You're missing about 80% of the data needed to make that estimation.

Are they Linux devices with a single interface, or Cisco 6513s with 700 interfaces?

Either way, you should reduce the number of poller threads, as you're probably causing load spikes on that server for no good reason.

Thanks,
Adam.

Sent from BlueMail

On 25 Aug 2016, at 04:09, Gabriel Zamorski <gabriel.zamorski@domeny.pl> wrote:
Hello,

I’m using Observium on a virtual machine, with rrdtool 1.6.0 and rrdcached. For now, it takes up to 80 seconds to poll 87 devices. I'm just curious - is it good or bad time?

poller-wrapper.py(10710): /opt/observium/poller-wrapper.py: polled 87 devices in 77.74 seconds with 12 workers

Regards,






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