100% CPU Utilization in Cat 6509..

[hyukin.kwon]

Hi all,

I have two Cat 6509 which have two Supervisor 2 modules each with MSFC for redundancy.

They are doing HSRP for reliability..

and Image version is c6msfc-ds-mz.120-7.XE1.

and then suddenly the CPU utiliztion is 100 % for 30 minutes.

Log follows...

Jan 19 15:58:34: %STANDBY-6-STATECHANGE: Standby: 120: Vlan120 state Active -> Speak

Jan 19 15:58:44: %STANDBY-6-STATECHANGE: Standby: 120: Vlan120 state Speak -> Standby

Jan 19 16:00:17: %STANDBY-6-STATECHANGE: Standby: 37: Vlan10 state Standby -> Active

Jan 19 16:00:17: %STANDBY-6-STATECHANGE: Standby: 37: Vlan10 state Active -> Speak

Jan 19 16:00:27: %STANDBY-6-STATECHANGE: Standby: 37: Vlan10 state Speak -> Standby

Jan 19 16:00:40: %STANDBY-6-STATECHANGE: Standby: 150: Vlan150 state Standby -> Active

Jan 19 16:00:40: %STANDBY-6-STATECHANGE: Standby: 150: Vlan150 state Active -> Speak

Jan 19 16:00:50: %STANDBY-6-STATECHANGE: Standby: 150: Vlan150 state Speak -> Standby

Jan 19 16:10:24: %STANDBY-6-STATECHANGE: Standby: 41: Vlan150 state Standby -> Active

Jan 19 16:10:24: %STANDBY-6-STATECHANGE: Standby: 41: Vlan150 state Active -> Speak

Jan 19 16:10:34: %STANDBY-6-STATECHANGE: Standby: 41: Vlan150 state Speak -> Standby

the log of the two Cat 6509 are same...

Simply, the hsrp is the reason. I don.t know .

Welcome any possilbilities..

Thanks in advance..

[skarundi]

HSRP is rarely the reason. it's normally just a victim of the problem. HSRP tends to flap alot when you have a bridging loop or network congestion between the HSRP neighbors.

Read the Troubleshooting HSRP problems document. Lots of good troubleshooting tips in that doc.

URL: http://www.cisco.com/en/US/tech/tk648/tk362/technologies_tech_note09186a0080094afd.shtml#topic19

[g.shashank]

Hi,

Did u added any new server in u r network, or any new application ( dealing with multicast) just look into it. I faced the same problem and the culprit turned out to be a server sending huge amount of traffic.

A sniffer might help u out to find out( span individual vlans and analyse the traffic for suspicious traffic)

regards,

shashank

[Rodney-roberts]

also check for spantree TCNs from the L2 do a sho spantree stat (port #) ( vlan) look for the Vlan based information & statistics section and lookk for the last topology change. and keep an eye on that. the topology change initiator will tell you where the TCN came from ,its best to start at your root bridge and work down.