Re: Agents Constantly Reporting Start Events

[Martin Pala]

On May 9, 2011, at 11:57 AM, Kevin Chadwick wrote:
> 
> I believe it's a bug introduced sometime before 5.2.3. With eventqueue
> enabled, it may well happen again the next time something changes.
> Either disable eventqueue or set up a clear up script to occur before
> the next round of checks or covering two. Or search for the bug of
> course.


The event queue stores the event in the case that the delivery to mmonit or 
SMTP server failed ... the delivery is retried next cycle and the event is 
removed as soon as the delivery succeeded. 

Excerpt from Christopher's monit log and request timeout on M/Monit side:
--8<--
[UTC May  8 20:51:11] error    : M/Monit: communication failed (event message)
[UTC May  8 20:51:11] error    : M/Monit handler failed, retry scheduled for 
next cycle
[UTC May  8 20:52:17] error    : M/Monit: error receiving data from 
http://{our_monit_url}:{our_monit_port}/collector -- 
--8<--

=> the eventqueue works as designed - the network trace could help to root 
cause why the request timed out.

Regards,
Martin