Richard A Nelson escribió:
> On Fri, 8 May 2009, Javier Kohan wrote:
>
>> Package: libmilter1.0.1
>> Version: 8.14.3-5
>> Severity: important
>>
>>
>> Sometimes I get errors like this:
>> May 8 19:15:16 rama1 kernel: [723703.161848] milter-greylist[4544]:
>> segfault at 130 ip 7f5d46f67900 sp 44ff10f0 error 4 in
>> libmilter.so.1.0.1[7f5d46f5c000+f000]
>> May 8 19:15:16 rama1 kernel: [723703.162967] mimedefang[4639]:
>> segfault at 130 ip 7fe6f9b20900 sp 427670f0 error 4 in
>> libmilter.so.1.0.1[7fe6f9b15000+f000]
>
>> I?'m using sendmail, milter-greylist and mimedefang.
>
> Ditto, and haven't seen that error yet (on either x86 or amd64).
Yep. I have several servers running this config with etch, and a few (2
o 3) with lenny, and this one is the only where it happens.
>
>> /home, /var/spool/mq* and /etc/mail are on a drbd volume but neither
>> /var/spool/MIMEDefang nor /var/run/milter-greylist are.
>> The hardware is a HP DL360, 3x 146G SAS (hardware) RAID5.
Forgot to mention, we use heartbeat combined with drbd to give HA to
mail and web.
As I mentioned before, mimedefang an greylist sockets are on plain block
devices, not drbd.
> Nice
>>
>> Thanks in advance.
>
> At this point, all I do is ask you to install the libmilter1.0.1-dbg
> package will give more information on the next segfault.
>
Done (both libmilter and libmilter-dbg installed, is it right ?).
Do you need me to do something else to help find the problem ?
It didn´t happen again, but last time took more than a week to repeat.
For now, I installed a script running from cron to restart the processes
in case they go down.
> I am on the road for the next two weeks, but will check mail as often as
> possible
>
Thanks again.
Regards,
Javier
--
To UNSUBSCRIBE, email to [email protected]
with a subject of "unsubscribe". Trouble? Contact [email protected]
