Comware

We have been experiencing high cpu usage on 3 HPE 5130-48G-PoE IRF stacks the last few days.

The switches were recently upgraded to 10GB uplinks to the TOR which is an HP A5500-24G EI.

Since that time we have been experiencing issues with CPU spikes on the master switrches within the IRF Stacks - the top offending processes are drvr, psed, llpd and snmpd.   I have reduced the number of items within our zabbix server being monitored.

When trying to backup the devices or synchronize to our iMC server the CPU on the master spikes and the switches.  They have an unknown status within iMC. 

Suggestions as to where to look?

Switches are running version Comware 7.1.070, Release 3208P03.  iMC is version 7.3 (E0506)

Thanks

Hi,

To start with upgrade the code on HPE 5130 to the latest release R3208P12 and test.

Hi,

Do you have more than one NMS server monitoring his IRF stack, if yes, you may restrict it to single one so that the snmp process comes down, At he same time, psed process is related to PoE, if there are no PoE devices, disabling the poe of hte ports.