IP SLA measurement using Cisco SAA

[josephn]

Hi,

I have problem with the source or shadow router which was used as the source router for the IP SLA measurement. The test and oncfiguration being setup through an exernal application called Concord eHealth.

I had noticed error messages tied to CPU HOG on the log of the shadow router. There is no data being polled by the eHealth and currently it is reporting an error during the poll and this had greatly impact the SLA reporting.

Have anybody encountered this issue before ?

I'm using Cisco 3620 and IOS version 12.3(12a).

Thanks.

Joseph

[David Stanford]

What is the specific CPUHOG message? Is there an oid specified if it's related to SNMP?

[josephn]

Hi Davistan,

I'm seeing the following error messages on the log of the router:

%SNMP-3-CPUHOG: Processing GetNext of rttMonStatsTotalsInitiations.2018128175.0

%SNMP-3-CPUHOG: Processing GetNext of rttMonStatsTotalsInitiations.422568533.0

%SNMP-3-CPUHOG: Processing GetNext of rttMonStatsTotalsInitiations.152459434.0

%SYS-3-CPUHOG: Task ran for 2208 msec (2/1), process = SNMP ENGINE, PC = 60A41988.

%SNMP-3-CPUHOG: Processing GetNext rttMonStatsTotalsInitiations.754241940.0

%SYS-3-CPUHOG: Task ran for 2116 msec (4/3), process = SNMP ENGINE, PC = 60A41988.

Thanks.

Joseph

[David Stanford]

How many probes do you have configured on the device? Are you polling with snmp bulk gets?

This looks similar to a DDTS I have seen - CSCse19414

[josephn]

Hi Davistan,

There are about 361 entries configured on this router and out of this, we have about 45 probes which are active.

[David Stanford]

Thats not too bad, shouldn't be too many which would cause these CPUHOG messages. I would suggest opening a TAC SR to see if this may be another DDTS causing this.