This section applies to only the Service Assurance Manager.
Escalation is a function within SAM that executes various automatic, user-defined actions at specified time intervals. Escalation, when it is used, requires care when implementing the hot-Standby failover solution. While the Standby SAM should mimic a parallel behavior in terms of moving the notifications along the levels of an escalation path, it should not allow the same actions to be executed in the Standby SAM as in the Active SAM. If it did, actions would be duplicated, such as sending an email back to back for the same reason. On the other hand, escalation in the Standby SAM cannot be completely disabled because escalation needs to work as configured after a failover occurs.
For example, a given escalation path has three levels: Level-0 (action invoked immediately), Level-1 (action invoked 1 minute after Level-0), and Level-2 (action invoked 10 minutes after Level-1). A notification occurs; it matches the path, in both the Active and Standby SAMs, and Level-0 actions are executed immediately (for example, setting a UserDefined field), on the Active SAM. A minute passes and the Level-1 actions are executed (for example, sending an email to an engineer). During the succeeding minute, the Active SAM dies and a failover occurs.
The Standby SAM did not execute Level-0 and Level-1 actions because they were disabled. But the same notification is now scheduled to invoke Level-2 actions in the Standby SAM (after the specified interval). In the meantime, the Failover Manager reconfigures the path definition so that the automatic escalation actions are enabled in the Standby SAM after the failover. When Level-2 is reached, its actions are executed (for example, sending an email to an engineering manager) by the Standby SAM. In this manner, the notification was not affected by the failover during the transition in terms of its escalation actions.
The Failover Manager periodically transfers Global Manager configurations (Notification Lists, User Profiles, Actions, Escalation Policies, and so on) from the Active Service Assurance Manager to the Standby counterpart. However, the importing of escalation definitions with certain changes, such as a policy filter, may be restricted by the system.
You may observe that the Global Console connected to the Standby SAM in the Failover setup disconnects for a brief duration of 30-seconds while taking a backup of the action script, which transfers the Escalation policy from the Active SAM to the Standby SAM. This disconnect is expected as Failover Manager runs the importConfigFromFile on the ICS_ConfigurationManager or sm_config import with the --force --replace option. The Domain Managers will be reconfigured after this step.
An escalation policy or path must be in a disabled state when a filter is changed. Likewise, it must be disabled to add or remove levels, or change level intervals in an existing path. Refer to the Service Assurance Manager Configuration Guide for detailed information about how to modify existing escalation policy and paths.
When escalation is used with the hot-Standby failover solution, the escalation definitions in the Standby SAM may be inconsistent with the ones in the Active SAM. They may be inconsistent because changes made in the Active SAM could not be reflected in the Standby SAM.
If, for example, an escalation path is disabled, changed, and re-enabled on the Active SAM, and is then automatically transferred to the Standby SAM, the automatic transfer by the Failover Manager will fail. It fails because the path is still enabled on the Standby SAM, and any changes to the path (filters or level modifications) cannot be applied. Thus, all of the corresponding escalation paths in the Standby SAM must be disabled after they are modified in the Active SAM unless they are retired.
If escalation changes are accomplished by retiring a current path, instead of disabling and enabling it, there are no issues with copying them to the Standby SAM.
Disabling an escalation policy or path cancels all scheduled actions for the existing notifications that currently match an entity.
When it is necessary to disable an escalation path in an Service Assurance deployment where the hot-Standby failover solution is in effect, perform the following:
Attach to the Active Service Assurance Manager by using the Global Manager Administration Console.
Disable the escalation path.
Make the necessary filter changes in the path, and re-enable it.
Attach to the Standby Service Assurance Manager using the Global Manager Administration Console.
Disable the escalation path.
Detach from the Standby SAM.