Syslog-ng using a spectacular amount of CPU time... (I'm using sshguard)
On Mon, Mar 23, 2009 at 2:49 PM, Steve <firstname.lastname@example.org> wrote:
> Steve wrote:
>> Do others get this behaviour - is this a bug in syslog-ng?
> Sorry for the multiple posts... a slight error on my part. The sshguard
> process wasn't running - a /bin/sh process trying to spawn it was running
> (there was no link from /usr/local... to the binary) and when the binary
> failed to execute - syslog-ng got itself into a tiz. Everything seems to
> work fine when I correct the path to the program.
> Problem solved - but, I guess, this is a flaw in syslog-ng... I'd have hoped
> it would generate an error message rather than behave as it did.
I had a possibly similar problem a while back with syslog-ng going
crazy when a certain daemon would crash (in my case it filled up the
log wit about 60 gigabytes of the same thing repeated over and over,
in addition to using massive CPU%). I switched to metalog and haven't
had any problems since.