Note this will make it very very long to poll all devices... it should not cross 5 minutes or your system will explode. Check full poll time in the webinterface under "polling information", bottom line.

Running with a parallel set each logging to its own file could help there.

On 10/22/2014 06:28 PM, Mike Stupalov wrote:

Can change your cron from poller-wrapper to base poller.php with debug?

ie:
*/5 * * * *  /usr/bin/php /opt/observium/poller.php -d -h all >> /tmp/poller.debug

(don't forget temporary disable poller-wrapper)

This is fully unoptimized polling and will creates very huge debug output, but it may be able to catch the error.

After you get some times device up/down (by snmp), return back poller-wrapper and send me file /tmp/poller.debug

On Wed, Oct 22, 2014 at 11:42 AM, Rutger Bevaart <rutger@xconnect24.com> wrote:

On 22 Oct 2014, at 09:38, observium-request@observium.org wrote:

?

On Wed, Oct 22, 2014 at 10:53 AM, Rutger Bevaart <rutger@xconnect24.com>
wrote:

Hi all,

I can confirm that the issue is gone in 5903!

Thanks Mike & Joe for the excellent input (and fix)!

Let me know if any additional input, debugging or testing.

Regards,
Rutger

I have been too quick to judge, the issue is still there, although it seems to be less. 

Ideas?

--
Mike Stupalov
http://observium.org/

_______________________________________________
observium mailing list
observium@observium.org
http://postman.memetic.org/cgi-bin/mailman/listinfo/observium