ClamAV Module Core Dump

[Scott Silva]

Zivago Lee spake the following on 5/1/2007 3:01 PM:
>> Drew Marshall spake the following on 5/1/2007 1:56 PM:
>>> Hi all
>>>
>>> Ok another challenge that I hope you can help with. I am running a
>>> FreeBSD 6 box that I have just upgraded to ClamAV 0.90.2 which was
>>> upgraded from the ports tree. Knowing there can be issues, I also forced
>>> an update (Which in effect recompiles) of the Clam perl module.
>>> Restarted MailScanner and ever since MS core dumps. I can run MS using
>>> ClamAV only but it makes quite a performance hit so I want to get t
>>> reinstalled ASAP.
>>>
>>> Running in debug mode, I helpfully get:
>>>
>>> root at mx1 /usr/local/etc/MailScanner # mailscanner --debug
>>> In Debugging mode, not forking...
>>> Segmentation fault
>>>
>>> /var/log/messages gets:
>>>
>>> May  1 20:34:14 mx1 kernel: pid 43095 (perl5.8.8), uid 0: exited on
>>> signal 11 (core dumped)
>>>
>>> Again no clues that I can see. Can any one give me some ideas where to
>>> start with this?
>> I have a similar issue in Centos 4. While I am not getting cores,
>> MailScanner
>> will not run with the 0.20 module and 0.90.2 on one of my servers. It runs
>> fine with the commandline scanner, and I haven't had the time to upgrade
>> to
>> the new version of MailScanner to test the clamd functionality. I get
>> nothing
>> in debug, either.
> 
> Wow.. I thought it was just my server and/or config on why the
> clamavmodule was not working.  It would just keep restarting the
> MailScanner processes and -debug would not display any issues.  clamd
> works just fine, however, and works really well!
> 
I have tried installing more than once, and I am near to wiping out the clam
library and installing again. I might try clamd first, as anything has to be
better than the commandline scanner. I am already running a few scanners, and
every bit of load adds up.

-- 

MailScanner is like deodorant...
You hope everybody uses it, and
you notice quickly if they don't!!!!