On 2 Jun 2009, at 03:23, Chuck Fisher wrote:
Robert wrote:My install of Clamd has been crashing regularly. 5 times yesterday, 28 times and counting today. # Version = ClamAV 0.95.1/9411/ from source. System = Linux 2.6.26.8-57.fc8 #1 SMP Thu Dec 18 19:19:45 EST 2008 i686 i686 i386 GNU/Linux # Can any one help ?? Much Thanks, Robert.Just to get a little more information.If the script and freshclam are not triggering the reload, what is? Andhow? By killing the pid or reload command, etc.
I think the SelfCheck option in clamd.conf was forcing a reload. All the crashes occured after an instance of SelfCheck running. (According to my interpretation of the logs).
What does top (or similar) say you have in free memory when all your processes are running?
A quick 'top' just now gives the following..... Mem: 1034544k total, 912736k used, 121808k free, 39240k buffers Swap: 2031608k total, 95004k used, 1936604k free, 347516k cached
My own clam crash experience is a little hard to put together because Idid so many things within a short time frame. And my system was not crashing very often. I doubled the servers memory, changed how itreloaded and upgraded Clam. It was after I changed from reload to a pidkill that my system stopped crashing, but the other upgrades cameshortly after. So I cannot point my finger. Plus I was in somewhat of arush because I needed to get the system as rock solid as I could so I could get back surgery. Chuck
What I ended up doing was the following.... I disabled 'SelfCheck' in the clamd.conf file.Then enabled 'reload' in freshclam.conf and also in Bill Landry's script.
(I also checked that neither Freshclam or Bill's Script would execute at the same minute of the hour via cron) So far no crashes in the last 4hrs. (Whereas before I was getting around 3-4 crashes per hour). Fingers firmly crossed.... Robert.