Menu
▾
▴
[Bacula-bugs] [bacula 0001097]: bacula-sd won't start after upgrade from 2.2.8 to 2.4.0
|
From: <bac...@li...> - 2008-06-30 08:33:01
|
The following issue has been CLOSED ====================================================================== http://bugs.bacula.org/view.php?id=1097 ====================================================================== Reported By: harleymk2 Assigned To: ====================================================================== Project: bacula Issue ID: 1097 Category: Storage Daemon Reproducibility: always Severity: crash Priority: normal Status: closed Resolution: fixed Fixed in Version: 2.4.1 ====================================================================== Date Submitted: 06-13-2008 13:24 BST Last Modified: 06-30-2008 09:33 BST ====================================================================== Summary: bacula-sd won't start after upgrade from 2.2.8 to 2.4.0 Description: Hi, i just upgraded from 2.2.8 to 2.4.0 on solaris10. same configure as before. starting bacula-sd gives following error: bs1# /etc/init.d/bacula-sd start Starting the Bacula Storage daemon: 10-Jun 10:08 bacula-sd: Fatal Error because: Bacula interrupted by signal 10: Bus Error Kaboom! bacula-sd, bacula-sd got signal 10 - Bus Error. Attempting traceback. Kaboom! exepath=/opt/local/app/bacula/sbin/ Calling: /opt/local/app/bacula/sbin/btraceback /opt/local/app/bacula/sbin/bacula-sd 1822 gcore: cannot grab on: no such process gcore: cannot grab bs1: no such process gcore: /opt/local/app/bacula/working/bacula-sd.1822 dumped /opt/local/app/bacula/sbin/btraceback: dbx: not found Traceback complete, attempting cleanup ... Segmentation Fault reinstalling 2.2.8 makes everything ok I am using the File-Storage-device Is there any hint or help ? bye, Markus ====================================================================== ---------------------------------------------------------------------- ebollengier - 06-13-08 18:39 ---------------------------------------------------------------------- Hello, Could you try to make dbx working on your server ? > /opt/local/app/bacula/sbin/btraceback: dbx: not found Without the backtrace, it could be very long to fix this unwanted behavior. You can also start this daemon with -d200 to see something. (edit the bacula-sd script, add -d200 to OPTION='' and redirect the output to a file) ---------------------------------------------------------------------- kern - 06-14-08 20:29 ---------------------------------------------------------------------- Since Bacula has been built and tested on Solaris, I suspect this is a build error -- i.e. not a Bacula bug. Please attach the full log output of the build including the ./configure command, the make and the make install. ---------------------------------------------------------------------- harleymk2 - 06-16-08 11:06 ---------------------------------------------------------------------- Hi ebollengier, Installation of dbx is on the run. Hi kern, SVN (pls. put version in Build field) => 2.4.0 2.4.0 is not in the pull-down-menu attached are the wanted log files bye markus ---------------------------------------------------------------------- harleymk2 - 06-17-08 08:30 ---------------------------------------------------------------------- Hi ebollengier, here is the traceback: For information about new features see `help changes' To remove this message, put `dbxenv suppress_startup_message 7.6' in your .dbxrc Reading bacula-sd Reading ld.so.1 Reading libsec.so.1 Reading libz.so.1 Reading libpthread.so.1 Reading libresolv.so.2 Reading libnsl.so.1 Reading libsocket.so.1 Reading libxnet.so.1 Reading libintl.so.1 Reading libstdc++.so.6.0.3 Reading libm.so.2 Reading libgcc_s.so.1 Reading libc.so.1 Reading libavl.so.1 Reading libc_psr.so.1 Attached to process 9277 t@1 (l@1) stopped in _waitid at 0xfef419dc 0xfef419dc: _waitid+0x0008: ta %icc,0x00000008 Current function is signal_handler 167 waitpid(pid, NULL, 0); /* wait for child to produce dump */ ******** RUNNING LWPS/THREADS: >l@1 running in _waitid() ******** STACK TRACE OF CURRENT LWP: current thread: t@1 [1] _waitid(0x0, 0x243e, 0xffbfeee8, 0x3, 0xff3f06e0, 0x0), at 0xfef419dc [2] _waitpid(0x243e, 0x0, 0x0, 0x243e, 0xff2c2000, 0x0), at 0xfeee840c [3] waitpid(0x243e, 0x0, 0x0, 0x0, 0xfc00, 0xfef6ce40), at 0xfef348e4 =>[4] signal_handler(sig = 10), line 167 in "signal.c" [5] __sighndlr(0xa, 0x0, 0xffbff240, 0x83ee8, 0x0, 0x1), at 0xfef40620 ---- called from signal handler with signal 10 (SIGBUS) ------ [6] init_resource(type = 3002, items = 0xbac88, pass = 1), line 215 in "parse_conf.c" [7] parse_config(cf = 0xbd358 "/opt/local/app/bacula/etc/bacula-sd.conf", scan_error = (nil), err_type = 11), line 860 in "parse_conf.c" [8] main(argc = 0, argv = 0xffbffb74), line 207 in "stored.c" ******** VARIABLES DUMP OF CURRENT LWP: argv = ARRAY exelen = 27 sigdefault = CLASS btpath = "/opt/local/app/bacula/sbin/btraceback" pid = 9278 buf = "/tmp/" pid_buf = "9277" already_dead = 1 sig = 10 Current function is signal_handler 167 waitpid(pid, NULL, 0); /* wait for child to produce dump */ t@1 (l@1) stopped in _waitid at 0xfef419dc 0xfef419dc: _waitid+0x0008: ta %icc,0x00000008 ******** STACK TRACE OF LWP 1: [1] _waitid(0x0, 0x243e, 0xffbfeee8, 0x3, 0xff3f06e0, 0x0), at 0xfef419dc [2] _waitpid(0x243e, 0x0, 0x0, 0x243e, 0xff2c2000, 0x0), at 0xfeee840c [3] waitpid(0x243e, 0x0, 0x0, 0x0, 0xfc00, 0xfef6ce40), at 0xfef348e4 =>[4] signal_handler(sig = 10), line 167 in "signal.c" [5] __sighndlr(0xa, 0x0, 0xffbff240, 0x83ee8, 0x0, 0x1), at 0xfef40620 ---- called from signal handler with signal 10 (SIGBUS) ------ [6] init_resource(type = 3002, items = 0xbac88, pass = 1), line 215 in "parse_conf.c" [7] parse_config(cf = 0xbd358 "/opt/local/app/bacula/etc/bacula-sd.conf", scan_error = (nil), err_type = 11), line 860 in "parse_conf.c" [8] main(argc = 0, argv = 0xffbffb74), line 207 in "stored.c" ******** VARIABLES DUMP OF LWP 1: argv = ARRAY exelen = 27 sigdefault = CLASS btpath = "/opt/local/app/bacula/sbin/btraceback" pid = 9278 buf = "/tmp/" pid_buf = "9277" already_dead = 1 sig = 10 dbx: no LWP with id 2 dbx: no LWP with id 3 dbx: no LWP with id 4 dbx: no LWP with id 5 dbx: no LWP with id 6 dbx: no LWP with id 7 dbx: no LWP with id 8 detaching from process 9277 hope this helps bye, Markus ---------------------------------------------------------------------- kern - 06-17-08 09:31 ---------------------------------------------------------------------- After looking at the dbx output (thanks), this looks like a real bug -- possibly because some, but not all files were updated from the trunk. What surprises me is that this was not picked up by the regression testing (perhaps regression was only done on the trunk rather than the Branch). We should have a patch for this shortly, but also need to make sure the regressions are run on the 2.4 Branch. PS: I added 2.4.0 to the list immediately after I saw your bug and also updated the bug report appropriately -- thanks. Eric: I will coordinate with you via email. ---------------------------------------------------------------------- kern - 06-20-08 17:01 ---------------------------------------------------------------------- Would you please checkout the latest version of Branch-2.4 and try the code? The problem *should* now be fixed. svn export https://bacula.svn.sf.net/svnroot/bacula/branches/Branch-2.4/bacula will put the source in the directory bacula in the current directory. If you want to checkout the source (so you can update it later with "svn update") replace the word "export" with "checkout" (without the double quotes). ---------------------------------------------------------------------- harleymk2 - 06-24-08 07:36 ---------------------------------------------------------------------- Hi Kern, thank you very much. Cause my backup-server is not connected to the internet, can i also reload http://downloads.sourceforge.net/bacula/bacula-2.4.0.tar.gz?modtime=1212659202&big_mirror=1 ? Is that fixed too ? Bye, Markus ---------------------------------------------------------------------- kern - 06-30-08 09:33 ---------------------------------------------------------------------- To answer your question, no bacula-2.4.0.tar.gz contains 2.4.0 code and that will never change once released. You can either get the current code by checking it out via the svn command that I previously gave you -- you can move the directory to another machine if you want, or you can wait for the 2.4.1 beta that will be released shortly. We were never able to satisfactorily reproduce your problem, so it must have something to do with the compiler you are using; its version; or the options. In any case, for version 2.4.1, there should be no bus error. If there are please re-open this bug report or create a new one. Issue History Date Modified Username Field Change ====================================================================== 06-13-08 13:24 harleymk2 New Issue 06-13-08 18:39 ebollengier Note Added: 0003364 06-14-08 20:29 kern Note Added: 0003367 06-14-08 20:29 kern Status new => feedback 06-14-08 21:31 kern version SVN (pls. put version in Build field) => 2.4.0 06-16-08 11:06 harleymk2 Note Added: 0003375 06-16-08 11:07 harleymk2 File Added: LOG 06-17-08 08:30 harleymk2 Note Added: 0003379 06-17-08 09:31 kern Note Added: 0003380 06-20-08 17:01 kern Note Added: 0003385 06-24-08 07:36 harleymk2 Note Added: 0003393 06-30-08 09:33 kern Note Added: 0003423 06-30-08 09:33 kern Status feedback => closed 06-30-08 09:33 kern Resolution open => fixed 06-30-08 09:33 kern Fixed in Version => 2.4.1 ====================================================================== |