After 11.2.1 Upgrade, ZCM agent generates excessive traffic due to errant soap checksumming

  • 7010585
  • 07-Aug-2012
  • 12-Aug-2013

Environment

Novell ZENworks Configuration Management 11.2.1

Situation

After being upgraded to 11.2.1, the ZCM Agent may repeatedly perform "Checksumming soap call with UID:" against the same UID hundreds of times per minute and never cease, if it fails over to a second primary server.
 
[DEBUG] [07/25/2012 10:09:57.436] [1892] [ZenworksWindowsService] [99] [] [SoapUtility] [] [Checksumming soap call with UID: Assignment:GetEffectiveAssignments] [] []
is seen in zmd-messages.log approximately 0.002 seconds apart (this message is normal, it is only a problem if seen often, in a short time period, and it never stops)

"Total Connections For Previous 24-hour Period" on the server's Statistics tab shows millions of connections

Resolution

This is fixed in version 11.2.2 - see KB 7010757 "ZENworks Configuration Management 11.2.2 - update information and list of fixes" which can be found at https://support.microfocus.com/kb/doc.php?id=7010757
 

Workaround: if it is not possible to upgrade to ZCM 11.2.2 at this time, in the interim, Novell has made a Patch available for testing, as part of a Monthly patch update: it can be obtained at https://download.novell.com/Download?buildid=t9vbJXG-P0g~ as "ZCM 11.2.1 Monthly Update 1 - see TID 7010464". This update should only be applied if the symptoms above are being experienced, and are causing problems.

If it is not possible to implement ZCM 11.2.1 Monthly Update 1 at this time, Novell has made a Patch available for testing, in the form of a Field Test File (FTF): it can be obtained at https://download.novell.com/Download?buildid=zKtdNWriegU~ as "ZCM 11.2.1 fix for ZCM 11.2.1 agents cause high server utilization - see TID 7010585". This Patch is recommended for all ZCM 11.2.1 zones with more than one Primary Server, that have not deployed ZCM 11.21. Monthly Update 1 or later

Workaround:  Temporarily  location rules to exclude all but one primary server for just the configuration role and not use default rule.  After the fix is applied, that should be changed back to the original settings for proper load balance and failover.