference materials Useful information such as patch ordering information and key word charts Software Reference Index Arranged by product category, provides a list of Software reference article titles and numbers Hardware Reference Index Arranged by product category, provides a list of Hardware reference article titles and numbers Request article number(s) Allows users to retrieve specific articles by article number To find the article you need, call SunSolve Fax, select the desired index and category, and enter your fax number when prompted. A list of articles will be sent to your fax machine within a few minutes. To select the desired article(s), call SunSolve Fax again and select the "Request Article Number" option. When prompted, request the article(s) you want, and enter your fax number. Your requested article(s) will be sent to your fax machine in a few minutes. FEATURES BENEFIT ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Touch-tone Phone access * Easy and convenient way to retrieve most accessed SunSolve articles ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Delivered directly to * Easy and most readily available a fax machine method of delivery. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Access to a collections of latest and most frequently requested articles developed by Sun Experts EarlyNotifier * Proactive notification of newly discovered key problems or bugs Quick reference * Useful information such as patch ordering materials information and key word charts Software references * Arranged by product category, provides index a list of Software reference article titles and numbers Hardware references * Arranged by product category, provides index a list of Hardware reference article titles and numbers SunSolve FAX will be available by October 1, 1993. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ A T T A C H M E N T SunSolve OnLine Welcome to SunService's SunSolve Online(tm) customer support application. SunSolve Online, an integral part of SunSpectrum(tm) support, is a terminal based system that contains up-to-the-minute information for Sun workstations. SunSolve Online provides a simple interface for performing complex searches through a wide variety of information collections. SunService(tm) is pleased to announce that SunSolve Online is now available via SprintNet(tm) dial-up service and the Internet. We have also added the ability to retrieve patches using 'ftp' access. The new system, the server name is SUNSOLVE1, contains all information found on the current SunSolve Online system plus online patch access, browse mode, and several other new features. With the introduction of the new SunSolve Online, access to the old system will be phased out starting Sep. 1 1993. You will find the new features very easy to use and access via SprintNet or Internet will put SunSolve Online at your finger tips. Access to these and many more new features are now available. Follow the simple instructions outlined below. Once you have logged into our new and expanded system we suggest you E-mail, 'ftp', or request a hardcopy of the latest SunSolve Online Users Guide. This PostScript(tm) document will help introduce you to the new SunSolve Online application. ============================================================================== v1.4 SunSolve Online Access Reference Card ============================================================================== WARNING - The access and use of SunSolve Online is restricted to SunService Inc. contract customers only. Unauthorized access is prohibited. Note: Your input is underlined with [***]. SprintNet access to SunSolve Online is a 2 step process: 1. logging into SprintNet, 2. log into Sun's Customer Network and SunSolve Online. Access via SprintNet Services ----------------------------- 1. For your local SprintNet access phone number call SprintNet customer service toll-free at 1-(800)-877-5045, (option #5) or (404)-859-7700. A. Use your modem to dial the closest SprintNet local access telephone number. Select and dial the appropriate number for the desired baud rate (300, 1200, 2400, 9600). SPEED MODEM TYPE ----- ---------- 300 bps Bell 103 compatible 1200 bps Bell 212A compatible 2400 bps V.22 bis compatible 9600 bps (V.29) Microcom V.29 compatible 9600 bps (V.32) CCITT V.32 compatible NOTE: For dial-out capabilities of your system, use applications such as 'tip' or other methods appropriate in your environment. Contact your system administrator for your modem functionality and dial-out access capabilities. 'tip' should always be run from a Shell Tool. B. Upon successful connection you will see this prompt: CONNECT If your modem is using 8 databits, no parity, 1 stop bit, type @, capital `D', (release the shift) then a return (@D). ********** OR If you are using 7 data bits, even or odd parity, one stop bit, type @, followed by carriage return. (@) ********* C. You then get this response: TELENET 415 3I (this number response may vary) Enter the terminal type information: TERMINAL= d1 (CRT's and PC's) *** D. Connecting to Sun: >From the SprintNet prompt (@) type the following information. @ c 408133,sunsolve ***************** PASSWORD = sunmicro ******** 408 133 connected 2. Log into the Sun Customer Network. You will now log into the Sun Customer Network. This server will allow you access to the features and functionality of SunSolve Online. login: sunsolve ******** Password: sunmicro ******** 3. Go to "SunSolve Customer Login Screen" section of this document. Internet Access to SunSolve Online ---------------------------------- A. From your workstation type: prompt% telnet sunsolve1.sun.com ************************ B. Go to step 2 above. Note - SUNSOLVE1 Internet IP address 192.9.9.24 FTP Access to Binary Patches ---------------------------- Note -For ftp access you must FIRST set up a SunSolve Online user account. Login via SprintNet or the Internet and complete all sections of the SunSolve Customer Login Screen. Then: A. From your workstation type: prompt% ftp sunsolve1.sun.com ********************* B. Login: sunsolve ******** Password: sunmicro ******** C. 332-User sunsolve logged in; account ID requested. 332-Enter your ID and password separated by a slash (/). This won't echo. Account: your sunsolve login name/your sunsolve password *********************************************** Note - SUNSOLVE1 Internet IP address 192.9.9.24 SunSolve Customer Login Screen ------------------------------- Last login: Wed July 7 22:17:05 from 31104150011830 SunOS Release 4.1.1 (X25) #1: Wed Feb 24 10:20:55 PST 1993 Last login: Wed July 7 10:27:21 from mailhost Welcome to SunSolve ONLINE version 1.3 PRESS [RETURN] TO START Terminal type [network]:sun (shell tool) or sun-cmd (command tool) *** ******* Note: The default terminal type appears in brackets [] prior to the colon. If this is not your terminal type, ask your system administrator for the UNIX `terminfo' identifier for your terminal and enter it before pressing [Return]. A. Customer Log on Screen If this is your first time accessing the system enter `newuser' to be given temporary access until your account is validated. For all future access please use your user defined login/password generated below. To gain access to SunSolve Online type the following information: LOGIN SunSolve ONLINE v1.3 login: newuser ******* Password: (no password required) ****** B. Setting up your own user account and password file. The example below shows a New User Account registration form. Please complete all sections. Then select your personal SunSolve Online login account information. New User Account SunSolve ONLINE v1.3 ContractID (if known): Company: Contact: Address: City: State: ZIP Code: Phone: E-mail address: Login:[enter your future SunSolve Online account name] Password:[enter your future SunSolve Online password] Note: Your new login/password will take effect on your next login to SunSolve Online. Please exit and login using your newly created account. This account information is also required for ftp access. Warning - Unauthorized access is prohibited. ^^^^^^^ How to obtain user documentation --------------------------------- There are four ways to obtain a copy of the latest SunSolve Online User's Guide. Depending on your method of connectivity and access to the Internet, select the appropriate method. a. If you have logged in via ftp, 'get' a copy of the UserGuide.ps file located in the patch directory. Also in this directory, is a README file that provides information on how to obtain patch binary files. b. If you have logged in via Internet, from the SunSolve Online main menu select item #7, "Send User Guide (via email)". From the preceding screen verify that your Internet E-Mail address is correct. You will be E-mailed a PostScript(tm) copy of the user's guide. c. If you have logged in via SprintNet and have an E-mail address that is accessible from the Internet, follow step "b" above. d. If you have logged in via SprintNet and do not have an Internet E-mail address, select menu item #8, "Send Hardcopy User Guide". Complete the request form. A copy will be mailed to you through the U.S. Post Office. Please allow 7 to 10 working days for delivery. /\ \\ \ \ \\ / / \/ / / / / \//\ SunService \//\ / / A Sun Microsystems, Inc Business / / /\ / 2550 Garcia Avenue, Mountain View 94043 / \\ \ 415 960-1300 fax 415 969-9131 \ \\ \/ -- ***************************************************************************** Dave Twinam --- Evolving Systems, Inc. 8000 East Maplewood Ave., Englewood, CO 80111 Internet: dct@evolving.com Telephone: (303) 689-1358 FAX: (303) 689-1399 From news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!math.ohio-state.edu!cs.utexas.edu!uunet!munnari.oz.au!newshost.anu.edu.au!dubhe.anu.edu.au!barnard.anu.edu.au!not-for-mail Wed Oct 6 13:45:08 EDT 1993 Article: 8149 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!math.ohio-state.edu!cs.utexas.edu!uunet!munnari.oz.au!newshost.anu.edu.au!dubhe.anu.edu.au!barnard.anu.edu.au!not-for-mail From: chrism@cs.anu.edu.au (Chris McDonald UWA) Newsgroups: comp.unix.solaris Subject: Re: what is /etc/.pwd.lock /etc/.mnt.lock? Date: 6 Oct 1993 20:58:31 +1000 Organization: Department of Computer Science, Australian National University Lines: 17 Distribution: inet Message-ID: <28u8cnINNlcc@barnard.anu.edu.au> References: <28sfetINN194@gap.caltech.edu> <28sja3$5vr@bradley.bradley.edu> NNTP-Posting-Host: barnard.anu.edu.au Status: RO pwh@bradley.bradley.edu (Pete Hartman) writes: >In <28sfetINN194@gap.caltech.edu> ogpaik@caltech.edu (Oon-Gil Paik) writes: >>Hi, I just noticed two dot files in /etc directory and I don't know >>how they are created and what they are for. /etc/.mnt.lock is the newest mechanism by which it becomes impossible to bring up a Solaris machine that has been transported between sites. If for some reason this file exists on a read-only partition (such as / when in single user mode), its being impossible to delete means that the only way to change /etc/vfstab entries using the "toolset" in /sbin to enable mounting of different file-systems from different machines is to reinstall Solaris from CD (I'm no longer bitter about this as it happened over a week ago and I've calmed down now). chris. From news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!cs.utexas.edu!uunet!mcsun!sun4nl!fwi.uva.nl!casper Wed Oct 6 13:45:56 EDT 1993 Article: 8153 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!cs.utexas.edu!uunet!mcsun!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: Installing new patch version w/o backing out old version Date: 6 Oct 1993 12:17:28 GMT Organization: FWI, University of Amsterdam Lines: 11 Distribution: inet Message-ID: <28ud0o$sat@mail.fwi.uva.nl> References: <28444oINN6f@cbs.ksu.ksu.edu> NNTP-Posting-Host: adam.fwi.uva.nl Keywords: Solaris 2.2 Status: RO tar@cbs.ksu.ksu.edu (Tim Ramsey) writes: >Say you have patch 100999-21 installed under Solaris 2.2. Is it safe to >installpatch 100999-33 without first backing out 100999-21? Is it safe in >the general case (that is, with any patch)? If the new patch changes the same (or more) files, there is no problem. This is generally true. On clients we don't save the old files and can't backout. No problems so far. Casper From news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!mcsun!sun4nl!fwi.uva.nl!casper Thu Oct 7 19:45:25 EDT 1993 Article: 8172 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!mcsun!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: Where to find tcsh that will compile un Date: 6 Oct 1993 22:07:38 GMT Organization: FWI, University of Amsterdam Lines: 20 Distribution: world Message-ID: <28vfja$479@mail.fwi.uva.nl> References: <28at6n$ncb@senator-bedfellow.MIT.EDU> <28vd6j$j5k@hal.com> NNTP-Posting-Host: adam.fwi.uva.nl Status: RO centaur@dillia.hal.com (Paul Blair) writes: > I've tried all the currently posted suggestions with no luck. It seems that >tcsh is unable to be compiled with any luck on Solaris 2.2. The previously mentioned >binary core dumps with an illegal instruction, and the source from which it >was compiled dies on the last step of the make with an error which looks much >like the following. It won't compile with gcc or the Solaris cc. >Undefined first referenced > symbol in file >tgetstr ed.screen.o Looks like you're picking up the wrong termcap/curses library. Get rid of /usr/ucblib in your LD_LIBRARY_PATH, and/or install /usr/ccs/lib/lib{termcap,curses}.a Casper From news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!howland.reston.ans.net!xlink.net!math.fu-berlin.de!uniol!Andreas.Bartelt Mon Oct 11 08:43:22 EDT 1993 Article: 8263 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!howland.reston.ans.net!xlink.net!math.fu-berlin.de!uniol!Andreas.Bartelt From: Andreas.Bartelt@arbi.informatik.uni-oldenburg.de (Andreas Bartelt) Subject: Re: FTP site: binaries for Solaris 2.2? Organization: University of Oldenburg, Germany Distribution: inet Date: Fri, 8 Oct 1993 14:49:54 GMT Message-ID: <1993Oct8.145311.20767@arbi.Informatik.Uni-Oldenburg.DE> References: <28q43u$odq@nwfocus.wa.com> Sender: news@arbi.Informatik.Uni-Oldenburg.DE Lines: 92 Status: RO markmc@halcyon.com (Mark McWiggins) writes: >Is there an FTP site around with Solaris 2.2 binaries ready to go? ============================================================================ This is a list of FTP Sites where you can find Solaris related files: (Version 1.3 / 03.06.1993) Site Directory Remarks ---------------------------------------------------------------------------- farside.rutgers.edu / /usr/local of this machine! /bin /cygnus /lib /src ftp.luth.se /pub/unix/sun/Papers /pub/unix/sun/SunPapers /pub/unix/sun/sun_os_patch ftp.prz.tu-berlin.de /pub/docs/SunPapers Documentation Site ftp.Uni-Oldenburg.DE * /pub/unix/solaris/info Some informational stuff, e.g. FAQ /pub/unix/solaris/mirror Many useful mirrors! ftp.uu.net /systems/sun/ * /systems/sun/solaris Solaris descriptions (*.ps) * /systems/sun/sun-dist /systems/sun/sun-fixes * /systems/sun/sunflash /systems/sun/sunview ftp.uwtc.washington.edu /pub/SunFixes /pub/Sun_Software opcom.sun.ca /pub /pub/Cygnus /pub/binaries /pub/patches/sunos /pub/scripts /pub/tars /pub/x11r5 ra.mcs.anl.gov /sun-managers X11R5 patch kit, format.dat, sun-managers FAQ sunline.epfl.ch /pub /pub/sun-papers /pub/sun-patches sunsite.unc.edu /pub/sun-info /pub/wais /pub/packages/cygnus /pub/packages/terminal-emulators A site for x86 binaries is still missing :( An asterix (*) at the beginning of the line indicates that this site/directory always carries the most actual versions. Send corrections/additions to 'Andreas.Bartelt@Informatik.Uni-Oldenburg.DE'. ============================================================================ -- *~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~* Name : Andreas Bartelt / Nickname: Tares Organization: Fachbereich 10 Informatik Softwarelab., Universitaet Oldenburg EMail : Andreas.Bartelt@Informatik.Uni-Oldenburg.DE FTP-Adm/Unix: bartelt@aix01.HRZ.Uni-Oldenburg.DE *~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~* From news.umbc.edu!haven.umd.edu!darwin.sura.net!newsserver.jvnc.net!igor.rutgers.edu!geneva.rutgers.edu!hedrick Mon Oct 11 08:53:57 EDT 1993 Article: 8306 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!newsserver.jvnc.net!igor.rutgers.edu!geneva.rutgers.edu!hedrick From: hedrick@geneva.rutgers.edu (Charles Hedrick) Newsgroups: comp.unix.solaris Subject: tuning Solaris 2 Message-ID: Date: 9 Oct 93 22:44:26 GMT Organization: Rutgers Univ., New Brunswick, N.J. Lines: 68 Status: RO We've recently been through a crisis that points up the importance of tuning with Solaris 2. I'm sharing our experience in hopes that others will be moved to do something similar. I've also requested our salesman to see if he can get Sun to issue a tuning document giving some of their experience. We have a Sparccenter 2000, which is intended for general student timesharing. The original goal was to run 2000 simultaenous users. I personally have some qualms about that. It's hard to see how we'd go about managing a system where "who" goes on for 100 screens, but that's the goal. At the moment I don't think we have even 100 users on at once, but it's still a fairly large-scale system. Anyway, this system has turning up a number of interesting scaling problem. We brought it to its knees last week after increasing the number of pty's. First, we found that with more than about 440 pty's, it won't come up. OK, so we left it at 440 pty's. It came up, but performance was abysmal. It turns out to be due to ttyname. A lot of our interactive software uses ttyname. Since it's never been thought of as a an expensive call, some software calls it several times during startup. Well, on our system ttyname was taking 10 sec. of real time. Simple interactive programs would take 30 sec. to start. The reason is sort of obvious: ttyname works by looking through all of /dev for a terminal that matches the one it was passed. In the old days with 32 pty's that wasn't a problem. But on a system with 440 pty's, linear search is not a reasonable approach. All those stats were bringing the system to its knees. Anyway, to make a long story short, we've managed to work around the problem by using the parameters someone recently posted. They came indirectly from Sun, as part of an NFS benchmark. They increase various caches. Presumably the effect is to keep that search of pty's from having to go to disk for every one. I include the parameters we used, in case you missed the earlier posting. As we've got plenty of memory on the machine, I haven't tried tweaking the parameters individually to see which ones have the most effect. These parameters took ttyname from 10 sec back down to its usual fraction of a second. Here's the previous posting I mentioned. Note that I'm currently using a reduced version of these parameters on my Classic, in hopes of fixing a serious performance problem on it. (I can't use these parameters as is, because they're clearly intended for a machine with more memory.) Over the course of a few days, the system slows down, until starting emacs take 20 seconds. I'm about to throw the thing out the window and get a PC. (My 486 box at home running Linux is wonderful.) No, I can't move to 4.1.3. I have to know if Solaris 2 can be made to work, since otherwise we've got to keep our faculty from buying any Sun hardware that might depend upon Solaris 2. ---- The *documented* list of /etc/system parameters is in the manual section "Administering SunOS 5.2 Security Performance and Accounting - Appendix A - Tuning Kernel Parameters". Some benchmarks require that tuning parameters are disclosed, so the SPEC LADDIS report "documents" some tweaks for Solaris 2.3alpha that helps NFS performance on an SS1000 using FDDI. set maxusers=32 ( some of the dependent parameters are set explicitly below) set ufs_ninode=32768 set ncsize=32768 set nsegkp=1024 set nhbuf=256 set p_nbuf=2048 set bufhwm=20480 set bf:bf_nbufs=2048 set bf:bf_nrmds1=256 From news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!elroy.jpl.nasa.gov!ames!koriel!male.EBay.Sun.COM!jethro.Corp.Sun.COM!exodus.Eng.Sun.COM!jetsun!robinson Mon Oct 11 08:55:33 EDT 1993 Article: 8318 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!elroy.jpl.nasa.gov!ames!koriel!male.EBay.Sun.COM!jethro.Corp.Sun.COM!exodus.Eng.Sun.COM!jetsun!robinson From: robinson@jetsun.Eng.Sun.COM (David Robinson) Newsgroups: comp.unix.solaris Subject: Re: tuning Solaris 2 Date: 10 Oct 1993 23:24:21 GMT Organization: SunSoft, Inc. Lines: 45 Message-ID: References: NNTP-Posting-Host: jetsun Status: RO In article hedrick@geneva.rutgers.edu (Charles Hedrick) writes: >The *documented* list of /etc/system parameters is in the manual >section "Administering SunOS 5.2 Security Performance and Accounting - >Appendix A - Tuning Kernel Parameters". > >Some benchmarks require that tuning parameters are disclosed, so the >SPEC LADDIS report "documents" some tweaks for Solaris 2.3alpha that >helps NFS performance on an SS1000 using FDDI. Beware of using LADDIS tunings. It is a highly specilaized benchmark that only mildly resembles a real world NFS server environment. Using it's tunings for an interactive timeshare machine may not produce the desired results. For example, for an entire LADDIS run not a single system call is made! As you can see below almost all of these tunings are to maximize disk caches. >set maxusers=32 ( some of the dependent parameters are set explicitly below) Normally maxusers is tuned to the amount of memory in the system. Most of the tunings are based off of this number. Think of it as a global knob. >set ufs_ninode=32768 This is the size of the in core inode cache. The larger it is the more memory is used and the less frequently you will have to go to disk to get a file's attributes. Tune this to reflect the number of files that are commonly in use. >set ncsize=32768 This is the directory name cache size. It caches file names in diretories to save disk I/O. Use vmstat -s, if the hit rate is below 90% you may want to tune it up. >set nhbuf=256 >set p_nbuf=2048 No longer does anything in Solaris 2.3. Does almost nothing in Solaris 2.2. >set bufhwm=20480 The maximum number of buffers for buffered disk I/O. >set bf:bf_nbufs=2048 >set bf:bf_nrmds1=256 These are tunings for the SUN Sbus FDDI card, if you don't have it it is not useful. From news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!pipex!uunet!news.univie.ac.at!fstgds15.tu-graz.ac.at!fstghp04.tu-graz.ac.at!schinagl Mon Oct 11 10:59:07 EDT 1993 Article: 8336 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!pipex!uunet!news.univie.ac.at!fstgds15.tu-graz.ac.at!fstghp04.tu-graz.ac.at!schinagl From: schinagl@fstghp04.tu-graz.ac.at (Hermann Schinagl) Newsgroups: comp.unix.solaris Subject: Re: Solaris 2.2 - xdm and chooser Date: 11 Oct 1993 14:34:44 GMT Organization: Graz University of Technology, Austria Lines: 13 Distribution: world Message-ID: <29bqu4INNn2f@fstgds15.tu-graz.ac.at> References: <29bert$hvq@fuhainf.fernuni-hagen.de> NNTP-Posting-Host: fstghp04.tu-graz.ac.at X-Newsreader: TIN [version 1.2 PL1] Status: RO Thomas Feuerstack (feuerst@bonsai.fernuni-hagen.de) wrote: : Has anyone out there made the same experiences or (which would be much better) has anyone got a : solution ? - I don't want to compile the whole MIT-Sources only to get a working chooser. Did you apply Pacth 101135-02 ? This patch deals with the broken xdm -- Ciao Hermann ------ Microsoft failing at something is not a sterling proof of impossibility ... From news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!pipex!uknet!mcsun!sun4nl!fwi.uva.nl!casper Thu Oct 14 10:02:46 EDT 1993 Article: 8433 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!pipex!uknet!mcsun!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris,comp.sys.sun.admin Subject: Re: SunOS 4.1.3C Performance on Sparc CLASSIC (?) Date: 13 Oct 1993 09:15:09 GMT Organization: FWI, University of Amsterdam Lines: 13 Distribution: inet Message-ID: <29ggut$rrd@mail.fwi.uva.nl> References: <1993Oct8.075218.914@uklirb.informatik.uni-kl.de> <29busdINN1re@galaxy.uci.agh.edu.pl> <29dorjINNgu4@signal.dra.hmg.gb> <29gd60INNocv@signal.dra.hmg.gb> NNTP-Posting-Host: adam.fwi.uva.nl Xref: news.umbc.edu comp.unix.solaris:8433 comp.sys.sun.admin:22227 Status: RO bilsby@hermes.mod.uk (David Bilsby) writes: > The dhrystones program was if I remember right just compiled with gcc >2.3.3 with -O2 or just -O. It was just the one IPX binary, ie. I did not >recompile on the LX for either Sol 1 or Sol 2. Is it the case that LX microSPARC >code can be optimized better if the compiler knows its a microSPARC and not an >IPX (?) SPARC processor.???? gcc 2.3.3 understands the -mv8 option. This will give you the mul and div etc instructions. Casper From news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!mcsun!sun4nl!fwi.uva.nl!casper Thu Oct 14 10:03:32 EDT 1993 Article: 8434 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!mcsun!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: catman -w core dumping Date: 13 Oct 1993 10:35:26 GMT Organization: FWI, University of Amsterdam Lines: 39 Distribution: world Message-ID: <29glle$soi@mail.fwi.uva.nl> References: <1993Oct13.090447.24299@iiasa.ac.at> NNTP-Posting-Host: adam.fwi.uva.nl Status: RO herb@iiasa.ac.at (Herb HASLER) writes: >When I try to build an index of our systems local man pages I get the >following message: ># catman -w >Output line too long. >Output line too long. >Segmentation Fault - core dumped ># >It would appear that it is choking on one of the man pages (or perhaps there >are too many pages). If anyone has seen this problem before and/or has >a simple way to fix it, drop me some email. catman/man and friend are sort of broken in Solaris 2.0 - 2.2. (Looking at the source code of man/catman I'm not surprised) The particular problem you're seeing is because catman requires very specific formatting of manual pages. (In particular, it likes the nroff source files to look like ``.SH NAME\n\n.SH xxxxx'' getNAME, the program that gets the synopsis out of all the man files, copies everything between .SH NAME and the following .SH directive and concatenates this to one line. It also wants the exact naem of the manual page to appear in the synopsis line. For some manual pages this amounts to several kilobytes, because of bad structuring. Such long lines are then fed in the makewhatis script to sed and awk. The message ``Output line too long'' is printed by sed. ``Segmentation Fault - core dumped'' is printed by the shell. It's awk that crashed. You can edit the whatis script and substitute nawk for awk. This still won't fix the problem. Only editing manual pages really helps. For kicks, try /usr/lib/getNAME /usr/openwin/man/man3/classtree.3. Piped through wc it gives: 1 757 4039 Casper From news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!gatech!gsusgi1.gsu.edu!gsusgi1.gsu.edu!matess Mon Oct 18 11:05:31 EDT 1993 Article: 8526 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!gatech!gsusgi1.gsu.edu!gsusgi1.gsu.edu!matess From: matess@gsusgi1.gsu.edu (Eliza Strickler) Subject: Solaris 2.2 direct printing to a SparcPrinter - help Message-ID: Organization: Georgia State University Date: Fri, 15 Oct 1993 16:02:19 GMT Lines: 39 Status: RO Is there something special I need to do to get my Sparc2 running Solaris 2.2 to print directly to its SparcPrinter? Here is the configuration: # lpstat -p sun -l printer sun now printing sun-2. enabled since Fri Oct 15 11:48:31 EDT 1993. available. Form mounted: Content types: any Printer types: PS Description: local sun printer Connection: direct Interface: /usr/lib/lp/model/standard On fault: write to root once After fault: continue Users allowed: (all) Forms allowed: (none) Banner not required Character sets: Default pitch: Default page size: 80 wide 66 long Default port settings: 19200 When I send a print job, nothing ever happens; the file just sits in the queue. Any help is appreciated. Elizabeth -- \|/--_ -_- ---- ### _- ---------------------- -0 -_- -- -__ %~- ____#0 _- Elizabeth Strickler |\ ^ 0\~ /\ /\ - 1-2245 |_(___/ \_ ||_________/ _/ |_/ \_ matess@gsusgi1.gsu.edu From news.umbc.edu!haven.umd.edu!purdue!lerc.nasa.gov!magnus.acs.ohio-state.edu!math.ohio-state.edu!howland.reston.ans.net!europa.eng.gtefsd.com!uunet!mcsun!sun4nl!fwi.uva.nl!casper Mon Oct 18 11:10:39 EDT 1993 Article: 8554 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!purdue!lerc.nasa.gov!magnus.acs.ohio-state.edu!math.ohio-state.edu!howland.reston.ans.net!europa.eng.gtefsd.com!uunet!mcsun!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: I worry about tcsh Date: 16 Oct 1993 11:36:05 GMT Organization: FWI, University of Amsterdam Lines: 15 Distribution: inet Message-ID: <29omb5$smd@mail.fwi.uva.nl> References: <29nec3$s6v@majestix.cs.uoregon.edu> NNTP-Posting-Host: adam.fwi.uva.nl Status: RO keith@cs.uoregon.edu (Keith Steiger) writes: >I ftp'ed to opcom.sun.ca to get tcsh for Solaris 2.2. First, I got the >binary; it dumped core. Then, I went back to get the sources. I uncompress >and untar that, and it turns out that the tar file contained a core file. >I'm not sure I even want to compile this.... We've been running tcsh 6.0.4 from tesla.ee.cornell.edu for quiote some time now and are quite happy with it. (You need a small fix to prevent it from eating too much typeahead, I'll see if I can find that somewhere, if anyone is interested) Casper From news.umbc.edu!haven.umd.edu!uunet!cs.utexas.edu!hermes.chpc.utexas.edu!jones Mon Oct 18 11:23:19 EDT 1993 Article: 8560 of comp.unix.solaris Newsgroups: comp.sys.sun.hardware,comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!uunet!cs.utexas.edu!hermes.chpc.utexas.edu!jones From: jones@chpc.utexas.edu (William L. Jones) Subject: Re: dirent problem Message-ID: <1993Oct16.185653.20055@chpc.utexas.edu> Organization: The University of Texas System - CHPC References: <93269.010536U16072@uicvm.uic.edu> <1993Oct16.014620.9713@greco-prog.fr> Date: Sat, 16 Oct 93 18:56:53 GMT Lines: 7 Xref: news.umbc.edu comp.sys.sun.hardware:12983 comp.unix.solaris:8560 Status: RO If you use dirent then you must not linked against /usr/ucblib libraries. You will link against /usr/ucblib if you use /usr/ucb/cc. If you link against /usr/ucblib you use direct in /usr/ucbinclude/sys/dir.h. From news.umbc.edu!haven.umd.edu!purdue!lerc.nasa.gov!usenet.ins.cwru.edu!agate!doc.ic.ac.uk!uknet!mcsun!sun4nl!fwi.uva.nl!casper Mon Oct 18 11:24:47 EDT 1993 Article: 8566 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!purdue!lerc.nasa.gov!usenet.ins.cwru.edu!agate!doc.ic.ac.uk!uknet!mcsun!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: I worry about tcsh Date: 17 Oct 1993 16:07:27 GMT Organization: FWI, University of Amsterdam Lines: 29 Distribution: inet Message-ID: <29rqjv$3q3@mail.fwi.uva.nl> References: <29nec3$s6v@majestix.cs.uoregon.edu> <29omb5$smd@mail.fwi.uva.nl> NNTP-Posting-Host: adam.fwi.uva.nl Status: RO Previously I wrote: >(You need a small fix to prevent it from eating too much typeahead, >I'll see if I can find that somewhere, if anyone is interested) Since a number of people requested it, here it is: *** tcsh-6.04/sh.h.org Thu Jul 8 06:52:56 1993 --- tcsh-6.04/sh.h Sun Sep 12 13:55:03 1993 *************** *** 270,276 **** #define CSWTCH _POSIX_VDISABLE #endif ! #if !defined(FIOCLEX) && defined(SUNOS4) # include #endif /* !FIOCLEX && SUNOS4 */ --- 270,276 ---- #define CSWTCH _POSIX_VDISABLE #endif ! #if !defined(FIOCLEX) && (defined(SUNOS4) || (defined(sun) && SYSVREL >= 4)) # include #endif /* !FIOCLEX && SUNOS4 */ From cosc4hf@menudo.uh.edu Tue Oct 19 18:50:19 EDT 1993 Article: 8597 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!usc!elroy.jpl.nasa.gov!swrinde!menudo.uh.edu!not-for-mail From: cosc4hf@menudo.uh.edu (Aditya Talwar) Newsgroups: comp.unix.solaris Subject: 2.3 early experiences Date: 18 Oct 1993 18:58:54 -0500 Organization: University of Houston Lines: 93 Distribution: world Message-ID: <29vaju$rn8@menudo.uh.edu> NNTP-Posting-Host: menudo.uh.edu Status: RO Grave warning, the following maybe be bit long :( but i need your advice quickly ! Last week, i was doing some practice Auto-net-remote 2.3 early installs. (the sysadmin's 2.x dream) Our setup is like this: Assume the cdrom from which i am going to boot off is on the 670, 4.1.3 server below: --- x.29.1 ------ Backbone ----------------- 670 server YP master 4.1.3, with 2.3cdrom | all disks are spread | by automounter x.29.2 interface | | -- x.29.10 my standalone client being installed To test, the Optional boot option on subnet, i dumped the sun4c architecture on my subnet's 670 server using command ./setup_install_server -b /export/install sun4c this went okay; further I added my client on the 670 on my subnet using command: ./add_install_client -s 670:/cdrom -c 670:/export/install_config client sun4c this command checks the NIS maps and tells me to update the bootparams map on the YP master as follows: client root=670:/cdrom/export/exec/kvm/sparc.sun4c.Solaris_2.3 \ install=670:/cdrom install_config=670:/export/install_info This message i think is wrong, i think, if one uses -b option with setup_install_server, it should prompt me with the correct path for root ! client root=670:/export/install/export/exec/kvm/sparc.sun4c.Solaris_2.3 \ install=670:/cdrom install_config=670:/export/install_info What does the net think of this? is my assumption correct. 2. Then i do a boot net - install, my machine does not respond to this, NIS ethers, bootparams, hosts maps have my client info. Then by trial and error and old notes, i figure I have to start rarpd le0 on my subnet's interface x.29.2 ! This is NOT MENTIONED in THE MANUAL, nor is there a mention about ./rm_install_client script, whats it purpose etc, i think there should be minimum docs on this in the install and config manual. 3. after the installation begins, the auto-install fails as there are complaints about c0t1d0 being a duplicate mount point, BTW, my host class files looks like this install_type initial_install system_type standalone partitioning existing filesys c0t3d0s0 40 / filesys c0t3d0s4 10 /var filesys c0t0d0s6 existing /usr filesys c0t0d0s5 existing /opt preserve filesys c0t1d0s6 existing /home preserve filesys c0t3d0s1 free swap filesys Yp master:/var/spool/mail - /var/mail rw,intr,bg cluster SUNWCall on changing c0t1d0 to filesys c0t1d0s6 existing /export/home preserve i get the message, (after wathcing the install again for 10mins :( c0t3d0s0 too small for configuration, so i change this to: filesys c0t1d0s6 existing /client_0 preserve then i get a message, after waiting again for 10 mins, Error: overlapping slices on disk c0t1s0 slice 2 and 6 ! All the above configs have been passed the ./check utility! BTW, if you use any of the 2.3 early install scripts, one cannot UMOUNT the cdrom on the 4.1.3 server unless it is rebooted, this is a bug! I am running out of time, I would appreciate your suggestions for the correct host_class config files or other watch out's for subnet install type configuration. Thanks aditya ************ yes, i speak for my own self******************* From brossard@sic.epfl.ch Tue Oct 19 18:51:36 EDT 1993 Article: 8606 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!xlink.net!scsing.switch.ch!swidir.switch.ch!epflnews!siisun!brossard From: brossard@siisun.epfl.ch (Alain Brossard EPFL-SIC/SII) Newsgroups: comp.unix.solaris Subject: Re: 2.3 early experiences Date: 19 Oct 1993 10:01:37 GMT Organization: Ecole Polytechnique Federale de Lausanne Lines: 83 Sender: brossard@siisun (Alain Brossard EPFL-SIC/SII) Distribution: world Message-ID: <2a0du1$n0@info.epfl.ch> References: <29vaju$rn8@menudo.uh.edu> Reply-To: brossard@sic.epfl.ch NNTP-Posting-Host: siisun.epfl.ch Status: RO In article <29vaju$rn8@menudo.uh.edu>, cosc4hf@menudo.uh.edu (Aditya Talwar) writes: |> this went okay; further I added my client on the 670 on my subnet |> using command: |> ./add_install_client -s 670:/cdrom -c 670:/export/install_config client sun4c |> |> this command checks the NIS maps and tells me to update the bootparams |> map on the YP master as follows: |> |> client root=670:/cdrom/export/exec/kvm/sparc.sun4c.Solaris_2.3 \ |> install=670:/cdrom install_config=670:/export/install_info You are correct, this is wrong all instances of "/cdrom" should be in your case /export/install. I haven't had any problems with 2.3 with respect with JumpStart, however I use my own script rather than "add_install_client". Maybe there is an extra option with add_install_client which tells it where the Solaris_2.3_EA is located? |> |> client root=670:/export/install/export/exec/kvm/sparc.sun4c.Solaris_2.3 \ |> install=670:/cdrom install_config=670:/export/install_info wrong here too ^^^^^^ replace with /export/install |> |> 2. Then i do a boot net - install, my machine does not respond to |> this, NIS ethers, bootparams, hosts maps have my client info. |> |> Then by trial and error and old notes, i figure I have to start |> rarpd le0 on my subnet's interface x.29.2 ! Normaly rarpd is started at boot time if there is a /etc/ethers with the -a flag (-a: all interfaces), however it uses the contents of the NIS ethers map. The add_install_client should restart the daemon. In the case of bootparams, the local file /etc/bootparams is consulted first and always re-read by the rpc.bootparamd daemon so no need to fuss with remaking NIS map. |> This is NOT MENTIONED in THE MANUAL You are using a BETA release, you should expect things to break and be missing, in no cases are you supposed to go to the net with these things. Haven't they made you sign a non-disclosure agreement? |> |> 3. after the installation begins, the auto-install fails as there |> are complaints about c0t1d0 being a duplicate mount point, |> filesys c0t1d0s6 existing /home preserve In Solaris 2, /home is under the automounter control. |> then i get a message, after waiting again for 10 mins, |> Error: overlapping slices on disk c0t1s0 slice 2 and 6 ! This is presumably a bug since slice 2 is supposed to be the whole disk and hence to be overlapping. You should be mentionning this to whomever provided you with this Beta/Early Access version of 2.3. |> BTW, if you use any of the 2.3 early install scripts, one cannot |> UMOUNT the cdrom on the 4.1.3 server unless it is rebooted, |> this is a bug! Most probably more a problem on your part since I had no problems umounting the CDROM. You mentionned that you started manually rarpd, could it be that you were under /cdrom at the time. You almost certainly started one or more programs while under /cdrom which are still running and hence stopping you from doing the umount. |> I am running out of time, I would appreciate your suggestions |> for the correct host_class config files or other watch out's |> for subnet install type configuration. When installing an Early Access version, by definition, you should expect problems and allocate plenty of time. You sound like you want to install a production environment, in that case you would be better off installing Solaris 2.2 for now. -- Alain Brossard, Ecole Polytechnique Federale de Lausanne, SIC/SII, MA-Ecublens, C.P. 121, CH-1015 Lausanne, Suisse, +41 21 693-2211 brossard@sic.epfl.ch From news.umbc.edu!europa.eng.gtefsd.com!darwin.sura.net!jhunix.hcf.jhu.edu!tomahawk.welch.jhu.edu!keithc Wed Oct 20 16:09:39 EDT 1993 Article: 8662 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!darwin.sura.net!jhunix.hcf.jhu.edu!tomahawk.welch.jhu.edu!keithc From: keithc@tomahawk.welch.jhu.edu. (Keith Christopher) Newsgroups: comp.unix.solaris Subject: Re: Pop Date: 20 Oct 1993 19:13:25 GMT Organization: Sun Microsystems, Inc. Lines: 8 Distribution: world Message-ID: <2a42klINNmah@jhunix.hcf.jhu.edu> References: <2a1kl8$9gk@caslon.CS.Arizona.EDU> Reply-To: keithc@tomahawk.welch.jhu.edu. NNTP-Posting-Host: tomahawk.welch.jhu.edu Status: RO Check opcom.sun.ca:/pub/binaries If it's not there, email me and I'll send it to you. Keith From news.umbc.edu!europa.eng.gtefsd.com!uunet!mcsun!sun4nl!fwi.uva.nl!casper Thu Oct 21 09:20:35 EDT 1993 Article: 8682 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!uunet!mcsun!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: perl 4.036 compile problem Date: 21 Oct 1993 08:51:59 GMT Organization: FWI, University of Amsterdam Lines: 11 Distribution: world Message-ID: <2a5ijf$em3@mail.fwi.uva.nl> References: NNTP-Posting-Host: adam.fwi.uva.nl Status: RO jtcox@gumby.syr.edu writes: >I am trying to get perl 4.036 to compile using gcc-2.4.5 on my Sparc >10 running Solaris 2.2. During the configure script I told it to use >the "hints" file for Solaris 2.1. After doing a 'make depend' and then >'make' I get: Get rid of the ucb stuff (all, library & includes), don't use vfork and perhaps you'll need to compile util.c and tutil.c w/o -O Casper From news.umbc.edu!europa.eng.gtefsd.com!uunet!mcsun!sun4nl!fwi.uva.nl!casper Thu Oct 21 09:21:18 EDT 1993 Article: 8684 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!uunet!mcsun!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris,comp.unix.admin Subject: Re: too many lpNet processes Date: 21 Oct 1993 08:57:15 GMT Organization: FWI, University of Amsterdam Lines: 24 Distribution: world Message-ID: <2a5itb$eoi@mail.fwi.uva.nl> References: <2a3oon$m7p@msuinfo.cl.msu.edu> NNTP-Posting-Host: adam.fwi.uva.nl Xref: news.umbc.edu comp.unix.solaris:8684 comp.unix.admin:16158 Status: RO vuppala@parsley.msu.edu (Vibhavasu Vuppala) writes: >Hi, >Env : sparc-10, 2.2, mandatory patches, 101025-07 >There are about 50 lpnet prodcesses running on the print server. lpshut >make them disappear but they come back after starting the LP sceduler. >It is creating a process for each client we have. This is due to the timeout you have said in the /etc/lp/Systems file. To get rid of the lpNet processes, you'll need to put a" :x:-:s5:-:0:10:-:-: in the /etc/lp/Systems file on each of your print clients. This makes the clients drop the connection when the work is done. We also noticed that the print service became more reliable once we used a ``never'' timeout on the server and a 0 timeout on the clients. Casper From news.umbc.edu!europa.eng.gtefsd.com!paladin.american.edu!darwin.sura.net!news.duc.auburn.edu!usenet Fri Oct 22 13:44:49 EDT 1993 Article: 8712 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!paladin.american.edu!darwin.sura.net!news.duc.auburn.edu!usenet From: elling@eng.auburn.edu (Richard Elling) Subject: Re: too many lpNet processes Message-ID: <1993Oct21.220639.12446@news.duc.auburn.edu> Sender: usenet@news.duc.auburn.edu (News Account) Nntp-Posting-Host: netman.eng.auburn.edu Reply-To: elling@eng.auburn.edu Organization: Auburn University Engineering References: <2a3oon$m7p@msuinfo.cl.msu.edu> Date: Thu, 21 Oct 1993 22:06:39 GMT Lines: 33 Status: RO In article m7p@msuinfo.cl.msu.edu, vuppala@parsley.msu.edu (Vibhavasu Vuppala) writes: >Env : sparc-10, 2.2, mandatory patches, 101025-07 > >There are about 50 lpnet prodcesses running on the print server. lpshut >make them disappear but they come back after starting the LP sceduler. >It is creating a process for each client we have. > >Any help will be greatly appreciated. I kinda doubt there are 50. By default lpsched will croak if you have too many printers (the magic number seems to be about 48) and do an lpstat. It opens a file descriptor for every printer and since the default file descriptor limit is 64, it will crash unceremoniously. You can fix it by increasing the file descriptor limit in the /etc/init.d/lp script as follows: ulimit -n 256 [ -f /usr/lib/lpsched ] && /usr/lib/lpsched This really ought to make it into one of the lp patches, better yet change lpsched so it doesn't leave bunches of open files, or at least doesn't crash when it can no longer open more files. For my own network, I'd rather not have so many printers -- what a waste of money. Alas, every time I recommend saving $200k/year by not buying more printers I get shouted down :-( --- Richard Elling, Manager of Network Support, Engineering Administration Auburn University richard.elling@eng.auburn.edu kb4hb.ampr.org Where is Auburn? Go to Margaritaville, take a right, and go until you have to ask yourself where in the hell you are. From news.umbc.edu!haven.umd.edu!darwin.sura.net!martha.utcc.utk.edu!martha.utcc.utk.edu!rich Fri Oct 22 13:57:35 EDT 1993 Article: 8735 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!martha.utcc.utk.edu!martha.utcc.utk.edu!rich From: rich@martha.utcc.utk.edu (Steve Rich) Subject: Re: Increasing PTY's on Solaris X86 Message-ID: <1993Oct21.234443.21970@martha.utcc.utk.edu> Keywords: pty Sender: rich@utkux1.utk.edu Organization: University of Tennessee Computing Center References: <1993Oct21.122035.23928@visionware.co.uk> Date: Thu, 21 Oct 1993 23:44:43 GMT Lines: 17 Status: RO >From the Sun technical people, this is the way (this got me 255 ptys): (in /etc/system) set pt_cnt=255 set npty=255 set sadcnt=512 set nautopush=255 (and then a boot w/reconfigure) Steve -- ---------------------------------------------- Steve Rich rich@utkux1.utk.edu "I think the past/the past is behind us/ be real confusing if not/but anyway... "But Anyway", Blues Traveller ---------------------------------------------- From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!spool.mu.edu!uwm.edu!msuinfo!parsley!vuppala Fri Oct 22 13:58:29 EDT 1993 Article: 8738 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!spool.mu.edu!uwm.edu!msuinfo!parsley!vuppala From: vuppala@parsley.msu.edu (Vibhavasu Vuppala) Newsgroups: comp.unix.solaris Subject: Re: too many lpNet processes Date: 22 Oct 1993 13:46:47 GMT Organization: Michigan State University, CPS Department Lines: 37 Sender: vuppala@parsley (Vibhavasu Vuppala) Distribution: world Message-ID: <2a8o87$19s@msuinfo.cl.msu.edu> References: <2a3oon$m7p@msuinfo.cl.msu.edu> <1993Oct21.220639.12446@news.duc.auburn.edu> NNTP-Posting-Host: parsley.cps.msu.edu Status: RO In article <1993Oct21.220639.12446@news.duc.auburn.edu>, elling@eng.auburn.edu (Richard Elling) writes: |> In article m7p@msuinfo.cl.msu.edu, vuppala@parsley.msu.edu (Vibhavasu Vuppala) writes: |> >Env : sparc-10, 2.2, mandatory patches, 101025-07 |> > |> >There are about 50 lpnet prodcesses running on the print server. lpshut |> >make them disappear but they come back after starting the LP sceduler. |> >It is creating a process for each client we have. |> > |> >Any help will be greatly appreciated. |> |> I kinda doubt there are 50. By default lpsched will croak if you have |> too many printers (the magic number seems to be about 48) and do an |> lpstat. It opens a file descriptor for every printer and since the |> default file descriptor limit is 64, it will crash unceremoniously. |> You can fix it by increasing the file descriptor limit in the |> /etc/init.d/lp script as follows: |> ulimit -n 256 |> [ -f /usr/lib/lpsched ] && /usr/lib/lpsched |> no, the problem lies with the number of clients and not the printers. each client initiates a connection with each printer server which results in a lpNet on the client as well as the server side. the clients do not drop the connection since timeout in /etc/lp/Systems is set to "n" (never timeout). So the server has an lpNet for each client and the clients have an lpNet for each printer. this can be solved by having the timeout as 0 in /etc/lp/Systems. - vasu PS : 50 printers !! I am frustrated to death with just one !! From news.umbc.edu!haven.umd.edu!darwin.sura.net!news.duc.auburn.edu!usenet Sun Oct 24 17:37:57 EDT 1993 Article: 8751 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!news.duc.auburn.edu!usenet From: elling@eng.auburn.edu (Richard Elling) Subject: Re: xdm does not log to /etc/utmp Message-ID: <1993Oct22.184140.22402@news.duc.auburn.edu> Sender: usenet@news.duc.auburn.edu (News Account) Nntp-Posting-Host: netman.eng.auburn.edu Reply-To: elling@eng.auburn.edu Organization: Auburn University Engineering References: <2a87mlINN82u@fstgds15.tu-graz.ac.at> Date: Fri, 22 Oct 1993 18:41:40 GMT Lines: 30 Status: RO In article 2a87mlINN82u@fstgds15.tu-graz.ac.at, schinagl@fstghp04.tu-graz.ac.at (Hermann Schinagl) writes: >Hi > >Is there any way under Solaris 2.2 I can force xdm ( patch 101135-02 ) >to log to /var/adm/utmp or /var/adm/wtmp, so that I can get info of users, >which logged in via xdm ? I don't know if thats a bug or a feature, but >it seems to me more a bug. > >I have searched for some resources, which might xdm to force to do that, >but haven't found any. XDM itself won't do that. >I have heard of sessreg, and already compiled it, it seems to me better >to do that via xdm sessreg is what I use for such purposes. The problem is that the MIT sessreg doesn't understand SVR4 utmpx/wtmpx entries. I have modified sessreg to do so and placed the diffs on ftp.eng.auburn.edu in /put/xdm/xdm.config.solaris2.1.tar.Z if you are interested. Some of the stuff there is rather dated, but I probably won't do much about it since coesession is looming on the horizon... --- Richard Elling, Manager of Network Support, Engineering Administration Auburn University richard.elling@eng.auburn.edu kb4hb.ampr.org Where is Auburn? Go to Margaritaville, take a right, and go until you have to ask yourself where in the hell you are. From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!vixen.cso.uiuc.edu!bradley.bradley.edu!bradley.bradley.edu!not-for-mail Sun Oct 24 17:38:28 EDT 1993 Article: 8754 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!vixen.cso.uiuc.edu!bradley.bradley.edu!bradley.bradley.edu!not-for-mail From: pwh@bradley.bradley.edu (Pete Hartman) Newsgroups: comp.unix.solaris Subject: Re: x_cdplayer Date: 22 Oct 1993 16:32:12 -0500 Organization: Bradley University Lines: 12 Distribution: inet Message-ID: <2a9jgs$fsk@bradley.bradley.edu> References: <2a4e6p$icj@elroy.jpl.nasa.gov> NNTP-Posting-Host: bradley.bradley.edu Status: RO What you really want to get is workman. There are instructions for dealing with Solaris (it was originally written under SunOS) and how to set it up with the volume manager which is nifty. Archie should be able to tell you where to find it. -- Pete Hartman Bradley University pwh@bradley.bradley.edu Who else but a Mennonite would be alone, naked in a garden with a woman, and only want an apple? From news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!pipex!uknet!ukc!eagle.ukc.ac.uk!db15 Sun Oct 24 17:40:18 EDT 1993 Article: 8764 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!pipex!uknet!ukc!eagle.ukc.ac.uk!db15 From: db15@ukc.ac.uk (Damiano Bolla) Newsgroups: comp.unix.solaris Subject: Re: Should local software go in /opt or /opt/local? Date: Sat, 23 Oct 93 11:31:02 GMT Organization: Computing Lab, University of Kent at Canterbury, UK. Lines: 47 Sender: db15@ukc.ac.uk Distribution: inet Message-ID: <6226@eagle.ukc.ac.uk> References: <293p6a$ifh@dux.dundee.ac.uk> <29u6tsINNrsh@phage.cshl.org> <1993Oct21.133155.25957@dragon.acadiau.ca> Reply-To: db15@ukc.ac.uk (Damiano Bolla) NNTP-Posting-Host: eagle.ukc.ac.uk Status: RO In article <1993Oct21.133155.25957@dragon.acadiau.ca> peter@dragon.acadiau.ca (Peter Steele) writes: >>>The point of /opt is for mounting SVR4 software packages, the rationale > >>i would make /opt be an automount point. that way i can place packages >>anywhere and just update the map. is anyone doing this? does it seem >>correct to you? > >This sounds like a good approach to me. Could you elaborate? I had to set up a mix of Solaris 2.2 and SunOS 4.1.2 systems and therefore I had to solve the problem of giving access to a mixed env. different local compiled programs. What I have done is: create an automount map auto_nfsd that holds data for the directory /nfsd The map is as follows ftp -ro,nosuid,soft arianna_hyde:/arianna3/ftp local4.1 -ro,nosuid,soft arianna:/usr/local local2.2 -ro,nosuid,soft paola:/ext2/local sunvision -ro,nosuid,soft arianna:/arianna3/sunvision-1.2 dsp -nosuid,soft arianna:/dspdisk opt -ro,nosuid,soft paola:/opt usr2.2 -ro,nosuid,soft paola:/usr And of course depending on what machine you are you make the right simlink to the right point. Ex: if I am a SunOS 4.1.2 and I want the local I will do ln -s /nfsd/local4.1 /usr/local So when I try to get to /usr/local I automount the right place. This allows you to move around things and in the same time support different OS. Also.. since is an indirect map automount will notice changes in the table as soon as you make them.. and since only needed directory are mounted you don't have much troubles if you do find / -name something -print Seems to work ok so far... :-) Damiano From news.umbc.edu!europa.eng.gtefsd.com!paladin.american.edu!news.univie.ac.at!fstgds15.tu-graz.ac.at!fstghp04.tu-graz.ac.at!schinagl Sun Oct 24 17:42:37 EDT 1993 Article: 8774 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!paladin.american.edu!news.univie.ac.at!fstgds15.tu-graz.ac.at!fstghp04.tu-graz.ac.at!schinagl From: schinagl@fstghp04.tu-graz.ac.at (Hermann Schinagl) Newsgroups: comp.unix.solaris Subject: Re: xdm does not log to /etc/utmp Date: 24 Oct 1993 10:01:42 GMT Organization: Graz University of Technology, Austria Lines: 25 Distribution: inet Message-ID: <2adjq6INN8nv@fstgds15.tu-graz.ac.at> References: <2a87mlINN82u@fstgds15.tu-graz.ac.at> <1993Oct22.184140.22402@news.duc.auburn.edu> NNTP-Posting-Host: fstghp04.tu-graz.ac.at X-Newsreader: TIN [version 1.2 PL1] Status: RO Richard Elling (elling@eng.auburn.edu) wrote: : In article 2a87mlINN82u@fstgds15.tu-graz.ac.at, schinagl@fstghp04.tu-graz.ac.at (Hermann Schinagl) writes: : > : >Is there any way under Solaris 2.2 I can force xdm ( patch 101135-02 ) : >to log to /var/adm/utmp or /var/adm/wtmp, so that I can get info of users, : sessreg is what I use for such purposes. The problem is that the MIT : sessreg doesn't understand SVR4 utmpx/wtmpx entries. I have modified : sessreg to do so and placed the diffs on ftp.eng.auburn.edu in : /put/xdm/xdm.config.solaris2.1.tar.Z if you are interested. Some of : the stuff there is rather dated, but I probably won't do much about it : since coesession is looming on the horizon... Thanks a lot, it works fine now for me. I have tar'd the patched source and a very short description how to use it together, and put it at our ftp server: ftp.tu-graz.ac.at:/pub/util/sessreg-sol22.tar.gz -- Ciao Hermann ------ Microsoft failing at something is not a sterling proof of impossibility ... From news.umbc.edu!europa.eng.gtefsd.com!uunet!caen!usenet.coe.montana.edu!news.uoregon.edu!news.uoregon.edu!meyer Mon Oct 25 11:02:10 EDT 1993 Article: 8795 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!uunet!caen!usenet.coe.montana.edu!news.uoregon.edu!news.uoregon.edu!meyer From: meyer@cambium.uoregon.edu (David M. Meyer 503/346-1747) Newsgroups: comp.unix.solaris Subject: traceroute patches Date: 25 Oct 1993 13:15:32 GMT Organization: University Network Services, University of Oregon, Eugene, OR 97403 Lines: 203 Distribution: world Message-ID: NNTP-Posting-Host: cambium.uoregon.edu Status: RO Here are some patches for traceroute that make it run on Solaris 2.x. The whole archive can be found on ftp.uoregon.edu on pub/Solaris2.x/src. Thanks, Dave David M. Meyer Voice: 503/346-1747 Senior Network Engineer Pager: 503/342-9458 Office of University Computing FAX: 503/346-4397 Computing Center Internet: meyer@ns.uoregon.edu University of Oregon 1225 Kincaid Eugene, OR 97403 ----- #!/bin/sh # This is a shell archive (produced by shar 3.49) # To extract the files from this archive, save it to a file, remove # everything above the "!/bin/sh" line above, and type "sh file_name". # # made 10/25/1993 13:14 UTC by meyer@cambium # # existing files will NOT be overwritten unless -c is specified # # This shar contains: # length mode name # ------ ---------- ------------------------------------------ # 3676 -rw-r--r-- patch01 # 183 -rw-r--r-- traceroute.h # # ============= patch01 ============== if test -f 'patch01' -a X"$1" != X"-c"; then echo 'x - skipping patch01 (File already exists)' else echo 'x - extracting patch01 (Text)' sed 's/^X//' << 'SHAR_EOF' > 'patch01' && *** Makefile.orig Mon Oct 25 05:59:27 1993 --- Makefile Mon Oct 25 06:13:54 1993 *************** *** 27,33 **** X X # ARCH_K should be set to the output you get when you type "arch -k" X # on your system. (Should be one of sun3, sun3x, sun4, or sun4c.) ! ARCH_K= sun4c X X # Set DESTDIR to the directory in which the traceroute executable will be X # installed. /usr/etc is a good place to put a network debugging tool such --- 27,33 ---- X X # ARCH_K should be set to the output you get when you type "arch -k" X # on your system. (Should be one of sun3, sun3x, sun4, or sun4c.) ! ARCH_K= sun4m X X # Set DESTDIR to the directory in which the traceroute executable will be X # installed. /usr/etc is a good place to put a network debugging tool such *************** *** 37,50 **** X X # You shouldn't need to change anything below this line. X ! CC= cc ! CFLAGS = -O ! KFLAGS= -fsoft -DKERNEL -D$(ARCH_K) X X # At the moment, the INCL variable isn't really needed for anything. ! INCL = -I. ! LIBS = ! MANDIR= /usr/man/man8 X X # SUN_INCL_DIR should be set to the full pathname of the system include X # directory in which ip_icmp.h lives. For SunOS 4.0.3 and 4.0.3c, this is --- 37,51 ---- X X # You shouldn't need to change anything below this line. X ! INSTALL = /usr/ucb/install ! CC = gcc ! CFLAGS = -O ! # KFLAGS= -fsoft -DKERNEL -D$(ARCH_K) X X # At the moment, the INCL variable isn't really needed for anything. ! INCL = -I. ! LIBS = -lnsl -lsocket ! MANDIR = /usr/local/man/man8 X X # SUN_INCL_DIR should be set to the full pathname of the system include X # directory in which ip_icmp.h lives. For SunOS 4.0.3 and 4.0.3c, this is *************** *** 57,63 **** X SYS_OBJ_DIR= /sys/$(ARCH_K)/OBJ X X ! all: traceroute ip_icmp.o raw_ip.o X X traceroute: traceroute.c X $(CC) $(CFLAGS) $(INCL) -o traceroute traceroute.c $(LIBS) --- 58,64 ---- X SYS_OBJ_DIR= /sys/$(ARCH_K)/OBJ X X ! all: traceroute X X traceroute: traceroute.c X $(CC) $(CFLAGS) $(INCL) -o traceroute traceroute.c $(LIBS) *************** *** 69,78 **** X $(CC) $(CFLAGS) $(KFLAGS) $(INCL) -c raw_ip.c X X install: all ! install -o root -g bin -m 4755 traceroute ${DESTDIR} ! install -o root -g staff -m 755 traceroute.8 ${MANDIR} ! install -o root -g staff -m 444 ip_icmp.o $(SYS_OBJ_DIR) ! install -o root -g staff -m 444 raw_ip.o $(SYS_OBJ_DIR) X @ echo "" X @ echo "Binaries and man page installed." X @ echo "" --- 70,79 ---- X $(CC) $(CFLAGS) $(KFLAGS) $(INCL) -c raw_ip.c X X install: all ! $(INSTALL) -o root -g bin -m 4755 traceroute ${DESTDIR} ! $(INSTALL) -o root -g staff -m 755 traceroute.8 ${MANDIR} ! # $(INSTALL) -o root -g staff -m 444 ip_icmp.o $(SYS_OBJ_DIR) ! # $(INSTALL) -o root -g staff -m 444 raw_ip.o $(SYS_OBJ_DIR) X @ echo "" X @ echo "Binaries and man page installed." X @ echo "" *** traceroute.c.orig Mon Oct 25 05:58:35 1993 --- traceroute.c Mon Oct 25 06:11:47 1993 *************** *** 191,197 **** --- 191,201 ---- X X #include X #include + #ifdef __STDC__ + #include + #else X #include + #endif X #include X X #include *************** *** 208,213 **** --- 212,219 ---- X #include X #include X + #include "traceroute.h" + X #define MAXPACKET 65535 /* max ip packet size */ X #ifndef MAXHOSTNAMELEN X #define MAXHOSTNAMELEN 64 *************** *** 394,400 **** --- 400,412 ---- X Printf(usage); X exit(1); X } + + + #ifdef __STDC__ + setvbuf(stderr, NULL, _IOLBF, 0); + #else X setlinebuf (stdout); + #endif /* __STDC__ */ X X (void) bzero((char *)&whereto, sizeof(struct sockaddr)); X to->sin_family = AF_INET; SHAR_EOF chmod 0644 patch01 || echo 'restore of patch01 failed' Wc_c="`wc -c < 'patch01'`" test 3676 -eq "$Wc_c" || echo 'patch01: original size 3676, current size' "$Wc_c" fi # ============= traceroute.h ============== if test -f 'traceroute.h' -a X"$1" != X"-c"; then echo 'x - skipping traceroute.h (File already exists)' else echo 'x - extracting traceroute.h (Text)' sed 's/^X//' << 'SHAR_EOF' > 'traceroute.h' && X #ifdef __STDC__ #define bcopy(src,dest,len) (memmove((dest), (src), (len))) #define bzero(dest,len) (memset((dest), (char)0, (len))) #define index(s,c) strchr((s),(c)) #endif SHAR_EOF chmod 0644 traceroute.h || echo 'restore of traceroute.h failed' Wc_c="`wc -c < 'traceroute.h'`" test 183 -eq "$Wc_c" || echo 'traceroute.h: original size 183, current size' "$Wc_c" fi exit 0 From ta@trinity.uk.sun.com Tue Oct 26 15:43:02 EDT 1993 Article: 8808 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!uunet!olivea!koriel!newscast.West.Sun.COM!seven-up.East.Sun.COM!uk-usenet.uk.sun.com!news From: ta@trinity.uk.sun.com (Tim Addison - Sun UK - SE Cambridge) Newsgroups: comp.unix.solaris Subject: Re: out of ptys Date: 25 Oct 1993 17:28:11 GMT Organization: Sun Microsystems (UK) Lines: 29 Distribution: world Message-ID: <2ah2bb$12l@uk-usenet.uk.sun.com> References: <29bguj$125l@rs6000.bham.ac.uk> Reply-To: ta@trinity.uk.sun.com NNTP-Posting-Host: trinity.uk.sun.com Status: RO Hi Jonathon, You can increase the number of pty's by making the following change to /etc/system. e.g. Set pt_cnt=250 and then reboot with the -r flag. --Tim In article 125l@rs6000.bham.ac.uk, harleyjw@sun1.bham.ac.uk (Jon Harley) writes: > My SS10 (solaris2.2) keeps running out of ptys now that students have > started using it. The man page for pts says they are configured at install > time - can I add extra ptys now or is it too late? > > > /Jonathan > ______________________ _______________________________________________ ____ > / -- Jonathan Harley -- / Computer Officer (Systems Group), Academic __\_ / > / J.W.Harley@bham.ac.uk / Computing Service, University of Birmingham \ X/ > /_______________________/ _Phone: 021 414 6575________Fax: 021 414 3952___\/ > > From rehmann@cscs.ch Tue Oct 26 15:43:13 EDT 1993 Article: 8809 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!xlink.net!scsing.switch.ch!pobox!usenet From: rehmann@cscs.ch (Rene M. Rehmann) Subject: Re: Problem with Xfig 2.1 pl 8 under Solaris2.1 Message-ID: <1993Oct25.171109.24435@cscs.ch> Sender: usenet@cscs.ch (NEWS Manager) Nntp-Posting-Host: vevey.cscs.ch Reply-To: rehmann@cscs.ch Organization: Centro Svizzero di Calcolo Scientifico (CSCS), Manno, Switzerland References: Date: Mon, 25 Oct 1993 17:11:09 GMT Lines: 31 Status: RO In article 93Oct25095321@eastman1.mit.edu, lfk@eastman1.mit.edu (Lee F. Kolakowski) writes: > >Dear Wizards of Solaris2.x and Xfig. > >Over the weekend I ftp'd xfig2.1pl8 from pith.uoregon.edu and >compliled it. > >Details: > >SunClassic 16M 16" mr color >Graphics system is OpenWindows as dist (Not X111.R5) >GCC 2.3.3 built with boot version from prep. > >Basically no problems with compilation >errors >bcopy in two files - fixed with #define bcopy memcpy >and a few indirection warnings. > I use #define bcopy(s1, s2, n) memmove(s2, s1, n) (< --- note that s1 and s2 are exchanged) and it works fine. Rene --- Rene Rehmann phone: +41 91 50 82 34 Section for Research and Development (SeRD) fax : +41 91 50 67 11 Swiss Scientific Computing Center CSCS email: rehmann@cscs.ch Via Cantonale, CH-6928 Manno, Switzerland From Thomas.Tornblom@Nexus.Comm.SE Tue Oct 26 15:48:51 EDT 1993 Article: 8844 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!uunet!pipex!sunic!seunet!seunet!comm!news!Thomas.Tornblom From: Thomas.Tornblom@Nexus.Comm.SE (Thomas Tornblom) Subject: Re: Automatic Startup of Workman In-Reply-To: robm@aplcore.jhuapl.edu's message of Tue, 19 Oct 1993 18: 10:47 GMT Message-ID: Sender: news@nexus.comm.se Organization: Communicator Nexus AB References: Date: Tue, 26 Oct 1993 12:53:42 GMT Lines: 34 Status: RO In article robm@aplcore.jhuapl.edu (Robert E. Mentle 40-213) writes: Has anyone gotten workman to start up automatically including a workman data base using volume management. I am using gcc-2.4.5 and Sol 2.2 and after talking to several people found that vold will work if we use the -R flag for compilation, but I have never found out how to also set the WORKMANDB environment value, within volume management. Any help? My /etc/rmmount.conf looks like this, and workman starts up whenever I insert an audio CD. ------------- # @(#)rmmount.conf 1.2 92/09/23 SMI # # Removable Media Mounter configuration file. # # File system identification ident hsfs ident_hsfs.so cdrom ident ufs ident_ufs.so cdrom floppy ident pcfs ident_pcfs.so floppy # Actions action cdrom action_workman.so /usr/openwin/bin/workman action cdrom action_filemgr.so action floppy action_filemgr.so -- Real life: Thomas Törnblom Email: Thomas.Tornblom@Nexus.Comm.SE Snail mail: Communicator Nexus AB Phone: +46 18 171814 Box 857 Fax: +46 18 696516 S - 751 08 Uppsala, Sweden From news.umbc.edu!europa.eng.gtefsd.com!paladin.american.edu!news.univie.ac.at!fstgds15.tu-graz.ac.at!fstghp04.tu-graz.ac.at!schinagl Thu Oct 28 11:44:00 EDT 1993 Article: 8920 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!paladin.american.edu!news.univie.ac.at!fstgds15.tu-graz.ac.at!fstghp04.tu-graz.ac.at!schinagl From: schinagl@fstghp04.tu-graz.ac.at (Hermann Schinagl) Newsgroups: comp.unix.solaris Subject: Re: Xdm (Openwin 3.2 Solaris 2.2) not working ? Date: 27 Oct 1993 21:13:19 GMT Organization: Graz University of Technology, Austria Lines: 16 Distribution: world Message-ID: <2amo9fINN25b@fstgds15.tu-graz.ac.at> References: <2ajas8$en2@dlsn31> NNTP-Posting-Host: fstghp04.tu-graz.ac.at X-Newsreader: TIN [version 1.2 PL1] Status: RO Alisdair Craig (ahcraig@mepsi.mobil.com) wrote: : I am trying to setup Xdm as supplied with Solaris 2.2 (openwin 3.2) to manage a group of Xterminals. I gan get the nice xdm "Welcome" message to come up via xdmcp. But when a user tries to login, xdm dies with "error (pid 231): Unknown session exit code 2816 from process 304" (where 304 PID of the spawned xdm). : Anyone seen this ? Does Xdm as supplied work ? No, stock xdm under Solaris 2.2 does NOT work properly, you may get patch 101135-02, which helps. Ciao Hermann ------ Microsoft failing at something is not a sterling proof of impossibility ... From news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!pipex!uknet!mcsun!sun4nl!fwi.uva.nl!casper Thu Oct 28 11:45:59 EDT 1993 Article: 8931 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!howland.reston.ans.net!pipex!uknet!mcsun!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: creating many files using "find" Date: 28 Oct 1993 09:50:02 GMT Organization: FWI, University of Amsterdam Lines: 30 Distribution: world Message-ID: <2ao4ka$9sk@mail.fwi.uva.nl> References: <2amumd$phi@elroy.jpl.nasa.gov> NNTP-Posting-Host: adam.fwi.uva.nl Status: RO Timothy Dry writes: > I'm trying to create a particular file in every directory >within a particular tree. I believe I should be able to do >the following: >smogalert(~)% find . -type d -exec touch {}/xyz \; >touch: /xyz cannot create >touch: /xyz cannot create >touch: /xyz cannot create >^C > Unfortunately, as you can see, it tries to create the files >in "/" as if the braces were never used. The directory in >the example has many directories under it. The system is >Solaris 5.2 on a SparcServer 1000. The {} is only expanded in the argument to exec if it appears as a lone argument. Not when it appears as part of an argument. Fix: upgrade to GNU find: % /usr/bin/find /var/tmp -type d -exec echo '{}/foo' \; {}/foo {}/foo % /opt/gnu/bin/find /var/tmp -type d -exec echo '{}/foo' \; /var/tmp/foo /var/tmp/unbundled/foo Casper From news.umbc.edu!haven.umd.edu!nova.umd.edu!not-for-mail Fri Oct 29 02:05:26 EDT 1993 Article: 8951 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!nova.umd.edu!not-for-mail From: caruso@nova.umd.edu (John Caruso) Newsgroups: comp.unix.solaris Subject: Re: Patch 100999-33 Problems Date: 28 Oct 1993 13:22:05 -0400 Organization: University of Maryland University College Lines: 22 Distribution: inet Message-ID: <2aov3t$91s@nova.umd.edu> References: <2ao65b$ocn@ratatosk.uninett.no> NNTP-Posting-Host: nova.umd.edu Status: RO In article <2ao65b$ocn@ratatosk.uninett.no>, Rune Groenaas wrote: >I have a SparcServer 1000, running Solaris 2.2. Some patches have been >installed, including the 100999-33. But after installing this patch, the >program "sysinfo" (Public Domain), crashes the computer, with a long >traceback message. After uninstalling this patch, the program worked. > >Has anybody else had problems with this patch???? 100999-33 was very unstable; we had three system panics in the week after installing it. The current revision is -43 (that's as of the beginning of this week, and considering the frequency of updates to the kernel patch it's probably higher now). We currently have -38 installed and it has been running well for over two weeks. You may want to try updating to one of these later revisions and seeing if the problem goes away. --------------------------------------------------------------------- John Caruso caruso@nova.umd.edu Unix/VMS System Administrator caruso@UMUC (Bitnet) University of Maryland University College Work: (301) 985-7447 --------------------------------------------------------------------- From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!agate!ames!koriel!newscast.West.Sun.COM!news2me.EBay.Sun.COM!exodus.Eng.Sun.COM!appserv.Eng.Sun.COM!ffeine.Sun.COM!wsanders Tue Nov 2 13:44:21 EST 1993 Article: 9110 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!agate!ames!koriel!newscast.West.Sun.COM!news2me.EBay.Sun.COM!exodus.Eng.Sun.COM!appserv.Eng.Sun.COM!ffeine.Sun.COM!wsanders From: wsanders@ffeine.Sun.COM (Wiley Sanders) Newsgroups: comp.unix.solaris Subject: Re: Automounter 2.3 changes? Date: 1 Nov 1993 18:53:43 GMT Organization: A Jug of Wine, a Loaf of Bread, and Fifty Thousand Dollars Lines: 20 Distribution: na Message-ID: References: NNTP-Posting-Host: ffeine Keywords: automounter Status: RO There are only three really substantive changes from a user's point of view: 1) When you modify your maps, you reread the maps by typing "automount" instead of killing and restarting it. 2) The "/etc/auto_XXXXX" files should have permissions 644. If they are executable, the automounter will try to run them as executable shell scripts. 3) No more soft links to /tmp_mnt, this might be a gotcha if some of your applications or environment variables have references to /tmp_mnt/something. -w -- "There is absolutely no truth to the rumor that all employees are going to be required to have lobotomies . . . at least at the prices we were quoted." -Dilbert From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!pipex!uknet!mcsun!sun4nl!fwi.uva.nl!casper Fri Nov 5 01:47:39 EST 1993 Article: 9229 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!pipex!uknet!mcsun!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: /etc/system defaults Date: 4 Nov 1993 11:54:00 GMT Organization: FWI, University of Amsterdam Lines: 24 Distribution: world Message-ID: <2baqgo$137@mail.fwi.uva.nl> References: <2bac17$me8@disuns2.epfl.ch> NNTP-Posting-Host: adam.fwi.uva.nl Status: RO chris@lslsun7.epfl.ch (Christian Iseli) writes: >Anyone knows (short of using adb on the kernel) where to see the defaults >used in the /etc/system file? And what are the available tuning >parameters? sysdef lists some, but there seems to be more... >For example, it says in the man page of system(4) that maxusers is >"a useful tuning parameter"... But what is the default value (seems >to be 8, from using adb)? Nowhere in TFM does it seem to have a >complete list of those "useful tuning parameters". The default value isn't 8. The default value is scaled with the amount of memory. The maxusers parameter scales a number of other parameters, all of which can be set individually. One of the most commonly asked parameters: set pt_cnt = * number of SysV ptys Personnaly recommended: set nfs:nfs_portmon = 1 * for increased fileserver security. Casper From news.umbc.edu!europa.eng.gtefsd.com!uunet!Germany.EU.net!news.sni.de!nanette!nanette!news Fri Nov 5 05:53:51 EST 1993 Article: 9265 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!uunet!Germany.EU.net!news.sni.de!nanette!nanette!news From: Josef Moellers Newsgroups: comp.unix.solaris,comp.unix.misc Subject: Re: Where is bcopy,bzero,bcmp? Date: 5 Nov 1993 08:10:32 GMT Organization: Siemens Nixdorf Informationssysteme AG, Paderborn, Germany Lines: 31 Distribution: inet Message-ID: <2bd1poINNc8d@nanette.sto.pdb.sni.de> References: NNTP-Posting-Host: uranium.sto.pdb.sni.de Xref: news.umbc.edu comp.unix.solaris:9265 comp.unix.misc:11072 Status: RO In blair@world.std.com (Blair P. Houghton) writes: >In article hofu@renoir.nlp.physik.th-darmstadt.de (Joachim Holzfuss) writes: >>Helo, >>I'm trying to compile some program with >>gcc 2.4.5 on Solaris 2.1 and just can't figure out >>in what library the following subroutines are located: >> >>bcopy,index,bzero,rindex,bcmp Sorry, I don't find the original posting, so I followup on this folloup. If you cannot find the functions on your machine, the following defines will do: # include # define bcopy(s,d,l) memcpy((d),(s),(l)) # define bzero(d,l) memset((d),0,(l)) # define bcmp(s1,s2,l) memcmp((s2),(s1),(l)) # include # define index(s,c) strchr((s),(c)) # define rindex(s,c) strrchr((s),(c)) Look into the manual for memory(3C) and string(3C). There are handwritten notes next to the mem- and str- functions that tell you (at least they are in my manual B-{) -- | Josef Moellers | c/o Siemens Nixdorf Informationssysteme AG | | USA: mollers.pad@sni-usa.com | Abt. STO-XS 113 | Riemekestrasse | | !USA: mollers.pad@sni.de | Phone: (+49) 5251 835124 | D-4790 Paderborn | From news.umbc.edu!europa.eng.gtefsd.com!uunet!iphase.com!iex!peters Sat Nov 6 18:54:54 EST 1993 Article: 9280 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!uunet!iphase.com!iex!peters From: peters@iex.UUCP (James A. Peters) Subject: Re: /etc/system defaults Message-ID: <1993Nov5.165956.11450@iex.uucp> Keywords: kernel parameters pt_cnt npty Sender: usenet@iex.uucp (USENET news) Organization: IEX Corporation References: <2bac17$me8@disuns2.epfl.ch> <2baqgo$137@mail.fwi.uva.nl> Date: Fri, 5 Nov 1993 16:59:56 GMT Lines: 56 Status: RO In article <2baqgo$137@mail.fwi.uva.nl> casper@fwi.uva.nl (Casper H.S. Dik) writes: >chris@lslsun7.epfl.ch (Christian Iseli) writes: > >>Anyone knows (short of using adb on the kernel) where to see the defaults >>used in the /etc/system file? And what are the available tuning >>parameters? sysdef lists some, but there seems to be more... > >>For example, it says in the man page of system(4) that maxusers is >>"a useful tuning parameter"... But what is the default value (seems >>to be 8, from using adb)? Nowhere in TFM does it seem to have a >>complete list of those "useful tuning parameters". > >The default value isn't 8. The default value is scaled it appears the default GENERIC value is 8, pg 116 SunOS 5.2 Administering Security, Performance, and Accounting. >with the amount of memory. The maxusers parameter scales >a number of other parameters, all of which can be set individually. the parameters ncallout, ufs_ninode, ncsize, max_nprocs, ndquot, and maxuprc are affected by maxusers, reference pg 116 and Table 6-1 pg 117. >One of the most commonly asked parameters: > > set pt_cnt = * number of SysV ptys the following is a bit of pseudo tty wisdom from bill wisner. none of this stuff is covered in the manuals/answerbook, go figure. there is supposed to be man page for pty(4) as given by man xterm(1) but it is vapor. "ntpy controls the number of BSD-style compatibility pty devices. These are only present for compatibility and none of the programs that are part of the Solaris system as shipped use them. You cannot go above 48 of these devices in the stock 2.2 release; Sun has issued a patch that corrects this problem and allows you to configure as many as you want. However, you must use mknod to manually create the device nodes. pt_cnt controls the number of pts devices. These are System V pseudo-ttys. in.telnetd and in.rlogind use these. Any software which was properly written to work on System V uses these. If you change this number and do a boot -r, the kernel will build the correct number of device nodes in /dev/pts. (Note, though, that it will not remove old ones, so if you *lower* pt_cnt you will have to manually remove the extra device nodes.)" >Personnaly recommended: > > set nfs:nfs_portmon = 1 * for increased fileserver security. the complete list of tunable kernel parameters is obtained by using /usr/ccs/bin/nm /kernel/unix. /usr/sbin/sysdef -i shows current values assigned to kernel parameters. reference Appendix A pg 197. regards, james, peters@iex.com From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!vixen.cso.uiuc.edu!sdd.hp.com!nigel.msen.com!well!wisner Sat Nov 6 18:56:18 EST 1993 Article: 9286 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!vixen.cso.uiuc.edu!sdd.hp.com!nigel.msen.com!well!wisner From: wisner@well.sf.ca.us (Bill Wisner) Subject: Re: /etc/system defaults Message-ID: Keywords: kernel parameters pt_cnt npty Sender: news@well.sf.ca.us Nntp-Posting-Host: well.sf.ca.us Organization: The Whole Earth 'Lectronic Link, Sausalito, CA References: <2bac17$me8@disuns2.epfl.ch> <2baqgo$137@mail.fwi.uva.nl> <1993Nov5.165956.11450@iex.uucp> Date: Fri, 5 Nov 1993 20:13:48 GMT Lines: 44 Status: RO Casper Dik says (of maxusers): >>The default value isn't 8. The default value is scaled >>with the amount of memory. The maxusers parameter scales >>a number of other parameters, all of which can be set individually. James A. Peters says: >it appears the default GENERIC value is 8, pg 116 SunOS 5.2 Administering >Security, Performance, and Accounting. My kernel says: # adb -k /kernel/unix /dev/mem physmem 7d9e maxusers? maxusers: maxusers: 20 = unimp 0x1e 20 hexadecimal is 32 decimal which is apparently the default value for maxusers in Solaris 2.2. So who's right? >there is supposed to be man page for pty(4) as given by man xterm(1) but it >is vapor. As far as I can determine the BSD-style pty devices in Solaris 2.2 are documented nowhere. If you really want to know, go packing back to your SunOS 4.1 manuals. It really is a good idea to make your software use the pts devices instead of ptys, though - the pts drivers are a standard part of SVR4 while the BSD prt driver was apparently grafted on by Sun as a compatibility afterthought. The pts drivers are probably much better supported. >the complete list of tunable kernel parameters is obtained by using >/usr/ccs/bin/nm /kernel/unix. /usr/sbin/sysdef -i shows current values >assigned to kernel parameters. reference Appendix A pg 197. sysdef -i only shows a handful of the tunable parameters -- those few that Sun officially condones the use of. I find it quite amusing that the manual instructs users to use nm to get a list of tunable parameters from the kernel, as nm will list every kernel symbol that can possibly be set, most with unpredictable and possibly quite disastrous results. That bit was a real gaffe on the part of Sun's technical writers. -- Bill Wisner Gryphon Gang San Francisco CA 94116 From news.umbc.edu!europa.eng.gtefsd.com!uunet!mcsun!sun4nl!fwi.uva.nl!casper Mon Nov 8 01:07:08 EST 1993 Article: 9323 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!uunet!mcsun!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: /etc/system defaults Date: 7 Nov 1993 11:26:32 GMT Organization: FWI, University of Amsterdam Lines: 52 Distribution: world Message-ID: <2bim18$1n0@mail.fwi.uva.nl> References: <2bac17$me8@disuns2.epfl.ch> <2baqgo$137@mail.fwi.uva.nl> <1993Nov5.165956.11450@iex.uucp> NNTP-Posting-Host: adam.fwi.uva.nl Keywords: kernel parameters pt_cnt npty Status: RO wisner@well.sf.ca.us (Bill Wisner) writes: }Casper Dik says (of maxusers): }>>The default value isn't 8. The default value is scaled }>>with the amount of memory. The maxusers parameter scales }>>a number of other parameters, all of which can be set individually. }James A. Peters says: }>it appears the default GENERIC value is 8, pg 116 SunOS 5.2 Administering }>Security, Performance, and Accounting. }My kernel says: }maxusers: 20 = unimp 0x1e }20 hexadecimal is 32 decimal which is apparently the default value }for maxusers in Solaris 2.2. So who's right? I am. It says so in the source. First, I believe what I observe, then I believe the source and the least reliable source of information is the documentation. }As far as I can determine the BSD-style pty devices in Solaris 2.2 are }documented nowhere. If you really want to know, go packing back to }your SunOS 4.1 manuals. It really is a good idea to make your software }use the pts devices instead of ptys, though - the pts drivers are a }standard part of SVR4 while the BSD prt driver was apparently grafted }on by Sun as a compatibility afterthought. The pts drivers are probably }much better supported. And they're much easier to use. Nor do you need to hardcode the number of ptys you expect on a system, nor do you need to be root, nor are there any problems with race conditions during allocations. BSD ptys are a hack. I don't miss them. }sysdef -i only shows a handful of the tunable parameters -- those few }that Sun officially condones the use of. I find it quite amusing that }the manual instructs users to use nm to get a list of tunable parameters }from the kernel, as nm will list every kernel symbol that can possibly }be set, most with unpredictable and possibly quite disastrous results. }That bit was a real gaffe on the part of Sun's technical writers. It also lists all the functions and what not in the kernel. nm prints the names of the symbols, as they appear in /dev/ksyms. Variables such as nfs:nfs_portmon show up as nfs_portmon, but you can't set nfs_portmon. Often you need to specify the module name first. Casper From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!pipex!uknet!mcsun!sun4nl!fwi.uva.nl!casper Mon Nov 8 07:58:12 EST 1993 Article: 9339 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!pipex!uknet!mcsun!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: set nfs:nfs_portmon = 1? (was Re: /etc/system defaults) Date: 8 Nov 1993 09:11:13 GMT Organization: FWI, University of Amsterdam Lines: 26 Distribution: inet Message-ID: <2bl2fh$74n@mail.fwi.uva.nl> References: <2bac17$me8@disuns2.epfl.ch> <2baqgo$137@mail.fwi.uva.nl> <2bk7vdINN362@cbs.ksu.ksu.edu> NNTP-Posting-Host: adam.fwi.uva.nl Status: RO tar@cbs.ksu.ksu.edu (Tim Ramsey) writes: >casper@fwi.uva.nl (Casper H.S. Dik) writes: >>Personnaly recommended: >> set nfs:nfs_portmon = 1 * for increased fileserver security. >What does this do? Checks that all NFS requests come from a port # < 1024. What does this buy you? Simple, w/ nfs_portmon set to 0, any user can implement and run their own nfs client code on their workstation. This allows them access to all files that can possibly accessed from this workstation. (I.e, effectively ``su user'' for any user id when it comes to NFS file access) Most crackers nowadays have ready-to-run copies of such NFS user-mode clients. In SunOS 4.x, the variable nfs_portmon controls the same. (In rc.local this variable is set to 1 if /etc/security/passwd.adjunct exists, you can turn it on by default.) Casper From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!sol.ctr.columbia.edu!jabba.ess.harris.com!mlb.semi.harris.com!controls.ccd.harris.com!apnsund!kap Thu Nov 11 14:58:30 EST 1993 Article: 9497 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!sol.ctr.columbia.edu!jabba.ess.harris.com!mlb.semi.harris.com!controls.ccd.harris.com!apnsund!kap From: kap@sun.com (Kurt Post) Subject: Re: xntpd on Solaris 2.2 Message-ID: <1993Nov11.002741.25682@ccd.harris.com> Sender: usenet@ccd.harris.com Reply-To: kap@sun.com Organization: Sun Microsystems, Inc. References: <2bqrmvINNl6b@maz4.sma.ch> Date: Thu, 11 Nov 1993 00:27:41 GMT Lines: 18 Status: RO In article 2bqrmvINNl6b@maz4.sma.ch, gdb@sma.ch (Guido Della Bruna) writes: >Hello, >does anybody know if a version of xntpd for Solaris 2.2 exists? >I have tried with xntp version 3 with the configuration for Solaris 2.1, but >it seems that some kernel variables changed their name in Solaris 2.2. The >daemon compiles fine but it gives a runtime error. > What you need is xntp3.3. You can get it from louie.udel.edu in the /pub/ntp directory (xntp3.3.tar.Z). This version of xntp supports Solaris 2.2. All you have to do is remove the -DKERNEL_PLL from DEFS_LOCAL in the Config.local file and build the the distribution. Kurt From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!xlink.net!fauern!rrze.uni-erlangen.de!nessy!kardel Fri Nov 12 10:28:07 EST 1993 Article: 9550 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!xlink.net!fauern!rrze.uni-erlangen.de!nessy!kardel From: kardel@nessy.informatik.uni-erlangen.de (Frank Kardel) Newsgroups: comp.unix.solaris Subject: Re: xntp for solaris 2.2 Date: Fri, 12 Nov 1993 11:10:15 GMT Organization: CSD., University of Erlangen Distribution: inet Message-ID: <2bvqunEc7m@uni-erlangen.de> References: <5081@rztsun.rz.tu-harburg.de> NNTP-Posting-Host: faui45x.informatik.uni-erlangen.de Lines: 14 Status: RO rztjk@tu-harburg.dbp.de (Jrgen Kabelitz) writes: >You need the xntp3.3b-no-us-des version. You need to say that you got that from: ftp.uni-erlangen.de:/pub/NTP/* This version is xntp3.3b with lib/authdes.c replaced by an australian version. All US citizens can use plain xntp3.3b.tar.Z from louie.udel.edu. Frank Kardel (time@informatik.uni-erlangen.de) From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!EU.net!sun4nl!fwi.uva.nl!casper Fri Nov 19 12:44:04 EST 1993 Article: 9803 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!EU.net!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: Using adb on Solaris loadable kernel Date: 19 Nov 1993 08:54:16 GMT Organization: FWI, University of Amsterdam Lines: 22 Distribution: usa Message-ID: <2ci1jo$e64@mail.fwi.uva.nl> References: <1968@pacvax.UUCP> NNTP-Posting-Host: adam.fwi.uva.nl Status: RO fields@pacvax.UUCP (Daniel A. Fields) writes: >Is it possible to use "adb" to examine variable in kernel modules on >Solaris. How is this done. Ive heard that the only satisfactory way to >do it is to boot kadb and break into the debugger to examine variables. It's actually quite easy: adb -k /dev/ksyms /dev/mem This will let you examine the kernel while it's running. There are limits to what you can get away with, but this works essentially the same as under SunOS 4.x. /dev/ksyms is the device that presents you with a snapshot of the system's symbol table at the time you open it. /kernel/unix does not include all symbols you may want to examine. Casper From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!cs.utexas.edu!swrinde!elroy.jpl.nasa.gov!ncar!noao!rstevens Fri Nov 19 13:07:24 EST 1993 Article: 9822 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!cs.utexas.edu!swrinde!elroy.jpl.nasa.gov!ncar!noao!rstevens From: rstevens@noao.edu (W. Richard Stevens) Subject: Re: Keepalive using tcp/ip Message-ID: <1993Nov19.175413.28340@noao.edu> Summary: use ndd(1) Sender: news@noao.edu Nntp-Posting-Host: gemini.tuc.noao.edu Organization: National Optical Astronomy Observatories, Tucson, AZ, USA References: <2civ6s$1hc@zebedee.ingres.co.uk> Distribution: inet Date: Fri, 19 Nov 1993 17:54:13 GMT Lines: 15 Status: RO > How do you set the keepalive value on solaris 2.x? I can't seem to work > our how to build a kernel (if you can). The keepalive value seems to be > shown in /usr/include/netinet/tcp_timer.h and is set to (120*60*PR_SLOWHZ) > where PR_SLOWHZ is 2 (ie twice a second). TCP/IP configurable parameters are set/examined using ndd(1). What you want is ndd /dev/tcp tcp_keepalive_interval and the value printed is 7200000 (2 hours in ms). Change that to whatever you'd like, *but* it's a system-wide parameter and will affect all TCP connections that set the keepalive option. Rich Stevens (rstevens@noao.edu) From news.umbc.edu!europa.eng.gtefsd.com!uunet!spool.mu.edu!sgiblab!wetware!iggy.GW.Vitalink.COM!ns!hammer!dap Sat Nov 20 16:54:22 EST 1993 Article: 9826 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!uunet!spool.mu.edu!sgiblab!wetware!iggy.GW.Vitalink.COM!ns!hammer!dap From: dap@hammer.network.com (Dave Peterson) Subject: Re: Keepalive using tcp/ip Message-ID: <1993Nov19.184221.14946@ns.network.com> Sender: news@ns.network.com Nntp-Posting-Host: hammer Reply-To: dap@hammer.network.com Organization: Network Systems Corporation References: <2civ6s$1hc@zebedee.ingres.co.uk> Date: Fri, 19 Nov 93 18:42:21 GMT Lines: 37 Status: RO In article 1hc@zebedee.ingres.co.uk, jonm@ingres.co.uk (Jon Machtynger) writes: > >How do you set the keepalive value on solaris 2.x? I can't seem to work >our how to build a kernel (if you can). The keepalive value seems to be >shown in /usr/include/netinet/tcp_timer.h and is set to (120*60*PR_SLOWHZ) >where PR_SLOWHZ is 2 (ie twice a second). > >If you can't build a kernel, and /etc/system is just used on boot, will >modifying this value change the keepalive (as other things in this >area) or will an 'update' program need to be run to make this known. >I would have thought there would be some utility to do this. Answerbook is >not too helpful in this area. > >Thanks in advance > > >Jon > > >-- >-- >Jon Machtynger (jonm@ingres.com) Tech Consultant (Instn and Perf. Team) >Ingres Technology Centre. Wharfedale Rd, >Winnersh, Wokingham. Berks UK. (0734)-496481 Try: ndd -set /dev/tcp tcp_keepalive_interval --- =================================================================== Dave Peterson phone : 612-493-1008 Senior Engineer Network Systems Corp. email: dap@network.com =================================================================== From can@cam-orl.co.uk Mon Nov 22 13:22:22 EST 1993 Article: 9884 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!pipex!pavo.csi.cam.ac.uk!nntp-serv.cam.ac.uk!can From: can@cam-orl.co.uk (Cosmos Nicolaou) Subject: solaris 2.3: good and not so good Message-ID: Sender: news@infodev.cam.ac.uk (USENET news) Nntp-Posting-Host: shrimp.cam-orl.co.uk Organization: Olivetti Research Ltd. Date: Mon, 22 Nov 1993 15:36:34 GMT Lines: 52 Status: RO Here are my initial experiences with solaris 2.3 - some good and some not so good! The installation went pretty smoothly - I did a complete re-install (using a second hard disk to back up my root disk) and all went well. On the good side, /kernel/drv/st.conf allows you add 3'rd party drives, here's my entry for a sony sdt-4000 8G dat tape. It took about 15 minutes to get it all going - including testing ufsdump and ufsrestore - modunload certainly saves a lot of time over hacking st_conf.c, relinking and rebooting as in 4.1. tape-config-list = "SONY SDT-4000", "", "sony_sdt_4000"; sony_sdt_4000 = 1, 0x34, 0, 0x679, 4, 0, 0, 0x8c, 0x8c, 0x8c, 3; I understand all of this apart from the density values (0x8c) which I copied from the man page. Any ideas? Printing - still as broken as ever, the release notes say that sun will junk the v.4 code in the next release - about time too! I've managed to get it to work, but for some reason I always get an error report when printing postscript files! The files print fine, so I can live with it, but it this clearly a bug. The problem is caused by postio outputing messages to stdout out, such as ``printer startup'' as it goes about its business. The standard output of postio (or postcomm or whatever filter you use) is fed to lp.tell's standard input - these messages then trick lp.tell into thinking that postio is reporting an error - it complains to the sysadmin, but meanwhile the file is printed. Any ideas what's going on - I've applied patch 101317-02, but it made no difference? Other than that the upgrade went pretty smoothly apart from finger/brain trouble on my part! Cheers, Cos. Cosmos Nicolaou email: can@dsl-cam.co.uk Distributed Solutions Ltd (DSL), can@cam-orl.co.uk 64 Garden Walk, Cambridge, tel: +44 223 566 300 (DSL) United Kingdom CB4 3EN fax: +44 223 566 301 (DSL) tel: +44 223 343 307 (Olivetti) fax: +44 223 313 542 (Olivetti) --- From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!math.ohio-state.edu!magnus.acs.ohio-state.edu!csn!news.sinet.slb.com!rag.austin.nam.slb.com!slcs.slb.com!bemont Wed Dec 1 13:01:26 EST 1993 Article: 9993 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!math.ohio-state.edu!magnus.acs.ohio-state.edu!csn!news.sinet.slb.com!rag.austin.nam.slb.com!slcs.slb.com!bemont From: bemont@slcs.slb.com (Les Bemont) Newsgroups: comp.unix.solaris Subject: Re: solaris 2.3 upgrade from 2.2 Date: 24 Nov 1993 16:29:13 GMT Organization: Schlumberger Laboratory for Computer Science Lines: 82 Distribution: world Message-ID: <2d024p$3is@rag.austin.nam.slb.com> References: NNTP-Posting-Host: grimm.austin.nam.slb.com Status: RO In article , can@cam-orl.co.uk (Cosmos Nicolaou) writes: |> |> Hi, |> |> has anyone tried to upgrade from 2.2 to 2.3 and if so did you experience |> any problems - in particular will the 2.3 boot cdrom recognise that |> you've already got 2.2 installed and perform an incremental upgrade or |> will it just go ahead and re-format your disk for you? |> Here's some info we got from Sun. Hope it helps. --Les >From Robert.Rebiejo@Corp.Sun.COM Wed Nov 24 10:03:40 1993 Date: Wed, 24 Nov 93 08:05:42 PST From: Robert.Rebiejo@Corp.Sun.COM (Robert Rebiejo) To: holle@austin.wireline.slb.com Subject: Regarding Upgrade from 2.2 -> 2.3 Cc: Robert.Rebiejo@Corp.Sun.COM Content-Length: 1848 The criterion for whether or not Upgrade is a choice in suninstall. Enjoy! ----- Begin Included Message ----- --> One customer boots the cd and the Upgrade option does not appear as an --> Installation choice, The book says it checks to make sure you have a Solaris --> 2.1 disk, Exactly what is it checking for? Does it actually look at the disk --> partitions or do some kind of grep on OS version? An unpgradeable disk is defined to have the file /var/sadm/softinfo/INST_RELEASE This file must have the following entries: OS=Solaris VERSION=2.3 REV=0 VERSION must be > 2.0 for upgrade to work. We assume that the / file system is on slice 0 of the disk. A separate /var file system is dealt with properly provided that / is mountable. Here is the algorithm for upgrades to 2.2 - I'm sure it is pretty close for 2.3 (this was provided by Lori Alt a while back). for each disk on the system { attempt to mount slice 0 at /a; /* note that if slice 0 is already * mounted somewhere, or for some other reason the mount * would fail, this attempt will fail */ if (mount succeeded) { read /a/etc/vfstab to determine whether /var is a separate file system; if ( /var is on a separate file system) mount that file system at /a/var; read the file /a/var/sadm/softinfo/INST_RELEASE; if (the read succeeds and the value of VERSION in the file is 2.1 or later) { add disk to list of upgradeable disks } unmount /a/var if it was mounted ummount /a } } if (there is at least one disk in the list of upgradeable disks) present the upgrade button So, if the user had deleted the /var/sadm/softinfo/INST_RELEASE file, the disk would not be recognized as upgradeable. ----- End Included Message ----- From news.umbc.edu!haven.umd.edu!purdue!lerc.nasa.gov!usenet.ins.cwru.edu!gatech!news-feed-2.peachnet.edu!darwin.sura.net!jhunix.hcf.jhu.edu!skysrv!szgyula Wed Dec 1 13:03:37 EST 1993 Article: 10005 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!purdue!lerc.nasa.gov!usenet.ins.cwru.edu!gatech!news-feed-2.peachnet.edu!darwin.sura.net!jhunix.hcf.jhu.edu!skysrv!szgyula From: szgyula@skysrv.pha.jhu.edu (Gyula Szokoly) Newsgroups: comp.unix.solaris Subject: kernel tuning -- STATUS (far from SUMMARY) Date: 24 Nov 1993 22:25:25 GMT Organization: Johns Hopkins University Lines: 61 Distribution: inet Message-ID: <2d0n0lINNt6r@jhunix.hcf.jhu.edu> NNTP-Posting-Host: skysrv.pha.jhu.edu Keywords: kernel, /etc/system, configuration, solaris Status: RO I was posting about a week ago about kernel configuration concerns. At the moment I still don't feel good. What I figured out so far (thanks to the many who responded and to a sunsolve document): 1. The kernel is configured automagically by doing a 'boot -r'. This should take care of the necessary drivers, adjusts the tunable parameters. The first part seems to be OK. It does what it should, if for any reason you don't want to load a driver (but you HAVE the hardware), you can easily figure out what to do: the '/usr/sbin/modinfo -i' should tell you what drivers are loaded. You can play with the 'modunload' command to figure out if you can really remove it. The second part (tunable parameters) are not so clean. 2. You can get the names of these parameters using the 'nm' on '/kernel/unix'. After greping for 'OBJT' and 'GLOB', I ended up with more than 1000 names. Some of them are obvious (like maxusers), but I guess at least 95% is not very interesting. So this is not the way. The claim by Sun that the automatic kernel configuration does what it should is just not true. There are some built in thumb of rules but that's all. I recall VMS 4.7 (at least 5 years old): it was able to collect information and adjust the parameters according to the usage. The Solaris is not capable to do this. 3. I was playing with 'sar'. To use this, you can do many things, but the ab sugegsted way is pretty easy: become sys ('su sys') and edit the crontab file. The entries are already there so it should be easy. Than wait a few days to collect info. To examine the collected info, do a 'sar -A | more'. This will report EVERY collected info for the current day (do a 'man sar' to figure out how to get the previous days). Thanks Sun, the report makes sense, i.e. the statistics are reported in groups so you don't need a 2000 char long display. Here you might catch what is the source of your problems. 4. Well, you are pretty close. Go to 2.) and try to figure out which parameter should be adjusted. Sometimes it's easy, sometimes it's nearly hopeless. 5. Check out the current setting. Start adb: adb -k /dev/ksyms /dev/mem (You don't get a prompt). To check a parameter (e.g. maxusers): maxusers/D maxusers/X The first gives the decimal, the second the hex value. Depending on the variable, you might need different commands (do a man adb). 6. Guess the correct value. Put it into /etc/system. The file is self descriptive, e.g. to increase the maxusers, insert: set maxusers=16 You can also disable the loading of modules here. After you are done, you have to do a halt and a 'boot -r' (the touch /reconfigure seems to be broken in some cases), than a reboot (I'm not quite sure if this is needed but since you already restarted the machine, an additional restart should not hurt). Still needed: what are the usefull kernel parameters? The maxusers is obviously one of them, while s_cache_flushpagectx probably is not. Hope somebody can enlighten me, Gyula From news.umbc.edu!eff!news.kei.com!uuspew.uu.net!uunet!EU.net!julienas!news.fnet.fr!ilog!parquier Thu Dec 2 16:50:37 EST 1993 Article: 10070 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!eff!news.kei.com!uuspew.uu.net!uunet!EU.net!julienas!news.fnet.fr!ilog!parquier From: carre@ens-lyon.fr (Giles Carre) Subject: Re: mandatory patches for Solaris 2.3 ? Message-ID: <1993Nov26.140958.28113@ilog.fr> Sender: parquier@ilog.fr (Pierre Parquier) Nntp-Posting-Host: whisky.ens-lyon.fr Reply-To: carre@ens-lyon.fr Organization: Ecole Normale Superieure de Lyon Date: 26 Nov 1993 11:25:06 GMT Lines: 36 Status: RO In article 000DE2E4@infrasys.ascom.ch, BUECHE@infrasys.ascom.ch (Charles Bueche) writes: > > This morning, I heard that there are mandatory patches to > Solaris 2.3. Does anybody along you has a list ? > Mandatory patches for Solaris 2.3 (on 22 Nov 1993), from Sun Hotline -------------------------------------------------------------------- patch id 101316 patch id 101318 patch id 101319 patch id 101327 patch id 101329 patch id 101344 patch id 101346 patch id 101347 At this time, there are 29 patches (not only mandatory) for Solaris 2.3 (see sunsolve). About Solaris 2.2 patches ------------------------- . do not install any of them after a 2.3 full install . or absolutely remove all of them before an upgrade to 2.3 From news.umbc.edu!europa.eng.gtefsd.com!gatech!swrinde!sgiblab!rtech!amdahl!netcomsv!netcomsv!stokely!celeste Thu Dec 2 17:08:14 EST 1993 Article: 10182 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!gatech!swrinde!sgiblab!rtech!amdahl!netcomsv!netcomsv!stokely!celeste From: celeste@xs.com (Celeste Stokely) Subject: Solaris 2.x modem/terminal how-to, Rev 10/9/93 Message-ID: <1993Nov30.142955.11423@xs.com> Sender: celeste@xs.com (Celeste Stokely) Organization: Stokely Consulting, Mountain View, CA Date: Tue, 30 Nov 1993 14:29:55 GMT Lines: 578 Status: RO (It seems to be time for me to post this again, judging from the number of requests I'm getting for it.) The Answerbook instructions on setting up modems and terminals under Solaris 2.x are wrong, because they turn a few back-quotes into forward quotes. The Answerbook's procedure is also not very easy to understand. Here is a more cookbook approach, with a couple of useful scripts you can easily modify to suit your needs. Enjoy! ..Celeste Stokely Unix System Administration Consultant, Stokely Consulting EMAIL: celeste@xs.com Voice Line: 415-967-6898 / FAX: 415-967-0160 USMAIL Address: Stokely Consulting 211 Thompson Square / Mountain View CA 94043 Oct 9 07:57 1993 Setting up Terminals and Modems under Solaris 2.x Page 1 Configuring Terminals and Modems under Solaris 2.x ================================================== Revision 10/9/93 INDEX ----- Setting up a terminal on Solaris 2.x.................Section A.....Page 2 BEFORE YOU BEGIN: SET UP THE TERMINAL............Section A.1...Page 2 ENABLING SOLARIS FOR A TERMINAL..................Section A.2...Page 2 add_terminal script............................................Page 3 Setting Up a Bidirectional Modem on Solaris 2.x......Section B.....Page 4 BEFORE YOU BEGIN: SET UP THE MODEM...............Section B.1...Page 4 ENABLING SOLARIS FOR A BIDIRECTIONAL MODEM.......Section B.2...Page 4 add_modem script...............................................Page 6 More on tip and /etc/remote..........................Section C.....Page 7 /etc/remote example............................................Page 8 Setting other serial port modes......................Section D.....Page 9 /etc/ttydefs examples..........................................Page 10 Random notes of various uses of serial ports.........Section E.....Page 11 Parity in UUCP..................................................Page 11 Flow control with UUCP..........................................Page 12 Oct 9 07:57 1993 Setting up Terminals and Modems under Solaris 2.x Page 2 A. Setting up a terminal on Solaris 2.x ********************************************************************* *** IMPORTANT NOTES: *** *** 1. All single quotes in these directions are BACK QUOTES, *** *** not FORWARD QUOTES, except where specifically noted. *** *** *** *** 2. Be sure and execute these commands as root, from a *** *** BOURNE shell (sh), not ksh, tcsh, or csh. (If you put the *** *** commands into a Bourne shell script, that's fine, too.) *** *** *** *** 3. If, after following these instructions, you get *** *** terminal failures, and you're not using a script, then *** *** make sure that root's default shell is /sbin/sh and not *** *** csh. *** ********************************************************************* A.1 BEFORE YOU BEGIN: SET UP THE TERMINAL - Set the terminal for the speed you wish (example below assumes you are using 9600 baud) - The only lines you really need in the terminal-to-Sun cable are 2, 3, and 7. Be sure to cross lines 2 and 3 in your cable. - Set the terminal for Xon/Xoff flow control - Set the terminal for 7 bits, Even Parity. (1 stop bit, if you have that setting.) If you want a different parity or word length, see Section D of this document. A.2 ENABLING SOLARIS FOR A TERMINAL: - Save the following script as add_terminal - make it executable with "chmod u+x add_terminal" - Edit add_terminal to meet your particular needs - Execute add_terminal - If the terminal doesn't give you a login when you hit try rebooting the machine once. #!/bin/sh # ########################################################################## # add_terminal shell script. Must be run as root from bourne shell # Change these parameters as needed for your particular needs. # PARAMETER MEANING # --------- ------- # -s ttya terminal is on ttya. you may want ttyb # -d /dev/term/a actual port device. you may want /dev/term/b # -l 9600E speed/stty setting from /etc/ttydefs. choose an # entry that meets your needs # -T tvi925 your default terminal type for this port. This # is a terminfo terminal type # -i 'terminal disabled' message sent to the tty port if the port is # ever disabled # QUOTE-ALERT: The single quotes the port-disabled message are # FORWARD QUOTES ########################################################################## # Remove any existing port monitor on ttya Oct 9 07:57 1993 Setting up Terminals and Modems under Solaris 2.x Page 3 # (Change the "ttya" to "ttyb" if that's the port you're dealing with) # pmadm -r -p zsmon -s ttya # # Create the new port monitor # # QUOTE-ALERT: The single quotes in the port-disabled message # [terminal disabled] are FORWARD QUOTES # pmadm -a -p zsmon -s ttya -i root -fu -v 1 -m "`ttyadm -l 9600E \ -d /dev/term/a -T tvi925 -i 'terminal disabled' -s /usr/bin/login -S y`" Oct 9 07:57 1993 Setting up Terminals and Modems under Solaris 2.x Page 4 B. Setting Up a Bidirectional Modem on Solaris 2.x ================================================== B.1 BEFORE YOU BEGIN: SET UP THE MODEM Be sure that the modem is configured for the following: - Hardware DTR, when the Sun drops DTR the modem should hang up and reset the modem. - Hardware Carrier Detect, the modem only raises the CD line when there is an active carrier signal on the phone connection. When carrier drops, either when the other end of the connection terminated or if the phone connection is broken, the Sun will be notified and act appropriately. The CD signal is also used for coordinating dial-in and dial-out use on a single serial port and modem. - Respond with numeric result codes. (Usually this is ATV0) - Sends BASIC result codes ONLY. (Sometimes this is ATX0 or ATQ0) - Does not echo commands. (Usually this is ATE0) - Modem is locked at a single speed setting between Sun and modem. Speed may vary as needed between local and remote modem, but should remain constant between Sun and modem. B.2 ENABLING SOLARIS FOR A BIDIRECTIONAL MODEM ********************************************************************* *** IMPORTANT NOTES: *** *** 1. All single quotes in these directions are BACK QUOTES, *** *** not FORWARD QUOTES, except where specifically noted. *** *** *** *** 2. Be sure and execute these commands as root, from a *** *** BOURNE shell (sh), not ksh, tcsh, or csh. (If you use *** *** the a Bourne shell script, that's ok.) *** *** *** *** 3. If, after following these instructions, you get *** *** modem failures, or if step #7 gives an error message *** *** about "ttyadm: -V: invalid parameter", then make sure *** *** that root's default shell is /sbin/sh and not csh. *** ********************************************************************* 1. Login as root and type: prompt# eeprom ttya-ignore-cd=true prompt# eeprom ttya-rts-dtr-off=true >> go to step 2 2. Reboot the system by typing: prompt# init 6 >> Go to step 3 3. Make sure the modem is properly connected and configured. Make sure the cable is ok. >> Go to step 4 Oct 9 07:57 1993 Setting up Terminals and Modems under Solaris 2.x Page 5 4. Is zsmon configured and running? (The following instructions assume that the modem cable to serial port A.) Login as root and type: prompt# sacadm -l -t ttymon If you get a message like: "Invalid request, zsmon does not exist" then the zsmon port monitor is not configured. >> Go to step 5 If you get a result like: PMTAG PMTYPE FLGS RCNT STATUS COMMAND zsmon ttymon - 0 ENABLED /usr/lib/saf/ttymon >> Go to step 6 5. To configure the zsmon port monitor prompt# sacadm -a -p zsmon -t ttymon -c /usr/lib/saf/ttymon \ -v `ttyadm -V` -y "dial in/out on serial port a" (Note: zsmon = PMTAG "dial in/out on serial port a" = "any comment") >> Go to step 6 6. Is ttymon configured and running? Type: prompt# pmadm -l -s ttya (where ttya is the SVCTAG) a) If there is no service, you will get a result like: the Service does not exist. >> Go to step 7. b) If you got a result like: PMTAG PMTYPE SVCTAG FLGS ID zsmon ttymon ttya u root /dev/term/a I - / Then you need to remove the existing service by typing: prompt# pmadm -r -p zsmon -s ttya ^ ^ (PMTAG) (SVCTAG) >> Go to step 7. Oct 9 07:57 1993 Setting up Terminals and Modems under Solaris 2.x Page 6 7. Make a script which will remove any existing port monitor and create a new bidirectional port service. (Putting this in a script is a Very Good Idea because if you make a mistake, you can easily fix the problem and rerun the script. Don't try to do this without a script--there are too many typing errors waiting to be made in this command. Type in the following script as a file (for instance, /sbin/add_modem), make it executable, then run it: #!/bin/sh # ########################################################################## # add_modem shell script. Must be run as root from bourne shell # Change these parameters as needed for your particular needs. # PARAMETER MEANING # --------- ------- # -p zsmon PMTAG / name of this port monitor # -s ttya SVCTAG / modem is on ttya. you may want ttyb # -d /dev/term/a actual port device. you may want /dev/term/b # -l contty5H ttylabel, defined in /etc/ttydefs file (speed # setting. # contty5H=19200 baud. see ttydefs for other baud rates) # -b flag for bidirectional port use # -S n turn software carrier off (modem supplies hardware # carrier detect signal) # "dial in/out on serial port a" this is a comment you'll see on pmadm -l ########################################################################## # Remove any existing port monitor on ttya # (Change the "ttya" to "ttyb" if that's the port you're dealing with) # pmadm -r -p zsmon -s ttya # # Create the new port monitor # pmadm -a -p zsmon -s ttya -i root -v `ttyadm -V` -fu -m \ "`ttyadm -d /dev/term/a -s /usr/bin/login -l contty5H -b -S n -m \ ldterm,ttcompat`" -y "dial in/out on serial port a" # #end of add_modem script ##################################################################### Make the script executable by typing prompt# chmod 700 /sbin/add_modem then actually run it with: prompt# /sbin/add_modem >> Go to step 8. Oct 9 07:57 1993 Setting up Terminals and Modems under Solaris 2.x Page 7 8. TIP and /etc/remote Edit /etc/remote and find the entry beginning with "hardwire:". Change the portion which says: :dv=/dev/??????: to read :dv=/dev/cua/a: (or /dev/cua/b if you're using the B port) Change the portion which says: :br#somenumber: to read :br#9600: (or 19200, or whatever baud rate you plan to use) >> Go to step 9 9. Type: prompt# tip hardwire you should see "connected" type: ATE1V1 you should see "OK" >> Go to step 10 If you don't see "OK", you are not communicating with the modem. See if the modem is plugged in, verify the baud rate of the modem and the hardwire entry in /etc/remote, verify the modem cable. 10. Test the dialin Test your modem setup by dialing into the Solaris 2.x machine. If you get logged in, congratulate yourself! >> Go to step 11 11. Further setup for using the modem At this point, you have established basic outgoing computer-to-modem setup. If you plan to use uucp, then start setting up the /etc/uucp files. (This is left as an exercise for the reader.) If you are interested in more ways to use tip with your modem, see Section C of this document. >> Go To Section C Oct 9 07:57 1993 Setting up Terminals and Modems under Solaris 2.x Page 8 C. More on tip and /etc/remote ============================== 1. The cuaa entry (below) allows you to type "tip cuaa" and talk directly to a modem on port a at 2400 baud. 2. "tip phonenumber" will look for the tip0 entry and use that definition. (In the example below, it will use tip0, which points to UNIX-2400, which sets up a dialup 2400 baud hayes compatible modem. UNIX-2400 points to dialers, which references /dev/cua/a. 3. "tip mysystem" (below) looks up the "mysystem" entry, and dials the number 14155551212 via a 19200 baud connection on /dev/cua/a. The following sample file is set up for port a, 2400 baud hayes compatible modem port b, 9600 baud modem, may or may not be hayes compatible /etc/remote example: ==================== cuaa:dv=/dev/cua/a:br#2400 mysystem:pn=14155551212:tc=UNIX-19200: hardwire:\ :dv=/dev/cua/b:br#9600:el=^C^S^Q^U^D:ie=%$:oe=^D: tip1200:tc=UNIX-1200: tip0|tip2400:tc=UNIX-2400: tip9600:tc=UNIX-9600: tip19200:tc=UNIX-19200: cu1200:tc=UNIX-1200: cu0|cu2400:tc=UNIX-2400: UNIX-1200:\ :el=^D^U^C^S^Q^O@:du:at=hayes:ie=#$%:oe=^D:br#1200:tc=dialers: UNIX-2400:\ :el=^D^U^C^S^Q^O@:du:at=hayes:ie=#$%:oe=^D:br#2400:tc=dialers: UNIX-9600:\ :el=^D^U^C^S^Q^O@:du:at=hayes:ie=#$%:oe=^D:br#9600:tc=dialers: UNIX-19200:\ :el=^D^U^C^S^Q^O@:du:at=hayes:ie=#$%:oe=^D:br#19200:tc=dialers: dialers:\ :dv=/dev/cua/a: -------------------------------------------------------------------- The attributes are: dv device to use for the tty du make a call flag (dial up) pn phone numbers (@ =>'s search phones file; possibly taken from PHONES environment variable) at ACU type ie input EOF marks (default is NULL) oe output EOF string (default is NULL) cu call unit (default is dv) br baud rate (defaults to 300) tc to continue a capability Oct 9 07:57 1993 Setting up Terminals and Modems under Solaris 2.x Page 9 D. Setting other serial port modes ================================== Serial port modes (the terminal I/O options found in stty(1) ) are set in the file /etc/ttydefs. The name of the serial port mode (like contty5H) is used as the argument to the "-l" option in the "ttyadm" portion of the "pmadm -a" command. (This is the "speed" setting you use when you add a new port monitor.) See the stty man page for all the available terminal I/O options. The format of the entries in ttydefs is defined as: ttylabel:initial-flags:final-flags:autobaud:nextlabel ttylabel The string ttymon tries to match against the TTY port's ttylabel field in the port monitor administrative file. It often describes the speed at which the terminal is supposed to run, for example, 1200. initial-flags Contains the initial termio(7) settings to which the terminal is to be set. For example, the system administrator will be able to specify what the default erase and kill characters will be. initial-flags must be specified in the syntax recognized by the stty command. final-flags final-flags must be specified in the same format as initial-flags. ttymon sets these final settings after a connection request has been made and immediately prior to invoking a port's service. autobaud If the autobaud field contains the character `A', autobaud will be enabled. Otherwise, autobaud will be disabled. ttymon determines what line speed to set the TTY port to by analyzing the carriage returns entered. If autobaud has been disabled, the hunt sequence is used for baud rate determination. nextlabel If the user indicates that the current terminal setting is not appropriate by sending a BREAK, ttymon searchs for a ttydefs entry whose ttylabel field matches the nextlabel field. If a match is found, ttymon uses that field as its ttylabel field. A series of speeds is often linked together in this way into a closed set called a hunt sequence. For example, 4800 may be linked to 1200, which in turn is linked to 2400, which is finally linked to 4800. WARNING: These stty modes are processed from left to right within each set of "initial-flags" or "final-flags". So, the order of the modes in each entry is significant. In particular, when setting the mode to 8bits, do NOT enter it as ".... -parity ... sane", but as "... sane... -parity" because "sane" sets the port to 7 bits. Oct 9 07:57 1993 Setting up Terminals and Modems under Solaris 2.x Page 10 /etc/ttydefs examples: ====================== 1. contty5H, as supplied by Sun, is 19200 bps, 7 bits even parity contty5H:19200 opost onlcr:19200 hupcl sane::conttyH This is equivalent to: contty5H:19200 evenp opost onlcr:19200 sane evenp hupcl::conttyH or contty5H:19200 parenb -parodd opost onlcr:19200 sane parenb -parodd \ hupcl::conttyH 2. Change contty5H to "7 bits, no parity" (-parenb) contty5H:19200 -parenb opost onlcr:19200 sane -parenb hupcl::conttyH 3. Change contty5H to "8 bits, no parity" (-parenb cs8, or -parity) contty5H:19200 -parenb cs8 opost onlcr:19200 sane -parenb cs8 hupcl::conttyH or contty5H:19200 -parity opost onlcr:19200 sane -parity hupcl::conttyH 4. Change contty5H to "7bits, odd parity" (parenb parodd, or oddp) contty5H:19200 parenb parodd opost onlcr:19200 sane parenb parodd \ hupcl::conttyH or contty5H:19200 oddp opost onlcr:19200 sane oddp hupcl::conttyH Oct 9 07:57 1993 Setting up Terminals and Modems under Solaris 2.x Page 11 E. Random notes of various uses of serial ports =============================================== 1. Parity in UUCP: People often get confused about running UUCP in 7bits vs. 8 bits vs. even or odd parity. Here's what's really going on and how to deal with it. Both the caller and receiver must agree on parity (even, odd, none) during the login sequence. 7 or 8 bit word length really doesn't matter, if you "do the right thing" with parity. Once the uucico's are running, UUCP's own protocol handles all these issues for you, but you need to be concerned with it during login. There are 4 parity-related options you can set up during the login chat: P_ZERO - No parity P_EVEN - Even parity P_ODD - Odd parity P_ONE - "1" or Mark parity It is usually safest to set up parity on a per-system-to-call basis, in the /etc/uucp/Systems file, but it can also be set for all uses of the port, in /etc/uucp/Dialers file. Example 1. Setting up a port within UUCP to run with no parity, on a per-system-connect basis. In the file /etc/uucp/Systems, include P_ZERO in the chat script, as in: outhost Any ACU 9600 5551212 "" P_ZERO ogin: mylogin ssword: mypass This means that when you first start trying to contact the machine "outhost" uucp expects nothing ("") and sets P_ZERO (no parity). Example 2. Setting up a port within UUCP to run with no parity, for all uses of this port. (This example assumes you are using the "tbfast" entry in your /etc/uucp/Devices file) In /etc/uucp/Dialers: tbfast =W-, "" P_ZERO "" \dA\pA\pA\pTE1V1X1...... Oct 9 07:57 1993 Setting up Terminals and Modems under Solaris 2.x Page 12 2. Flow control with UUCP: With UUCP, either use hardware flow control or no flow control. Please remember that there is no incoming hardware flow control on Sun CPU serial ports, only outgoing hardware flow control. (There is no such limitation on non-cpu ports which claim to have hardware flow control.) The lack of incoming hardware flow control is not a problem with UUCP, usually, because the UUCP protocol is robust enough to request retransmission of packets, when the data isn't correct. Using a good modem that has built-in UUCP protocol support (also known as spoofing) will also help. If all else fails, slow down the connection between the Sun and the modem to what you can support in your environment. The string "STTY=crtscts" tells UUCP to use hardware flow control. It's usually better to set this up on a per-port basis, rather than on a per-system basis, but UUCP will let you configure it however you want. Example 1. Setting up a port with UUCP to use hardware flow control for all uses of this port. (Preferred way) (This example assumes you are using the "tbfast" entry in your /etc/uucp/Devices file) In /etc/uucp/Dialers, add "STTY=crtscts" after the CONNECT\sFAST: tbfast =W-, "" \dA\pTE1.... \EATDT\T\r\c CONNECT\sFAST STTY=crtscts Example 2. Setting up a port with UUCP to use hardware flow control, on a per-system-connect basis. (Do this if you feel you must, but there are very few good reasons for it.) In the file /etc/uucp/Systems, include "STTY=crtscts" in the chat script, as in: outhost Any ACU 9600 5551212 ogin: mylog ssword: mypass "" STTY=crtscts ..Celeste Stokely Unix System Administration Consultant, Stokely Consulting EMAIL: celeste@xs.com Voice Line: 415-967-6898 / FAX: 415-967-0160 USMAIL Address: Stokely Consulting 211 Thompson Square / Mountain View CA 94043 From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!xlink.net!ira.uka.de!iras5!guenther Wed Dec 8 11:05:05 EST 1993 Article: 10398 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!xlink.net!ira.uka.de!iras5!guenther From: guenther@iras5.ira.uka.de (Guenther Schreiner) Newsgroups: comp.unix.solaris,comp.sys.sun.hardware,comp.sys.sun.admin Subject: Re: SPARCserver 1000 panics when booting Solaris 2.3 Date: 7 Dec 1993 14:12:58 GMT Organization: University of Karlsruhe, FRG Lines: 18 Distribution: world Message-ID: <2e231aINN7e9@iraun1.ira.uka.de> References: NNTP-Posting-Host: iras5.ira.uka.de Xref: news.umbc.edu comp.unix.solaris:10398 comp.sys.sun.hardware:14396 comp.sys.sun.admin:25272 Status: RO In article , wisner@well.sf.ca.us (Bill Wisner) writes: |> 2.3 the system would consistently panic near the end of the run state |> 2 startup. I finally got the system to boot by disabling the startup |> of vold. I don't think vold was crashing the system; I suspect it was The same problem occurs on our SC2000. It can be solved by removing the CD from the drive just before rebooting the system :( Guenther -- Guenther Schreiner | University of Karlsruhe, Germany Fakultaet fuer Informatik (CSD) ---+--- Am Fasanengarten 5 | Phone: (+49) 721 608-3980 or 4321 76128 Karlsruhe | FAX: (+49) 721 699284 | INTERNET: guenther@ira.uka.de X.400: G=Guenther/S=Schreiner/OU=Informatik/PRMD=UNI-KARLSRUHE/ADMD=D400/C=DE/; -- From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!vixen.cso.uiuc.edu!bradley.bradley.edu!bradley.bradley.edu!not-for-mail Mon Dec 20 16:36:05 EST 1993 Article: 10767 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!vixen.cso.uiuc.edu!bradley.bradley.edu!bradley.bradley.edu!not-for-mail From: pwh@bradley.bradley.edu (Pete Hartman) Newsgroups: comp.unix.solaris Subject: Re: fix for slow sendmail queues (Solaris2.3) Date: 16 Dec 1993 02:59:47 -0600 Organization: Bradley University Lines: 17 Distribution: usa Message-ID: <2ep823$hpf@bradley.bradley.edu> References: NNTP-Posting-Host: bradley.bradley.edu Status: RO bob@kahala.soest.hawaii.edu (Bob Cunningham) writes: >We had problems with email messages lingering in the queue for hours >and hours. The fix for Sun's sendmail is the same as in earlier >versions of Solaris2: > ln -s /unix /kernel/unix >Due to lingering braindamage in the distribution versions of >Sun's /usr/lib/sendmail.mx for Solaris. Care to explain why that's the necessary solution? If you can fix it that simply it sounds like you should be able to reconfigure the sendmail.mx file a lot more straightforwardly than doing symbolic links and turning your filesystems into spaghetti bowls... -- Pete Hartman Bradley University pwh@bradley.bradley.edu "It's the cobblestones." From news.umbc.edu!eff!usenet.ins.cwru.edu!agate!CS.Berkeley.EDU!eric Mon Dec 20 16:36:40 EST 1993 Article: 10802 of comp.unix.solaris Path: news.umbc.edu!eff!usenet.ins.cwru.edu!agate!CS.Berkeley.EDU!eric From: eric@CS.Berkeley.EDU (Eric Allman) Newsgroups: comp.mail.sendmail,comp.unix.solaris Subject: Re: Compiling sendmail 8.6.4 under Solaris 2.3 (sparc) Date: 16 Dec 1993 19:22:42 GMT Organization: UC Berkeley Mammoth Project Lines: 21 Sender: eric@mastodon.CS.Berkeley.EDU (Eric Allman) Distribution: inet Message-ID: <2eqci2$881@agate.berkeley.edu> References: <2eqash$7sm@agate.berkeley.edu> Reply-To: eric@CS.Berkeley.EDU NNTP-Posting-Host: mastodon.cs.berkeley.edu Xref: news.umbc.edu comp.mail.sendmail:10941 comp.unix.solaris:10802 Status: RO In article <2eqash$7sm@agate.berkeley.edu>, jma@tron.ssl.berkeley.edu (Jim Alexander) writes: |> I am having trouble getting sendmail 8.6.4 to compile under Solaris 2.3 - |> has anyone done this? The compile goes fine until the final link, |> where I get a couple of undefined symbols: |> |> gcc -o sendmail alias.o arpadate.o clock.o collect.o conf.o convtime.o daemon.o deliver.o domain.o envelope.o err.o headers.o macro.o main.o map.o mci.o parseaddr.o queue.o readcf.o recipient.o savemail.o srvrsmtp.o stab.o stats.o sysexits.o trace.o u|> db.o usersmtp.o util.o version.o -L/usr/sww/lib -L/usr/ucblib -L/usr/lib -L/usr/ccs/lib -lresolv -lsocket -lnsl -lelf |> Undefined first referenced |> symbol in file |> __switch_gethostbyaddr daemon.o |> __switch_gethostbyname daemon.o |> |> |> Has anyone seen these? I can't find these symbols in any of the |> libraries. I would think they'd be in libnsl.a, but they're not |> there. Can anyone assist? This is fixed in 8.6.5. Sorry for the delays on it -- I'm still trying to track down a couple of possibly very serious problems; however, like all good (i.e., bad) problems, I can't reproduce them. eric From news.umbc.edu!europa.eng.gtefsd.com!uunet!well!wisner Mon Dec 20 16:36:59 EST 1993 Article: 10806 of comp.unix.solaris Newsgroups: comp.mail.sendmail,comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!uunet!well!wisner From: wisner@well.sf.ca.us (Bill Wisner) Subject: Re: Compiling sendmail 8.6.4 under Solaris 2.3 (sparc) Message-ID: Sender: news@well.sf.ca.us Nntp-Posting-Host: well.sf.ca.us Organization: The Whole Earth 'Lectronic Link, Sausalito, CA References: <2eqash$7sm@agate.berkeley.edu> Distribution: inet Date: Thu, 16 Dec 1993 19:45:23 GMT Lines: 9 Xref: news.umbc.edu comp.mail.sendmail:10943 comp.unix.solaris:10806 Status: RO This will be fixed in Sendmail 8.6.5, which should be available by year end. If you want a temporary fix in the meantime, edit conf.h and find the bit where gethostbyname and gethostbyaddr are redefined for SOLARIS and change the definitions to #define gethostbyname res_gethostbyname #define gethostbyaddr res_gethostbyaddr -- Bill Wisner Gryphon Gang San Francisco CA 94116 From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!pipex!uknet!EU.net!ieunet!maths.tcd.ie!not-for-mail Wed Dec 22 05:09:44 EST 1993 Article: 10732 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!pipex!uknet!EU.net!ieunet!maths.tcd.ie!not-for-mail From: jm@maths.tcd.ie (Justin Mason) Newsgroups: comp.unix.solaris Subject: Re: Backup software for Solaris 2.3? Date: 15 Dec 1993 17:58:59 -0000 Organization: Universal Media Netweb (Ireland) Inc. Lines: 23 Distribution: inet Message-ID: <2enj93$fj1@salmon.maths.tcd.ie> References: NNTP-Posting-Host: salmon.maths.tcd.ie Status: RO wisner@well.sf.ca.us (Bill Wisner) writes: >I'm unsatisfied with Sun's Online: Backup (aka Backup Co-Pilot) software, >and am trying to find information on other backup packages that exist for >Solaris 2. There don't seem to be many. Anybody know of any such packages >or possibly have experience with them? Well, if you'll accept free software, you're sorted. Over here, we use the OSU-CIS backup system, from Ohio State U.'s CIS dept. It allows you to label 20 tapes, put 'em in a cupboard, and just stick tapes in the drive when the mail from cron says so, which is what I want. ;) Everything else is automatic and pretty user-friendly, all things considered. If you can compile perl, and there's a dump, ufsdump, efsdump or whatever, you can run it on your machine. Get it from ftp.cis.ohio-state.edu:/pub/backup/backup-2.6pre.tar.Z. -- Justin Mason (Iona Technologies' unix caretaker, fixer-upper and disk-filler) [email: ] -><- [w3 http://www.iona.ie/hyplan/jmason.html] "I was an infinitely hot and dense dot." -- Mark Leyner From news.umbc.edu!europa.eng.gtefsd.com!uunet!Germany.EU.net!infoac.rmi.de!didymus!kkaempf Wed Dec 22 05:16:23 EST 1993 Article: 10783 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!uunet!Germany.EU.net!infoac.rmi.de!didymus!kkaempf From: kkaempf@didymus.rmi.de (Klaus Kaempf) Subject: Re: Disabling L1-A and root passwds Message-ID: Organization: Klaus Kaempf Softwareentwicklung References: <9312152033.AA01762@vger.demon.co.uk> Date: Thu, 16 Dec 1993 13:46:57 GMT Lines: 46 Status: RO We had this discussion last year regarding Solaris 1.x. I still have the stuff here, if anyone is interested ... Carsten Dieste (cutter@marie.physik.tu-berlin.de) asked me about this a few weeks ago and mailed me his version of the program to enable/disable L1-A. The following program should work on Solaris 2.x (I haven't tested it, use at your own risk :-) /* * The reset sequence disable program. */ #include #include #include #include main() { struct kiockey key ; int fd ; int z,i ; fd = open( "/dev/kbd", 1 ) ; if( fd < 0 ) { perror("Open") ; exit( 1 ) ; } key.kio_tablemask = KIOCABORT1 ; ioctl( fd, KIOCGETKEY, &key ) ; /* read abort key entry */ if (key.kio_station == 0) printf("L1-A reset sequence already disabled\n"); else if (key.kio_station == 1) printf("L1-A reset sequence enabled\n"); key.kio_station = 0 ; ioctl( fd, KIOCSETKEY, &key ) ; /* map it to a 'hole' in map */ ioctl( fd, KIOCGETKEY, &key ) ; /* verify the change */ if (key.kio_station == 0) printf("L1-A reset sequence disabled\n"); else if (key.kio_station == 1) printf("L1-A reset sequence still enabled\n"); } -- Klaus Kaempf E-Mail: kkaempf@didymus.rmi.de Jakobstr. 181 Fax: 0241-403407 D-52064 Aachen Voice: 0241-403446 From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!vixen.cso.uiuc.edu!uwm.edu!psuvax1!cpumagic!mloewen Wed Dec 22 05:27:44 EST 1993 Article: 10883 of comp.unix.solaris Newsgroups: comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!vixen.cso.uiuc.edu!uwm.edu!psuvax1!cpumagic!mloewen From: mloewen@cpumagic.scol.pa.us (Michael C. Loewen) Subject: Re: How do you find the ethernet number Message-ID: <1993Dec18.004519.4775@cpumagic.scol.pa.us> Organization: The Centre Programmers Unit BBS References: <2es4iq$95@xenon.bt-sys.bt.co.uk> Date: Sat, 18 Dec 1993 00:45:19 GMT Lines: 16 Status: RO In article <2es4iq$95@xenon.bt-sys.bt.co.uk> mhinds@bt-sys.bt.co.uk (Mike Hinds) writes: >How do you find the ethernet number of a running machine ? > >There must be some simple way of doing this. > >it appears during the reboot. but I have to do a whole load of remote >machines. > Try the 'getethers' utility, which runs through all the addresses on a subnet, pings each one, and displays the resulting ethernet addresses. The last version I recall seeing was 1.5. -- Mike Loewen The Centre Programmers Unit BBS mloewen@cpumagic.scol.pa.us (814) 353-0566 V.32bis/HST From news.umbc.edu!europa.eng.gtefsd.com!library.ucla.edu!agate!doc.ic.ac.uk!dundee.ac.uk!dundee.ac.uk!not-for-mail Wed Dec 22 10:30:45 EST 1993 Article: 11053 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!library.ucla.edu!agate!doc.ic.ac.uk!dundee.ac.uk!dundee.ac.uk!not-for-mail From: pjclinch@dux.dundee.ac.uk (Pete Clinch) Newsgroups: comp.unix.solaris Subject: Re: Sun C 2.0.1 MakeTool Problems Date: 17 Dec 1993 13:51:13 -0000 Organization: Information Technology Services , Dundee University, Scotland, UK Lines: 22 Distribution: inet Message-ID: <2esdgh$q9e@dux.dundee.ac.uk> References: NNTP-Posting-Host: dux.dundee.ac.uk X-Newsreader: TIN [version 1.2 PL1] Status: RO Robert Morse (rmorse@rmsun.uucp) wrote: : Hi, : I upgraded one of our systems to Solaris 2.3, and now it seems that : the Maketool that came with SC2.0.1 core dumps when it is started. : The machine that is a Sparc 1 running 2.3 with some patches. Maketool : runs fine under Solaris 2.2. : Has anyone seen this problem, : If no, I will look into some problem with my machine. : If yes, is there a patch, solution to the core dumpping. It's a recognised problem. There is a patch now: Sun 'phoned me yesterday after I logged this problem a cople of weeks ago to say that the patch is ready and will be in the post today. Don't know the number of the patch, but Sun are aware and will post it to you. Pete. -- Peter Clinch University of Dundee voice 44 382 60111 x 2050 Department of Medical Physics fax 44 382 640177 Ninewells Hospital email pjclinch@dux.dundee.ac.uk Dundee, DD1 9SY, Scotland, UK From casper@fwi.uva.nl Sat Dec 25 15:22:24 EST 1993 Article: 11157 of comp.unix.solaris Path: news.umbc.edu!haven.umd.edu!darwin.sura.net!math.ohio-state.edu!cs.utexas.edu!uunet!zib-berlin.de!netmbx.de!Germany.EU.net!EU.net!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: UPGRADE 2.2 to 2.3 Date: 24 Dec 1993 13:54:15 GMT Organization: FWI, University of Amsterdam Lines: 30 Distribution: world Message-ID: <2fesa7$kln@mail.fwi.uva.nl> References: <1993Dec21.145043@ctmces1.hou130.chevron.com> NNTP-Posting-Host: adam.fwi.uva.nl Status: RO miwh@ctmces1.hou130.chevron.com (Mike Whitley) writes: >Greetings all... >Forgive me if someone has already asked this... >Has anybody successfully UPGRADED from 2.2 to 2.3 (ie. I don't >mean BULK installing 2.3 over 2.2)? When I choose "upgrade" >in suninstall, my root partition gets mounted ok, but my usr >partition mount never makes it. I get "a file system in your >vfstab cannot be mounted..." I would assume (dangerous) that my >original vfstab in my machine's /etc directory would get looked at >in order to find the usr part. for upgrade purposes (partition 6 on sd0). >After suninstall tells me to exit, I can see my root partition mounted on >the cdrom's /a but no usr partition. The physical device partition being >used by suninstall is correct (it shows in the error message) but there is >no explanation why it can't be mounted. The system works fine under 2.2...no >partition probs etc. Has anyone called suninstall with "manual" mounts? >I called Sun on this one, and they are throwing it in the "pile" of bug reports >on the 2.3 install procs. When upgrading, make sure that all non-ufs filesystems are no longer in your vfstab file (best make sure that only filesystems touched by the installation (/, /usr, /var, /usr/openwin, /opt) are present in /etc/vfstab. Perhaps you have cdrom, nfs mounted filesystems, etc, etc in /etc/vfstab. You may even need to remove /proc and /dev/fd. Casper From news.umbc.edu!eff!news.kei.com!bloom-beacon.mit.edu!grapevine.lcs.mit.edu!kaleb Fri Jan 7 03:57:39 EST 1994 Article: 11474 of comp.unix.solaris Path: news.umbc.edu!eff!news.kei.com!bloom-beacon.mit.edu!grapevine.lcs.mit.edu!kaleb From: kaleb@x.org (Kaleb Keithley) Newsgroups: comp.unix.solaris Subject: Re: building motif under sol 2.3 Date: 4 Jan 94 16:13:00 GMT Organization: X Consortium, Cambridge, Massachusetts Lines: 42 Distribution: world Message-ID: References: <2gbstv$bdi@transfer.stratus.com> NNTP-Posting-Host: kanga.x.org Status: RO tonyr@sw.stratus.com (Tony Rudie) writes: >Has anyone out there built Motif under solaris? Most assuredly. >, but in trying to build motif for sol2 I can't get >the libraries right. I have tried >all four combinations of defining or not defining NO_REGEX and NO_REGCOMP Combinations is the wrong approach. If you look at the source, since you have a site license, you'll see that -DNO_REGCOMP is all that you need. Of course you still need -DSTRINGS_ALIGNED. >but in all cases either the libraries won't build, or when they do, >they produce unreferenced symbol errors when used to build the rest >of the stuff. the unreferenced symbols are either regex and regcmp >or re_exec and re_comp depending on the compiler flags. Did you RTFM? 'man -s3g regcmp' says: regcmp(3G) C Library Functions regcmp(3G) NAME regcmp, regex - compile and execute regular expression SYNOPSIS cc [ flag ... ] file ... -lgen [ library ... ] #include char *regcmp(const char *string1, /* char *string2 */ ... , (char *)0); char *regex(const char *re, const char *subject, /* char *ret0 */ ... ); -- Kaleb Keithley From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!pipex!sunic!EU.net!sun4nl!fwi.uva.nl!casper Fri Jan 7 08:53:37 EST 1994 Article: 11517 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!pipex!sunic!EU.net!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: LP troubles in Solaris 2.2 despite patch 101025-08 Date: 5 Jan 1994 10:53:55 GMT Organization: FWI, University of Amsterdam Lines: 42 Distribution: world Message-ID: <2ge683$hj4@mail.fwi.uva.nl> References: <1994Jan5.121828.19963@uts.uni-c.dk> <2gcmu6$6gf$1@sdl.Warren.MENTORG.COM> NNTP-Posting-Host: adam.fwi.uva.nl Keywords: Solaris 2.2, lp system, patch 101025-08 Status: RO ohnielse@fysik (0000-Admin(0000)) writes: >We keep having troubles with the lp system in Solaris 2.2, despite >having installed patch 101025-08 (SunOS 5.2: Jumbo patch fixes for lp system). >The symptoms we find resemble very much these topics that 101025-08 >claims to solve: >1145448 too many print jobs can hang around and make /var very full >1136170 Jobs never disappear from lp queue >Searching in SunSolve Online did not reveal any further hints or help. This can also be a symptom of incorrectly setting up your print clients. Do you have any Solaris 2.x print clients? If so, you must: - install the lp jumbo patch there too. - make sure you do the printer set-up on the clients as well. sacadm -l should show something like: tcp listen - 999 ENABLED /usr/lib/saf/listen tcp # pmadm -l should show something similar to: tcp listen lp - root - - p - /var/spool/lp/fifos/listenS5 # tcp listen lpd - root \x00020203000000000000000000000000 - p - /var/spool/lp/fifos/listenBSD # tcp listen 0 - root \x00020ACE000000000000000000000000 - c - /usr/lib/saf/nlps_server # And the print server must be in the /etc/lp/Systems file of each client. The System V print spooler wishes to report success or failure to the print client. It can only do so if the client configured as a print server. (I.e., all the listening services are installed) Only after the client is notified about success or failure, the print jobs are removed. If teh client isn't listening, the jobs remain in the queue with lpstat saying ``finished''. Sun has made available an add_printer script that does all this quite well (Tom Limoncelli just posted this in comp.sys.sun.admin, ``Problems with printing on SunOS 5.2'' article <2gcmu6$6gf$1@sdl.Warren.MENTORG.COM>) Casper From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!agate!ames!koriel!news2me.EBay.Sun.COM!exodus.Eng.Sun.COM!appserv.Eng.Sun.COM!teal!mre Mon Jan 10 05:13:12 EST 1994 Article: 11694 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!agate!ames!koriel!news2me.EBay.Sun.COM!exodus.Eng.Sun.COM!appserv.Eng.Sun.COM!teal!mre From: mre@teal.Eng.Sun.COM (Mike Eisler) Newsgroups: comp.unix.solaris Subject: Re: nfs3 Followup-To: comp.protocols.nfs Date: 8 Jan 1994 00:52:25 GMT Organization: Sun Microsystems, Mountain View, CA USA Lines: 23 Message-ID: References: <1994Jan6.080943.5719@aristo.tau.ac.il> NNTP-Posting-Host: teal Keywords: nfs3 Status: RO In article <1994Jan6.080943.5719@aristo.tau.ac.il> david@bones.co.il writes: > >Someone told me about nfs3... >seems to be great ... >more info are welcome.... > > David. Tried email directly, but it bounced. You can get the NFS V3 spec via anonymous ftp from: ftp.uu.net:/networking/ip/nfs/NFS3.spec.ps.Z bcm.tmc.edu:/nfs/nfsv3.ps.Z gatekeeper.dec.com:/pub/standards/nfs/nfsv3.ps.Z Note that SunSoft (and to the best of my knowledge, no one else) has released products that use the NFS V3 protocol, but they are in various stages of development. Questions about the protocol ought to be directed to comp.protocols.nfs -Mike Eisler mre@Eng.Sun.Com From news.umbc.edu!europa.eng.gtefsd.com!uunet!news.mentorg.com!sdl!not-for-mail Mon Jan 10 05:16:56 EST 1994 Article: 11720 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!uunet!news.mentorg.com!sdl!not-for-mail From: tal@Warren.MENTORG.COM (Tom Limoncelli) Newsgroups: comp.unix.solaris Subject: Re: cachefs entries in vfstab / Solaris 2.3 Date: 8 Jan 1994 20:43:47 -0500 Organization: Mentor Graphics -- IC Group, Warren, NJ, USA Lines: 19 Sender: tal@Warren.MENTORG.COM Distribution: inet Message-ID: <2gnngj$7tu$1@sdl.Warren.MENTORG.COM> References: NNTP-Posting-Host: sdl.warren.mentorg.com Status: RO In nejdl@informatik.rwth-aachen.de (Wolfgang Nejdl) writes: >Could anybody send me working entries for vfstab (both for /usr and >/usr/kvm on a dataless client and for a CDROM)? I have not managed to >get them to work after some tries. cachefs (the most promising feature of Solaris 2.3) doesn't work as shipped. To make it work, get patch 101329-04. FTP to sunsite.unc.edu (152.2.22.81) and grab /pub/sun-info/sun-fixes/101329-04.tar.Z Tom -- Tom Limoncelli -- tal@warren.mentorg.com (work) -- tal@plts.org (play) "Psst! Hey, Anthony! Y'know what I | Disclaimer: I do not like about existing?" "Uh... uh... what?" | speak for Mentor Graphics. "Possessing a physical extension." -TSA | From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!pipex!sunic!EU.net!sun4nl!fwi.uva.nl!casper Mon Jan 10 05:18:02 EST 1994 Article: 11724 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!pipex!sunic!EU.net!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: cachefs entries in vfstab / Solaris 2.3 Date: 9 Jan 1994 11:45:47 GMT Organization: FWI, University of Amsterdam Lines: 21 Distribution: inet Message-ID: <2goqpb$f35@mail.fwi.uva.nl> References: <2gnngj$7tu$1@sdl.Warren.MENTORG.COM> NNTP-Posting-Host: adam.fwi.uva.nl Status: RO tal@Warren.MENTORG.COM (Tom Limoncelli) writes: >cachefs (the most promising feature of Solaris 2.3) doesn't work as >shipped. >To make it work, get patch 101329-04. FTP to sunsite.unc.edu >(152.2.22.81) and grab /pub/sun-info/sun-fixes/101329-04.tar.Z That patch does not contain fixes to cachefs. There are fixes to autofs, to make it work with cachefs and there are some library and NIS+ fixes. No parts of cachefs are replaced. (Though new features not working is pretty common in SunOS) This is what works for us: server:/usr/openwin - /usr/openwin cachefs - yes backfstype=nfs,ro,cachedir=/cache,noquota,intr You must initialise the cache with cfsadmin before using it for the first time. Casper From news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!agate!doc.ic.ac.uk!uknet!EU.net!sun4nl!fwi.uva.nl!casper Mon Jan 10 05:18:47 EST 1994 Article: 11725 of comp.unix.solaris Path: news.umbc.edu!europa.eng.gtefsd.com!howland.reston.ans.net!agate!doc.ic.ac.uk!uknet!EU.net!sun4nl!fwi.uva.nl!casper From: casper@fwi.uva.nl (Casper H.S. Dik) Newsgroups: comp.unix.solaris Subject: Re: SCSI speed on Solaris 2.3? Date: 9 Jan 1994 11:53:45 GMT Organization: FWI, University of Amsterdam Lines: 33 Distribution: world Message-ID: <2gor89$f3n@mail.fwi.uva.nl> References: <1994Jan8.172336.22683@pinet.aip.org> NNTP-Posting-Host: adam.fwi.uva.nl Status: RO wonder@pinet.aip.org (James Wonder) writes: >Hello, > Does anybody know what the speed settings on Solaris 2.3 >are (Synch. vs Asynch.)? I know on 4.1.3 thay are 58 >for Asynch. and 78 (or 178) for Synch., and you change >them using "adb". I looked at my 2.3 system with adb and >came up with a number that did not match the above. Do >you change the speed the same way you would on 4.1.3 or >is there a better/easier way of doing it. What are >the correct values for Synch. and Asynch. on 2.3? >Thanks for any help regarding this matter. in you'll find the options. #define SCSI_OPTIONS_DR 0x8 /* Global disconnect/reconnect */ #define SCSI_OPTIONS_LINK 0x10 /* Global linked commands */ #define SCSI_OPTIONS_SYNC 0x20 /* Global synchronous xfer capability */ #define SCSI_OPTIONS_PARITY 0x40 /* Global parity support */ #define SCSI_OPTIONS_TAG 0x80 /* " tagged command support */ #define SCSI_OPTIONS_FAST 0x100 /* " FAST scsi support */ The default value is most likely to be 0x1f8 (all options). You can change the options in /etc/systems: set scsi_options = 0x178 * no tagged commands set scsi_options = 0x1d8 * no sync scsi Just | the options together. Casper