OP-Manager not able to Poll DMZ zone network quite often

[virgoboy009]

Hello Guys,

We have two monitoring Tools OP-Manager 6.0 & Solar Winds Engg edition ver9.0 in our network which monitors Servers, Networks in OP-Manager & Network in Solar winds.

From past few we are seeing that our OP-manager sends communication alert in the tool that all the server in DMZ are down and it will immediately come-up after 2-3mins,

However None of the servers in reality were shut down nor rebooted and no production impact were observed.

Also when we check the DMZ switch connected to these server ports are looks ok and switch did not reboot when checked the up-time & cpu utilization is ok.

But however we are suspecting that may be due to high data transfer between servers in DMZ zone it is causing the OP-manager to  fail to talk the DMZ zone servers.

Note:-
OP-Manager & Solar winds servers are connected in the Inside network.
We are also not seeing any major utilization on the DMZ switch 1 connected to external firewall.
It was only the DMZ switch 2 which is connected to DMZ switch1 as uplink and connected to servers showing high BW  utilization.

Please assist me in Isolating this issue.

I would highly appreciate all your suggestions!

Regards,
KA

[contech-nelsong]

If you're losing monitoring traffic because things get too busy, then to my mind you've got 2 options:

Treat this as false positive: reduce the sensitivity of the monitoring, longer timeouts more retries etc..

Treat this as a symptom of a lack of capacity: check the servers and the local switch for resource issues and upgrade appropriately.

If you've got an excessive transfer rate causing monitoring packets to be lost, then in theory you should be seeing those drops somewhere, if the server is too slow then the output queue on the switch ports that connect to those servers should be dropping packets.

If the switch is too slow then you should be seeing excessive utilisation somewhere on the switch; CPU, memory or interfaces.