High CPU utilization is been observed in few cisco switches and that is not allowing to login to those switches and reason due to SNMP
Can i know whether any one faced the same issue and what was the solution for that.
are u polling those swithces from multiple sources? Have u tried disabling SNMP and see the outcome?
Yes i have disabled those switches , automatically CPU utilization got reduced.
There is another tool is also polling these devices via SNMP , once solarwinds is live it will be decommissioned
Basically the device can only support so many snmp requests at once, you need to reduce the amount that you are asking from it. Either disable some of the polling features like routing rables/vlans/neighbors or do not poll as many interfaces as you are and see how it goes. You might just add it into your Orion and then unmanage it until the other tool is decommed, then make sure that it stays within acceptable cpuloads with just the one tool.
Agree to your point, but simultaneously i need both tools to work and monitor the environment. There will be test go in the environment by keep another tool standby , so would not be able to disable the monitoring from solarwinds
I have added only the connected interfaces to the monitoring and these are high end switches which needs to have other features enabled as per the requirement.
Probably u need of to think of some workaround... if Solarwinds is still in POC stage, then u can monitor very minimum and not all parameters...
Many older tools, and even some current, aren't very polite when polling a switch. They'll read large tables in a way that causes significant CPU spikes. Here is an article on how to prevent them from doing it...
IP Simple Network Management Protocol (SNMP) Causes High CPU Utilization - Cisco
