TEPS eWAS consumes high CPU.

Technote (troubleshooting)


Problem(Abstract)

eWas server using all available CPU.

Symptom

SystemOut.log shows the following error:

=======
[7/5/11 8:23:30:845 CDT] 00000013 SystemOut O
(4e131052.2eca2640-(null)OMNI_CSPREMA02.updater:CEVService,0,"CEVService
.com.ibm.TEPS.CTCEV.connectors.CEVOmnibusConnector.CEVOmnibusUpdater.run
()") Failed to flush IDUC connection. The connection to the ObjectServer
is unstable.
[7/5/11 8:23:30:845 CDT] 00000013 SystemOut O
(4e131052.2eca2640-(null)OMNI_CSPREMA02.updater:CEVService,0,"CEVService
.com.ibm.TEPS.CTCEV.connectors.CEVOmnibusConnector.CEVOmnibusUpdater.run
()") An exception was thrown when flushing IDUC data.
[7/5/11 8:23:30:845 CDT] 00000013 SystemOut O
(4e131052.2eca2640-(null)OMNI_CSPREMA02.updater:CEVService,0,"CEVService
.com.ibm.TEPS.CTCEV.connectors.CEVOmnibusConnector.CEVOmnibusUpdater.run
()")
com.ibm.TEPS.CTCEV.connectors.micromuse.events.OmnibusEventConnectorExce
ption: An exception was thrown when flushing IDUC data.
at
com.ibm.TEPS.CTCEV.connectors.micromuse.events.impl.OmnibusEventConnecto
rImpl.flushIducData(OmnibusEventConnectorImpl.java:255)
at
com.ibm.TEPS.CTCEV.connectors.CEVOmnibusConnector$CEVOmnibusUpdater.run(
CEVOmnibusConnector.java:1188)
at java.lang.Thread.run(Thread.java:810)
Caused by: com.micromuse.omnibus.niduc.NIducException: IDUCvX: Failed to
flush iduc data. [JZ0T4: Write operation timed out. Timeout in
milliseconds: 0.]
at
com.micromuse.omnibus.niduc.NIducClient.osFlushData(NIducClient.java:865
)
at
com.ibm.TEPS.CTCEV.connectors.micromuse.events.impl.OmnibusEventConnecto
rImpl.flushIducData(OmnibusEventConnectorImpl.java:247)
... 2 more
=======



Resolving the problem

To fix this issue, you can increase the "Time interval (in minutes) for polling event

system".

This can be done either via 'itmcmd manage' or should be able to be done
using 'itmcmd config -A cq'.


1) Using the Manage Tivoli Enterprise Monitoring Services (MTEMS)
console to reconfigure the TEPS.

2) Accept the defaults until you reach the Common Event Console
Configuration. Select the OMNIbus Connector tab. Change the value of
"Time interval (in minutes) for polling event
system" from 1 to 60.

3) Finish the configuration and recycle the TEPS.

If you do not see the Java process spike the CPU again, the value can be
lowered to 20-30.

If you do use 'itmcmd config -A cq', during the process there should be
an option to configure the Common Event Console.

This is what the documentation says on this variable:

The number of minutes between each poll of the event system for new or
changed events. The Tivoli Netcool/OMNIbus ObjectServer automatically
sends new or changed events to the common event console as they become
available. Therefore, the primary purpose of this checking is to ensure
that the server and the connection to the server are functioning
properly.

P.S.: The Tivoli Netcool/OMNIbus ObjectServer automatically
sends new or changed events to the common event console as they become
available. The Omnibus events will appear on the TEP console instantly
regardless of this setting.

The primary purpose of this checking is to ensure that the server and
the connection to the server are functioning properly.


Rate this page:

(0 users)Average rating

Add comments

Document information


More support for:

Tivoli Monitoring

Software version:

6.2.1, 6.2.2

Operating system(s):

All Platforms

Reference #:

1505402

Modified date:

2013-08-14

Translate my page

Machine Translation

Content navigation