Can an auto-operator be configured to close a disk alarm based on the existence of the same alarm in another node of a clustered environment?

[Fernando Galbier]

The scenario is a cluster monitored by cdm and cluster probes properly configured.

• Node A is active and cdm raises an "alarm A" for a shared disk X:, assigned to the robot node_a
• The clustered Resource Group (and the shared disk X:) moves from Node A to Node B
• A new alarm, "alarm B" is generated to the same disk X:, now assigned to the robot node_b
• The previous alarm A, assigned to robot_a is still active and it is cleared only when the disk X: falls below the threshold AND the Resource Group moves back to Node A.

 

My intention is to close "alarm A" when "alarm B" is raised. Both of them are asserted to the same Source <cluster_name>.<resource_group_name>.

 

This is Alarm A:

 

This is Alarm B:

Any idea?

 

Thanks,

Fernando

[Gene Howard]

This would have to be done with a NAS Lua script.

When alarm B came in you would have to check all alarms to see if alarm A existed then close it.

Depending on your environment and number of open alarms this could take some time.

As nas is asynchronous this could cause an issue with processing alarms.