We have now switched over to use our standby CDM. The response
times on this device are much better.
We have attempted the following to restore normal GUI response times on the primary CDM
1) Reload the device
2) Perform full database maintenance
Neither of these two actions has improved the performance of the primary CDM
We have not recently made any changes to the CDM - the responses have
slowly deteriorated over a period of time. Cisco recommend you can support
up to 1800 CEs on a CE-7305 - we are currently running ~1200 devices, well
below that limit.
There errors messages we see in the CDM syslog.txt relate to all managed
CEs going into an "off-line" and then "on-line" status repeatedly.
We also see a lot of these:
Jun 27 04:32:54 SFCE10 java: %CE-CMS-4-700002: cdm(RpcWorker-31):
java.net.SocketException: Broken pipe: java.net.SocketException: Broken
pipe at java.net.SocketOutputStream.socketWrite0(Native Method) at
On the CEs we get this:
Jun 27 13:15:47 C1012017 java: %CE-CMS-4-716058: ce(DataFeedPoll): Cannot
get updates from 188.8.131.52, because of RpcException:Rpc timeout (300.0s)
waiting for response
Jun 27 13:15:48 C1012017 java: %CE-CMS-4-716057: ce(DataFeedPoll): Cannot
get updates from 184.108.40.206 unicorn.RpcException: This operation is not
available on standby CDM
When a similar problem occurred last year, a developer from Cisco logged on
to the CDM using an engineering logon and performed low level SQL database
maintenance. This resolved the issues.
Do you have ny ideas?