How are Chassis Down Alarms generated in Spectrum and how can this be debugged further?

Products

• CA Spectrum

Issue/Introduction

We need information on how Spectrum is detecting chassis down to generate the alarms.  What are the parameters/oids read at device end that cause the alarm to be generated.

Environment

Component : Spectrum Core / SpectroSERVER

Cause

Chassis down alarms (0x10f69) are created from Chassis Transition event 0x10f8c which is generated from IHExtendedFaultDomainContactManager.  The IH stands for Inference Handler which are a piece of code that lays dormant until a specific function is called.  There are no OIDs polled from the device to create this alarm.

Resolution

If you want to troubleshoot / debug this further you can log the transition  event (0x10f8c). i.e. enable the database saving option from Event Configuration (Event Options tab,  under Logging click the box for Store Event in Historical Database).


After enabling database option for event 0x10f8c, below is the list of items to be collected when issue occurs:

1. List of events from the device to include the event type exported in CSV.

2. Enable the IHExtendedFaultDomainContactManager debugging from VNMSH using below action code

Update action=0x00010281 mh=<device condition domain model handle>
To get device condition domain model handle
show models |grep -i <devicename_model> |grep -i ENTITY-APP

3. After the problem reoccurs you can disable the debugging
Note : to disable debug below are update action for FaultDomainContactManager :
Update action=0x00010281 mh=<device condition domain model handle>

4. Output of the debugging is to the VNM.OUT, don´t forget to disable logging for event 0x10f8c to prevent excessive events from being stored in the ddmdb