I had this problem too. It was a combination of increasing the number of vCPU for the virtual machine (to prevent CPU overload) and to increase concurrent pollers (to finish the polling in less than 5 minutes).

Is your CPU load higher than your number of CPUs? (you can see this using "top" command)

top - 18:47:51 up 16 days, 4:09, 1 user, load average: 8.41, 10.53, 9.64

In my example, I have 8 vCPU, and my load is above 8, which means I might need to add more vCPU.

The time to complete polling is seen by the output of your /etc/crontab entry. I redirect this to a file so I can read it.:

*/5 * * * * root /opt/observium/poller-wrapper.py 13 >> /opt/observium/poller-wrapper.log 2>&1

Tristan

Tristan Rhodes
Network Engineer

Weber State University

801.626.8549

On Wed, Sep 24, 2014 at 12:16 PM, chott@praha1.net <chott@praha1.net> wrote:

Hello guys

Last two week i have problem with Graph blank places for few times

Observium is connected by localy switch gigabit line same problem in localy SNMP for server

Where is problem? Disk? RAM ? SQL?

Tomas

Tomas Chott
technický ředitel

chott@praha1.net
Phone: +420 245 004 005
Mobile: +420 604 930 119

Certifikace:
Cisco CXFS, CCDA, SMBEN, SMBAM, CCNA
Ubiquiti UCWA, UEWA, UCRSS, UCT
Authorised Trainer Cisco, Ubiquiti, TP-Link, Synology

Metropolitní síť Praha 1 z.s.p.o.
Jindrisska 18
110 00 Praha 1
Czech Republic
www.praha1.net

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