I just had a report from one of my sysadmins of a similar problem
under FreeBSD.
The load was up around 40, almost all of it spent in interrupts, all
caused by clamav.
We were using the libmap.conf trick that had prevented bad behavior
under 0.9. It's the first time the trouble has occurre
no obvious evidence
of the problem in the clam logs.
I'd greatly appreciate any suggestions on how best to troubleshoot this
issue.
Kevin Way
InsideSystems
___
http://lurker.clamav.net/list/clamav-users.html
t just sits there ignoring TERM, burning CPU and not
logging anything in particular. I'm not sure what to do at that point,
to extract any useful debugging data.
Regards,
Kevin Way
___
http://lurker.clamav.net/list/clamav-users.html
René Berber wrote:
Kevin Way wrote:
[snip]
I'm using exim as my MTA, and calling clam on every message with a
simple av_scanner = clamd:/var/run/clamav/clamd
This is wrong, clamd is the daemon not the scanner.
You should be using either clamdscan or clamscan.
clamd indicate
tailspin.
Hmm... It's hard to tell precisely (or even which of about 30) messages
might have been the last to hit the scanner. I'll ask our resident exim
expert if he can archive all attachments before they're sent to the
scanner, so we might have a shot at isolating the probl
Clamd only seems to spiral out of control while running on SMP-enabled
machines. We've been able to toggle between stable and awful behavior
by turning SMP support on and off in the kernel.
___
http://lurker.clamav.net/list/clamav-users.html
Dennis Peterson wrote:
>> So, its been a few days. How is everyone feeling about the new version?
>> I've hesitated to upgrade just yet. I've seen alot of feedback indicating
>> problems and very little about smooth and great upgrades.
>>
>> What's the general concensous - You can't upgrade fast en
md dies? I have not reverted back
>>
>
> At least some basic information, eg. how you compiled ClamAV, did you use
> --enable-experimental, platform details, etc.
>
>
I had the same problems. FreeBSD/i386 6.2p1 and 6.2p0. Compiled
directly from the ports collection, with
Craig Green wrote:
>
> Kevin Way wrote:
>>
>> I had the same problems. FreeBSD/i386 6.2p1 and 6.2p0. Compiled
>> directly from the ports collection, without --experimental.
>>
>> As far as we could tell, it just scanned slower and slower and then
>> s
Kevin Way wrote:
> Craig Green wrote:
>
>> Kevin Way wrote:
>>
>>> I had the same problems. FreeBSD/i386 6.2p1 and 6.2p0. Compiled
>>> directly from the ports collection, without --experimental.
>>>
>>> As far as we could tell,
10 matches
Mail list logo
