Dennis Peterson wrote:
> Bill Landry wrote:
> 
>> I'm thinking that the times that clamd crashed on my systems most likely
>> correspond to the times that SelfCheck ran during a script update or a
>> freshclam update, which is why the crashes happened so randomly (I can't
>> prove this now because I didn't maintain a record of my clamd crash
>> date/times).
> 
> In my clamd log I see this when clamd starts:
> clamd daemon 0.95rc1 (OS: solaris2.9, ARCH: sparc, CPU: sparc)
> 
> That may be useful to locate the start times as an indirect method of finding 
> crash times.

I did find several corresponding log entries, however, the system that I
was experience the crashes on was a test system that I also used for
testing my scripts, where I might execute a "kill 9" against clamd's pid
to confirm that my script could restart clamd with the orphaned
lock/pid/socket files left, like a real clamd crash might do.

However, for others, this might help:

grep -B1 "Started at" /your/path/to/clamd.log

Typically if you manually restart clamd, you should see a stop line and
then a start line.  But when clamd crashes, you will typically see
something like:

Wed Jan  7 15:53:21 2009 -> Reading databases from /var/lib/clamav
Wed Jan  7 16:37:15 2009 -> +++ Started at Wed Jan  7 16:37:15 2009

There will usually be a time gap between the "Reading database" line and
the next line "+++ Started at", which is the time difference between
when clamd crashed and the user or script restarted clamd.

Hope this helps...

Bill
_______________________________________________
Help us build a comprehensive ClamAV guide: visit http://wiki.clamav.net
http://www.clamav.net/support/ml

Reply via email to