poor NFS performance in CURRENT
Hi :) I upgraded two boxes to FreeBSD-5.2-BETA a week ago and I noticed that NFS performance is very slow compared to 4.x-RELEASE. Before, NFS transfers were between 10 and 12 MB/s and now I don't go past 7 MB/s. My exports/mount settings did not change and the hardware is obviously the same. My kernel is compiled without the debugging options. Any idea where I should start looking for resolving this ? Thanks. Antoine ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
ntpd and kvm_getprocs
Hi :) I'm having a problem with ntpd under 5.2-BETA When I start ntpd, I get the following error message: $ /etc/rc.d/ntpd start ps: kvm_getprocs: No such process Starting ntpd. ntpd get started anyway but does not seem to work, my /var/db/ntp.drift file contains "0.000" and this never changes ... ! Here is my ntp.conf file: #---# server ntp.univ-lyon1.fr prefer server clepsydra.dec.com server ntp0.nl.net driftfile /var/db/ntp.drift restrict 127.0.0.1 restrict 192.168.0.0 mask 255.255.255.0 notrust nomodify notrap restrict default ignore #---# Thanks. -- Antoine Jacoutot [EMAIL PROTECTED] http://www.lphp.org PGP/GnuPG key: http://www.lphp.org/ressources/ajacoutot.asc ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Re: rpc.lockd core dumped
> I can only say that I had a core dump under current on sparc, but the core > file was unusable. Can you compile rcp.lockd with -g in CFLAGS and LDFLAGS > and try to find out with gdb where it aborts? > > harti Hi :) I investigate a little more this nfs locking issue. I recompile my system with CURRENT from "Wed Nov 19 12:53:04". rpc.lockd core dump still appear but I noticed something: each time, just before it crashed, I can hear a very low noise coming from my harddrive. It is hard to describe, but it looks like if my hardrive was powering on or coming back from hibernate mode (or something like this). Is there a way to check this ? Like maybe a command to force hardrive not to stop or something ? Thanks. Antoine ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: upgraded to CURRENT = system is dead
On Tuesday 18 November 2003 00:17, Garance A Drosihn wrote: > Users need to specify a single target of 'installkernel', > with no blanks between the two words. Antoine, was that > just a typo, or did you really do 'make install kernel'? I was a typo :) -- Antoine Jacoutot [EMAIL PROTECTED] http://www.lphp.org PGP/GnuPG key: http://www.lphp.org/ressources/ajacoutot.asc ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: upgraded to CURRENT = system is dead
Selon Michael Collette <[EMAIL PROTECTED]>: > I just ran into this very same thing today on a brand new installation. Upon > > reboot I can't even get to a shell. Was rather hoping someone had some more > > information about this. > > Running the installation on an IBM xSeries 305 with a P4. Going to try the > installation fresh with a cvsup back to Nov 1st to see how it goes. Thank > the Unix gods this wasn't a production machine. Someone told me a solution for this: " If you are looking for a fast and temporary solution, just to finish the installworld process, you can export the following value to the LD_LIBRARY_PATH variable: "$LD_LIBRARY_PATH:/lib:/usr/lib" " ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: upgraded to CURRENT = system is dead
Selon Aron Håkanson <[EMAIL PROTECTED]>: > If you are looking for a fast and temporary solution, just to finish the > installworld process, you can export the following value to the > LD_LIBRARY_PATH variable: "$LD_LIBRARY_PATH:/lib:/usr/lib" Hey :) Thanks a lot... I'll try that. Antoine ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: upgraded to CURRENT = system is dead
Selon Dylan Wylie <[EMAIL PROTECTED]>: > The idea is that you need to build a new kernel with the new sources before > you install > world. > You built your kernel before building the sources, so your kernel is based on > old source. > > Quoting from another message: > > make buildworld > > make buildkernel > > make installkernel > > reboot > > make installworld --> yes... but I get this during the installworld: /libexec/ld-elf.so.1: Shared object "libedit.so.4" not found Stop in /usr/src :( I just tried it again from scratch but got the same error... ! If anyone has an idea... Antoine ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: upgraded to CURRENT = system is dead
Selon long <[EMAIL PROTECTED]>: > reboot > make world > single user mode, make installworld > reboot > > and it's up and running... no more sig12 or anything like that... Yes, but no ! I don't have any sig12 error... I have: /libexec/ld-elf.so.1: Shared object "libedit.so.4" not found And I can't even mount anything :( I followed everything that was in UPDATING... So my box is really dead... right ? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
upgraded to CURRENT = system is dead
Hi :) I just upgraded with cvsup to CURRENT. I read UPDATING to make sure I would have no problem, but maybe I misunerstood something. Here is what I did: $ cvsup blablabla... $ make buildkernel KERNCONF=MYKERNEL && make install kernel KERNCONF=MYKERNEL $ reboot (in single user mode) $ make buildworld $ make installworld .. --> error during the installworld: /libexec/ld-elf.so.1: Shared object "libedit.so.4" not found Now, any command I try to launch gives me this exact same error and my system is then totally useless... :( Anything I could do ? Please please, tell me my box isn't dead... By the way, I compiled CURRENT with dynamic root. Thanks in advance. -- Antoine Jacoutot [EMAIL PROTECTED] http://www.lphp.org PGP/GnuPG key: http://www.lphp.org/ressources/ajacoutot.asc ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: rpc.lockd core dumped
> Check to make sure that rpc.statd is running. There was an old bug that > rpc.lockd would dump core if it couldn't find a statd. Ho, it is running :) Actually, all my homedir are mounted with NFS, so rpc.lockd get used a lot. That is why it is an important concern to me. I couldn't find any repeatable behaviour that would make rpc.lockd crashes. Sometimes, it will crash when I launch an application, sometime, it will just crash after 30 minutes of inactivities on my client... The server is running CURRENT and the clients are running 5.1-p10. I know it is not a very stable/secure configuration, but it is done on purpuse of testing the new functionnalities under a semi-production environment. Antoine ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: rpc.lockd core dumped
I will rpc.lockd with -ggdb as you said and see if it is repeatable. Unfortunately, I'm not home right now, so I'll do this in 3 o 4 days. Regards. Antoine ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: rpc.lockd core dumped
On Saturday 08 November 2003 12:26, Harti Brandt wrote: > I can only say that I had a core dump under current on sparc, but the core > file was unusable. Can you compile rcp.lockd with -g in CFLAGS and LDFLAGS > and try to find out with gdb where it aborts? Allright, as soon as I get home in 4/5 days, I'll do that. Thanks. Antoine ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
rpc.lockd core dumped
Hi :) Are they any know issues with rpc.lockd under -CURRENT. I had a look at the gnats database and did not find anything related. I'm asking this because I have a lot of: kernel: pid 70065 (rpc.lockd), uid 0: exited on signal 11 (core dumped) Any idea ? I would be pleased to send more information but I didn't see where to find more debuging options for rpc.lockd. 5.1-CURRENT #0: Tue Nov 4 01:44:35 CET 2003 [EMAIL PROTECTED]:/usr/ obj/usr/src/sys/KERNSRV01 i386 Thanks in advance. -- Antoine Jacoutot [EMAIL PROTECTED] http://www.lphp.org PGP/GnuPG key: http://www.lphp.org/ressources/ajacoutot.asc ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: NFS client mount options in CURRENT/5.1-
Scott W wrote: mount -tnfs -orw,rsize=8196,wsize=8196,bg,hard,intr,async sol:/export /mnt nfs: -o rsize=: option not supported Try "-r 8196 -w 8196" and have a look at man mount_nfs. Antoine ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: if_tun failed to register
On Wednesday 05 November 2003 23:25, Brian Lynn wrote: > will be run at startup if you have any ifconfig_tun* lines in rc.conf. > You can see if ifconfig is the culprit by booting single-user and doing > ifconfig tun0 (it is only the first attempt that gives the error message > in question). If so, the following (Untested!) patch would presumably But I don't have any "ifconfig_tun*" in rc.conf; although I do have the ppp_enable set on. Do you think I should try your patch anyway ? Antoine ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: if_tun failed to register
On Wednesday 05 November 2003 23:07, Michael Nottebrock wrote: > FWIW, tun _is_ in the kernel if you compile it in, but gets loaded a second > time. The module which gets loaded isn't actually used. Hum, strange... :) Do you know if the lastest -CURRENT fixes that ? Antoine ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: if_tun failed to register
On Wednesday 05 November 2003 22:47, Brooks Davis wrote: > > Thanks, I'm happy I'm not the only one seeing this :) > > It looks like ppp does not get compiled in the kernel either... :( > > ppp(4) is in GENERIC. You just have to create the devices like: No ! Although tun and ppp are in GENERIC, they get loaded as modules anyway under 5-CURRENT... at least under my box. This is so strange. Antoine ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: if_tun failed to register
On Wednesday 05 November 2003 18:37, you wrote: > It's been there for quite a while, I see that behaviour in 5.1-RELEASE, > too. Thanks, I'm happy I'm not the only one seeing this :) It looks like ppp does not get compiled in the kernel either... :( Antoine ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: if_tun failed to register
Matteo Riondato wrote: Well, it did not change anything :( What is really strange is that "tun" is compiled in the kernel, but the module is started anyway ??? I had the same problem last year and solved it by removing device tun from the kernel configuration file. Yes, I though about it. But still, it is a strange bug and I cannot believe I (well, and you :) ) am the only one seeing this. Thanks for the feedback Antoine ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: if_tun failed to register
Max Laier wrote: > AJ> > AJ> module_register: module if_tun already exists! > AJ> Module if_tun failed to register: 17 > AJ> can't re-use a leaf (if_tun_debug)! > AJ> > AJ> Is there a kernel option I missed ? I already have "device tun" in AJ> my kernel config file. This might be due to missing/incomplete "mergemaster" after installworld? Try to remove your rc.d related stuff from etc before running "mergemaster -i". The scripts should not try to "kldload if_tun" if you allready have tun interfaces (built in). Well, it did not change anything :( What is really strange is that "tun" is compiled in the kernel, but the module is started anyway ??? $ kldstat Id Refs AddressSize Name 16 0xc040 35aef8 kernel 21 0xc075b000 4cd10acpi.ko 31 0xc463d000 4000 logo_saver.ko 41 0xc46e2000 4000 if_tun.ko Here is a part of my kernel configuration: device miibus device fxp device random device loop device ether device ppp device tun device pty device bpf options ETGRAPH options NETGRAPH_ETHER options NETGRAPH_PPPOE options NETGRAPH_SOCKET options IPFIREWALL options IPFIREWALL_VERBOSE options IPFIREWALL_VERBOSE_LIMIT=100 options IPDIVERT options CPU_ENABLE_SSE options IPSTEALTH options RANDOM_IP_ID options SC_DISABLE_REBOOT I thank you very much for your answer. Regards. Antoine ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
if_tun failed to register
Hi :) I just migrated a 4.8 server to -CURRENT and I have one question. At boot time I get the following message: module_register: module if_tun already exists! Module if_tun failed to register: 17 can't re-use a leaf (if_tun_debug)! Then I can see a "if_tun.ko" in my loaded modules. What is this module ? I guess this has to do with the fact that I connect to the internet using ppp, but this module did not appear under 4.8. Is there a kernel option I missed ? I already have "device tun" in my kernel config file. Thanks in advance. -- Antoine Jacoutot [EMAIL PROTECTED] http://www.lphp.org PGP/GnuPG key: http://www.lphp.org/ressources/ajacoutot.asc ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: OpenLDAP/nss_ldap/pam_ldap
On Wednesday 29 October 2003 00:42, you wrote: > I just checked the FreeBSD site and do not see any release 5.2 It is not release yet :) > Once i can hurl this obsticle, i think FreeBSD might be a viable > solution for me. Well, let be it then... I'm running -CURRENT with dynamic root and it works great. Antoine ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: OpenLDAP/nss_ldap/pam_ldap
On Tuesday 28 October 2003 23:52, Steve Lee wrote: > I have gotten FreeBSD 5.1 to authenticate user remotely > ssh ing to the box. They can log in, but when they log > in, the system shows their userid instead of the username > when you do a ps. this tells me something might be > wrong with nss_ldap. but not sure what i am doing wrong. > > I can do "id " which it returns the user information. You need to build FreeBSD with dynamic libraries... It only works under -CURRENT with the "WITH_DYNAMICROOT=true" option in your make.conf. I think it is supposed to be the default for 5.2-RELEASE. Antoine ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Upgrading to FreeBSD 5.1
Kris Kennaway wrote: On Mon, Sep 15, 2003 at 03:13:30PM -0400, Didier Rwitura wrote: I am trying to upgrade my sysstem from 4.8-RELEASE-p4 to 5.1 and i getting this error message when I run "make -j4 buildworld" I had the same problem. Making without the "-j4" option works however. Antoine ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
pam_mkhomedir
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi ! I just set up a test server running FreeBSD-5.1 with pam_ldap and nss_ldap for authentication. It works really nice so far. I have a little question though: I wanted the user homedir to be created the first time a user logs in if it does not exist. I know I need pam_mkhomedir for this but I can't find it under 5.1. If I remember, pam_mkhomedir was in the contrib section under 4.x. Any idea why it is not part of FreeBSD anymore ? Or do you know any other way of auto-creating users homedir ? Thanks in advance, Regards. - -- Antoine Jacoutot [EMAIL PROTECTED] http://www.lphp.org PGP/GnuPG key: http://www.lphp.org/ressources/ajacoutot.asc -BEGIN PGP SIGNATURE- Version: GnuPG v1.2.3 (FreeBSD) iD8DBQE/UFuqY3Hnhkr+5cQRAoLpAJ9xFUoYKx6nre442AvA19+2kP8lyQCbBUIw q4sqpdjgOVwxUnhEqUkKdSY= =0zQy -END PGP SIGNATURE- ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Palm syncing over USB on FreeBSD, any hopes?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Tuesday 29 July 2003 19:24, Rajappa Iyer wrote: > ucom0: Palm, Inc. Palm Handheld, rev 1.00/1.00, addr 2 > ucom0: init failed, STALLED > device_probe_and_attach: ucom0 attach returned 6 Try this: http://www.lphp.org/popups/articleswindow.php?id=13 It's been working great for me since FreeBSD-4.7 (now running 4.8) with a Palm m505. Antoine -BEGIN PGP SIGNATURE- Version: GnuPG v1.2.2 (FreeBSD) iD8DBQE/JrB7Y3Hnhkr+5cQRAoMnAJ9AZzDHPCwo4msdpTIt7wcxK/6lUwCffo04 8zgFdNxeD3VzLgFC8dsqW7k= =/SEh -END PGP SIGNATURE- ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "[EMAIL PROTECTED]"
