Re: Server failure

[sasha]

Tomas Styblo wrote:
> 
> So, I tried to build the server from the sources, but it did not help. I
> also upgraded glibc to its latest version before. I also tried to find
> out if the problem is related to some particular query or part of code
> in my client, and I found, that it is not.
> 
> I checked kernel sources and tried to find what does the "sigpending
> lied" error message means, it is only once there, in this part of code
> in signal.c :
> 
>               /* XXX: Once CLONE_PID is in to join those "threads" that
> are
>                  part of the same "process", look for signals sent to
> the
>                  "process" as well.  */
> 
>               /* Sanity check... */
>               if (mask == &current->blocked &&
> signal_pending(current))
> {
>                   printk(KERN_CRIT "SIG: sigpending lied\n");
>                   current->sigpending = 0;
>               }
> 
> I am almost sure, that mysql conflicts in some way with some other
> threaded process, that uses signals - it could probably can be the
> Sybase, which is very thread intensive and runs on this machine too.
> 
> I dont know if this is bug in mysql or kernel, I will modify the
> signal.c to output the related values (signal_pending(current),
> &current->blocked) to the syslog, to find out more.
> 
> Tomas Styblo
> admin@stripped

Is the problem repeatable?

Have you tried it with Sybase down?

-- 
Sasha Pachev

+------------------------------------------------------------------+
|     /*/\*\/\*\   /*/ \*\ /*/ \*\ |*|     Sasha Pachev            |
|    /*/ /*/ /*/   \*\_   |*|   |*||*|     mailto:sasha@stripped  |
|   /*/ /*/ /*/\*\/*/  \*\|*|   |*||*|     Provo, Utah, USA        |
|  /*/     /*/  /*/\*\_/*/ \*\_/*/ |*|____                         |
|  ^^^^^^^^^^^^/*/^^^^^^^^^^^\*\^^^^^^^^^^^                        |
|             /*/             \*\ Developers Team                  |
+------------------------------------------------------------------+