How Can Alarms from Some Probes on a Device with a Maintenance Mode Schedule Fail to Be Excluded During the Configured Maintenance Window?

Document ID : KB000120644
Last Modified Date : 08/11/2018
Show Technical Document Details
Introduction:
Several servers were put into maintenance_mode the night before, but alarms from the jvm_monitor probe were not blocked when these servers during the maintenance windows. 
Question:
Why were alarms from some probes on a device not excluded during the active maintenance window configured for the device?
Environment:
Any supported UIM release through 9.02
Answer:
When devices are selected from the UMP USM portlet to be added to a configured maintenance schedule, the MAINTENANCE_SCHEDULE_MEMBERS table is updated with the cs_id of the device added to the schedule.

When probes generate alarms on a robot, the alarm message has a dev_id field associated with it.  If the dev_id of the alarm is not correlated to one of the cs_id entries in the MAINTENANCE_SCHEDULE_MEMBERS table for any of the active schedules, then the maintenance_mode probe will not consider this device as being in maintenance and the nas will not exclude these alarms during the active maintenance windows.

This typically occurs with probes configured to monitor remote devices (like the net_connect, rsp, and jvm_monitor probes).  The alarms raised by these probes contain the dev_id of the monitored device and if the discovery_server does not correlated the monitored device with an existing device in the CM_COMPUTER_SYSTEM table which corresponds to a device that has been put into maintenance, then alarms raised by these probes will not be excluded during the active maintenance window.