>>>>> On Fri, 12 Nov 2010 19:55:53 -0500, AndyW said:
>
> Martin Simmons wrote:
> >
> > >
> > > >
> > > > >
> > > > > >
> > > > > > > On Mon, 08 Nov 2010 16:50:46 -0500, AndyW said:
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> > > I have been using Bacula for a few months now happily on x86_64 FC12 &
> > > FC13.
> > > I have just now upgraded to FC14, duing which my Bacula packages got
> > > upgraded to 5.0.2-8.fc14, but am now unable to start bacula-sd (the
> > > storage
> > > daemon). The service dies as soon as I start it and I get a message
> > > "bacula-sd dead but subsys locked" if I query the status. Whenever I
> > > query
> > > the status or stop bacula-sd - but not when I (re)start it - I get the
> > > following syslog messages referencing libc-2.12.90.so:- root < at >
> > > localhost
> > > init.d]# ./bacula-sd status bacula-sd dead but subsys locked
> > >
> > > Nov 8 21:24:47 localhost kernel: [161155.229044] pidof[17514]: segfault
> > > at 0 ip 00007f2f9097d316 sp 00007fffb83e2888 error 4 in
> > > libc-2.12.90.so[7f2f908fc000+199000]
> > > Nov 8 21:24:47 localhost kernel: [161155.271610] pidof[17515]: segfault
> > > at 0 ip 00007f8835f08316 sp 00007fff931ff828 error 4 in
> > > libc-2.12.90.so[7f8835e87000+199000]
> > > libc-2.12.90.so is part of GNU libc. I have both x86_64 and i686
> > > versions, and a total of three instances of the file:
> > > [root < at > localhost etc]# find / -name libc-2.12.90.so -exec ls -l {}
> > > \;
> > > -rwxr-xr-x. 1 root root 1989560 Oct 22 18:50 /lib64/libc-2.12.90.so
> > > -rwxr-xr-x. 1 root root 1906468 Oct 22 18:59
> > > /lib/i686/nosegneg/libc-2.12.90.so
> > > -rwxr-xr-x. 1 root root 1886052 Oct 22 18:59 /lib/libc-2.12.90.so
> > >
> > > All are on version level glibc-2.12.90-18.
> > >
> > > Are there any known issues with this version of glibc? Anyone else have
> > > the
> > > same problem with bacula / FC14 - or alternatively have it all working
> > > fine?
> > > The other parts of bacula (bacula-dir and bacula-fd) are fine and the tape
> > > drive (Tandberg SLR7) works fine with mt and tapeinfo. Kernel is
> > > 2.6.35.6-48.fc14.x86_64
> > >
> >
> > I suggest you install the Fedora glibc-debuginfo package and then run
> > bacula-sd under gdb to get a backtrace:
> >
> > http://www.bacula.org/5.0.x-manuals/en/problems/problems/What_Do_When_Bacula.html#SECTION00640000000000000000
> >
> > __Martin
> >
>
>
> Thanks for the tip. This is the output I get
> [r...@localhost sbin]# cd /usr/sbin
> [r...@localhost sbin]# gdb ./bacula-sd
> GNU gdb (GDB) Fedora (7.2-23.fc14)
> Copyright (C) 2010 Free Software Foundation, Inc.
> License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
> This is free software: you are free to change and redistribute it.
> There is NO WARRANTY, to the extent permitted by law. Type "show copying"
> and "show warranty" for details.
> This GDB was configured as "x86_64-redhat-linux-gnu".
> For bug reporting instructions, please see:
> <http://www.gnu.org/software/gdb/bugs/>...
> Reading symbols from /usr/sbin/bacula-sd...Reading symbols from
> /usr/lib/debug/usr/sbin/bacula-sd.debug...done.
> done.
> (gdb) run -s -f -c /etc/bacula/bacula-sd.conf
> Starting program: /usr/sbin/bacula-sd -s -f -c /etc/bacula/bacula-sd.conf
> [Thread debugging using libthread_db enabled]
> Detaching after fork from child process 30113.
> [New Thread 0x7ffff2132700 (LWP 30114)]
> [New Thread 0x7ffff1931700 (LWP 30115)]
> [Thread 0x7ffff2132700 (LWP 30114) exited]
> [New Thread 0x7ffff2132700 (LWP 30124)]
> [Thread 0x7ffff2132700 (LWP 30124) exited]
>
> In the debugger bacula-sd starts and appears to work normally (in that Bat
> console can connect to it) although when I query the daemon status is says
> it is stopped, as opposed to dead previously.
>
> It turns out that if I add the "-s" command line option (for "Do not trap
> signals") to the init.d daemon script then bacula-sd will start and function
> correctly, though it then returns status of dead and cannot be stopped using
> the script & has to be killed instead. I also get the syslog error messages
> about libc when I query the status or try to stop the daemon.
Ah, that's a clue.
Looking at the output again, I see that the crash actually occurs in pidof,
not bacula-sd. Either the startup script is misusing it or pidof is broken.
I suggest you report it to Fedora.
__Martin
------------------------------------------------------------------------------
Centralized Desktop Delivery: Dell and VMware Reference Architecture
Simplifying enterprise desktop deployment and management using
Dell EqualLogic storage and VMware View: A highly scalable, end-to-end
client virtualization framework. Read more!
http://p.sf.net/sfu/dell-eql-dev2dev
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
