cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
2434
Views
10
Helpful
3
Replies

High CPU usage

abarb002
Level 1
Level 1

CPU 1 is functioning at 85% workload. I've researched different ways to lower this, like CEF and MLS all which are being used currently on this layer 3 switch. Also looked into changing my log statements in my ACLs to address the IP input (biggest user under cpu processes) but I can't do that because its required by STIG rules. I just started using Solarwinds to monitor my network so I don't know how long this high usage rate has been going.

 

Question, is this normal for a 6506 with one 720 SUP? Is there something that can be done to load balance traffic? We are currently in the middle of an operation so I can't shut nor restart any network devices. Any suggestions are greatly appreciated.

 

1 Accepted Solution

Accepted Solutions

abarb002
Level 1
Level 1

At some point during the past few days there was a spike that caused the CPU to reach a load usage of 85%. I removed the switch from Solarwinds and rediscovered it and the CPU load warning went away. I doubled checked the cpu, #sh proc cpu, and everything was normal.  However the cpu history did show spikes. I will keep monitoring this and update if this occurs again.

View solution in original post

3 Replies 3

Reza Sharifi
Hall of Fame
Hall of Fame

Hi,

CPU 1 is functioning at 85% workload.

That is not normal. The high CPU is most likely because of a bug in the software version you are running. I suggest you open a ticket with Cisco and have them investigate  a possible resolution.

HTH 

 

abarb002
Level 1
Level 1

At some point during the past few days there was a spike that caused the CPU to reach a load usage of 85%. I removed the switch from Solarwinds and rediscovered it and the CPU load warning went away. I doubled checked the cpu, #sh proc cpu, and everything was normal.  However the cpu history did show spikes. I will keep monitoring this and update if this occurs again.

SNMP discovery issues like this is usually because of a bug in the switch IOS.

I have seen this in Nexus series, where SNMP causes a serious slowness in show commands. You may also want to post this on the SW forum, just in case this is an NPM related issue.

HTH

Review Cisco Networking for a $25 gift card