One thing that confuses me here is that you have received a SIGPIPE
signal (OK, that is reasonable with the fclose), but Bacula is set to
ignore signals, so I don't understand it.
You might try running the debugger but not using the -s option on the
run command, and if any signal occurs simply type "cont". This may
produce better results and get us closer to the real problem.
As previously mentioned, you probably should also do a:
mv /lib/tls /lib/tls-broken
and try Bacula again to see if the new native Linux threads is the
On Sun, 2004-05-30 at 03:49, Jeff Layton wrote:
> Ok, here's the original message I sent. Let me know if there's any other
> info I can provide...
> BTW: I didn't make it clear in my original email, but I'm using
> -- Jeff
> From: Jeff Layton <jtlayton@...>
> To: bacula-users@...
> Subject: [Bacula-users] bacula-dir crash
> Date: Sun, 23 May 2004 09:04:38 -0400
> I'm having some trouble getting bacula-dir to stay running. Essentially
> what I'm seeing is that it's runs just fine until I try to kick off a
> backup, and then it crashes.
> I'm running bacula 1.34.2, on a 2.6.6 kernel.
> I ran a non-stripped binary under the debugger, and this is what I see:
> (gdb) run -s -f -c /etc/bacula/bacula-dir.conf -u bacula -g bacula
> Starting program: /tmp/bacula-dir -s -f -c /etc/bacula/bacula-dir.conf
> -u bacula -g bacula
> [Thread debugging using libthread_db enabled]
> [New Thread 1076544608 (LWP 6589)]
> [New Thread 1084939184 (LWP 6592)]
> [New Thread 1093331888 (LWP 6593)]
> [New Thread 1101724592 (LWP 6598)]
> [New Thread 1110117296 (LWP 6604)]
> [New Thread 1118903216 (LWP 6606)]
> Detaching after fork from child process 6609.
> Program received signal SIGPIPE, Broken pipe.
> [Switching to Thread 1118903216 (LWP 6606)]
> 0xffffe410 in __kernel_vsyscall ()
> (gdb) bt
> #0 0xffffe410 in __kernel_vsyscall ()
> #1 0x402409db in __write_nocancel () from /lib/tls/i686/cmov/libc.so.6
> #2 0x401e10c9 in _IO_new_file_write () from
> #3 0x401e02ef in new_do_write () from /lib/tls/i686/cmov/libc.so.6
> #4 0x401e0288 in _IO_new_do_write () from /lib/tls/i686/cmov/libc.so.6
> #5 0x401dfb87 in _IO_new_file_close_it () from
> #6 0x401d49ff in fclose@... () from /lib/tls/i686/cmov/libc.so.6
> #7 0x0808022c in close_bpipe (bpipe=0x80c1f00) at bpipe.c:168
> #8 0x08087a38 in dispatch_message (jcr=0x80c10e8, type=10, level=0,
> msg=0x80c34b4 "salusa-sd: Job salusa.2004-05-23_08.59.45 waiting.
> Cannot find any appendable volumes.\nPlease use the \"label\" command
> to create a new Volume for:\n Storage: DLT4000\n Media
> type: DLT4000\n "...)
> at message.c:624
> #9 0x080547a4 in bget_dirmsg (bs=0x80c48e0) at getmsg.c:167
> #10 0x08059d58 in msg_thread (arg=0x80c10e8) at msgchan.c:228
> #11 0x4002aae5 in start_thread () from
> #12 0x00000000 in ?? ()
> Any ideas what it could be? Please let me know if you need any other
> info to help diagnose the problem.