Device is polling but interfaces stop polling randomly in CA Performance Management (CAPM)

Products

• CA Performance Management - Usage and Administration
• DX NetOps

Issue/Introduction

A device is up and polling but most of its interfaces suddenly stop polling and they lose data. Polling is restored for the interfaces after "Update Metric Families".

A dcdebug->Discovery Logging of the device shows that the interface ifAdminStatus OID (1.3.6.1.2.1.2.2.1.7) value is set to 1 (UP):

 

SnmpResponseVariable [oid=1.3.6.1.2.1.2.2.1.7, type=INTEGER, value={1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}, isDelta=false, isList=true, error=SUCCESS, isDynamicIndex=false, indexList=[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15]]

So change detection (CD) should detect this and restore polling to the interface, but doesn't.

 

Environment

DX NetOps CAPM Release : 21.2

Cause

By default the change detection rate is 24 hours.  If CAPM detected the ifAdminStatus was down, it will stop polling the interface.  Then it will wait for 24 hours to check the ifAdminStatus again.  But we should be able to detect the status automatically.

If IsPolled=false on the interface item, the Status will be marked as Inactive.  Then Update Metric Families will put IsPolled back to true. 

This issue is caused by a bug where change detection fails to set IsPolled to TRUE.  For change detection, it only detects the external attribute changes.  If there is no change, it won’t check the internal attributes, such as IsPolled, IsFiltered, etc.  Since IsPolled is set to FALSE for whatever reason, change detection won’t set it back to TRUE unless there is an external attribute change detected.

Resolution

Fix created and delivered in DX NetOps CAPM release 22.2.8. Upgrade to 22.2.8 will resolve this problem.