From: Ralf G. <Ral...@ra...> - 2010-02-17 08:22:26
|
Marco van Wieringen schrieb: > > a private email as I read the bacula-dev list via the web. ok > You state in a mail to the mailing list this: > > > Hi, > > > > I can reproducible crash bacula-dir when starting a backup on a client with > > Accurate backups enabled. > > > > The dir is running debian lenny + bacula 3.0.2 from backports (did also try > > self compiled 3.0.3), psql 8.4. > > > > The client is running Red Hat Enterprise Linux 4.8, with bacula-fd build from > > bacula-3.0.3-1.src.rpm. I also tried a rpm package from > > > > ftp://ftp.pbone.net/mirror/ftp.pramberger.at/systems/linux/contrib/rhel4/i386 > > > > What me makes wonder first, I see 3 bacula-fd processes on the client: > > > > -bacula-fd(8128)---bacula-fd(8129)---bacula-fd(8130) > That is strange and on strange ports. That's part of the output of pstree, so the numbers are the pid's not ports. Sorry for not beeing clear. > > Without "Accurate = yes" the backup runs fine. Accurate backup was enabled > > weeks ago a couple of backup were running fine with this option enabled. > > So I don't know what has changed since last week. > > > > Here's the level 150 dir debug output. > > > > VUMEM004-dir: fd_cmds.c:322-19628 >filed: VUMEM004-dir: getmsg.c:138-19628 bget_dirmsg 14: 2000 OK level > > VUMEM004-dir: getmsg.c:138-19628 bget_dirmsg 47: 3010 Job VU0EM003.2010-02-17_00.06.00_03 start > > VUMEM004-dir: getmsg.c:138-19628 bget_dirmsg 56: Status Job=VU0EM003.2010-02-17_00.06.00_03 JobStatus=82 > > VUMEM004-dir: getmsg.c:138-19628 bget_dirmsg 16: 2000 OK storage > > VUMEM004-dir: fd_cmds.c:581-19628 bdird: sending runscripts to fd > > VUMEM004-dir: util.c:723-19628 edit_job_codes: %c > > VUMEM004-dir: fd_cmds.c:586-19628 bdird: runscript %c -> VU0EM003 > > VUMEM004-dir: fd_cmds.c:592-19628 bdird: sending runscripts to fd '/opt/scripts/bacula/find.sh server' > > VUMEM004-dir: fd_cmds.c:555-19628 bdird: sending old runcommand to fd \ > > '/opt/scripts/bacula/find.sh^Aserver' > > VUMEM004-dir: getmsg.c:138-19628 bget_dirmsg 162: Jmsg Job=VU0EM003.2010-02-17_00.06.00_03 type=6 \ > > level=1266361563 VU0EM003 JobId 19628: shell command: run ClientRunBeforeJob "/opt/scripts/bacula/find.sh \ > > server" > > VUMEM004-dir: getmsg.c:138-19628 bget_dirmsg 137: Jmsg Job=VU0EM003.2010-02-17_00.06.00_03 type=6 \ > > level=1266361563 VU0EM003 JobId 19628: ClientRunBeforeJob: Fuehre Skript fuer server aus > > VUMEM004-dir: getmsg.c:138-19617 bget_dirmsg -1: > > VUMEM004-dir: getmsg.c:138-19617 bget_dirmsg -1: > > VUMEM004-dir: getmsg.c:138-19627 bget_dirmsg 129: Jmsg Job=SMTCZB0003.2010-02-17_00.05.00_02 type=6 \ > > level=1266361622 SMTCZB0003-fd JobId 19627: ClientRunBeforeJob: Starte MKS... > > VUMEM004-dir: getmsg.c:138-19627 bget_dirmsg 18: 2000 OK RunBefore > > VUMEM004-dir: util.c:723-19627 edit_job_codes: %c > > VUMEM004-dir: fd_cmds.c:586-19627 bdird: runscript %c -> SMTCZB0003 > > VUMEM004-dir: fd_cmds.c:592-19627 bdird: sending runscripts to fd \ > > '/home/mksadmin/backup/script/db2_backup_offline_post.sh' > > VUMEM004-dir: fd_cmds.c:555-19627 bdird: sending old runcommand to fd \ > > '/home/mksadmin/backup/script/db2_backup_offline_post.sh' > > VUMEM004-dir: getmsg.c:138-19627 bget_dirmsg 17: 2000 OK RunAfter > > /usr/sbin/bacula-dir: symbol lookup error: /usr/sbin/bacula-dir: undefined symbol: \ > > _Z22db_accurate_get_jobidsP3JCRP4B_DBP7JOB_DBRPc > > These kind of errors point to the fact that your shared libraries are not inline with the version > of the director you are running. > > So if you say I compiled things myself and I still get this error make sure all stuff from > the rpms is gone (e.g. no left over bacula libraries libbac*) And then compile and > run a make install. Hm, I just installed my self build/compiled bacula 3.0.3 version again (debian deb package). On the dir: $find /usr/lib -name libbac* -print0 | xargs -0 ls -cl -rw-r--r-- 1 root root 923 17. Feb 09:09 /usr/lib/bacula/libbaccfg.la lrwxrwxrwx 1 root root 18 17. Feb 09:09 /usr/lib/bacula/libbaccfg.so -> libbaccfg.so.1.0.0 lrwxrwxrwx 1 root root 18 17. Feb 09:09 /usr/lib/bacula/libbaccfg.so.1 -> libbaccfg.so.1.0.0 -rw-r--r-- 1 root root 32848 17. Feb 09:09 /usr/lib/bacula/libbaccfg.so.1.0.0 -rw-r--r-- 1 root root 929 17. Feb 09:09 /usr/lib/bacula/libbacfind.la lrwxrwxrwx 1 root root 19 17. Feb 09:09 /usr/lib/bacula/libbacfind.so -> libbacfind.so.1.0.0 lrwxrwxrwx 1 root root 19 17. Feb 09:09 /usr/lib/bacula/libbacfind.so.1 -> libbacfind.so.1.0.0 -rw-r--r-- 1 root root 50512 17. Feb 09:09 /usr/lib/bacula/libbacfind.so.1.0.0 -rw-r--r-- 1 root root 912 17. Feb 09:09 /usr/lib/bacula/libbac.la -rw-r--r-- 1 root root 956 17. Feb 09:09 /usr/lib/bacula/libbacpy.la lrwxrwxrwx 1 root root 17 17. Feb 09:09 /usr/lib/bacula/libbacpy.so -> libbacpy.so.1.0.0 lrwxrwxrwx 1 root root 17 17. Feb 09:09 /usr/lib/bacula/libbacpy.so.1 -> libbacpy.so.1.0.0 -rw-r--r-- 1 root root 15672 17. Feb 09:09 /usr/lib/bacula/libbacpy.so.1.0.0 lrwxrwxrwx 1 root root 15 17. Feb 09:09 /usr/lib/bacula/libbac.so -> libbac.so.1.0.0 lrwxrwxrwx 1 root root 15 17. Feb 09:09 /usr/lib/bacula/libbac.so.1 -> libbac.so.1.0.0 -rw-r--r-- 1 root root 336776 17. Feb 09:09 /usr/lib/bacula/libbac.so.1.0.0 -rw-r--r-- 1 root root 947 12. Feb 14:52 /usr/lib/bacula/libbacsql.la lrwxrwxrwx 1 root root 18 12. Feb 14:52 /usr/lib/bacula/libbacsql.so -> libbacsql.so.1.0.0 lrwxrwxrwx 1 root root 18 12. Feb 14:52 /usr/lib/bacula/libbacsql.so.1 -> libbacsql.so.1.0.0 -rw-r--r-- 1 root root 122864 12. Feb 14:52 /usr/lib/bacula/libbacsql.so.1.0.0 How can I see that these are the right files? > > I've set the debug level on the client to 200, the debug log file is created but empty. > > > > I read Kaboom chapter in the manual but I can't get it to work on the dir. > > Before I put more time in this problem, is there anything obvious in the dir > > debug output (well, it's not that much...)? > Debugging this much further doesn't solve anything make sure your libraries are ok or > compile using --disable-libtool and get shared libraries out of the equation. Will try this next, thanks for your help. Ralf |