This section describes how to set the pending timeout and the failback properties for the primary companion’s cluster resource. If you are configuring a symmetric setup, you must set the properties for both companions.
When the Microsoft Cluster Server (MSCS) takes the cluster resource for the primary companion online or offline, it allows for a certain amount of time to perform its processing before assuming that the operation will not complete. By default, this amount of time is 180 seconds (3 minutes). This value is known as the “pending timeout,” and can be set for each resource in the MSCS cluster.
For the Sybase Companion Server resource, the pending timeout period must be long enough to start the Adaptive Server, run recovery on its databases, and possibly execute sp_companion resume. For companions that have large databases, it is likely that this processing will take more than 180 seconds, and you should set the pending timeout property to a higher number.
If you are repairing or restarting the primary node after a failover, MSCS automatically fails back to the primary node as soon as the primary node comes back up unless the MSCS group containing the Sybase Companion Server resource is set to not automatically fail back.
To configure these properties:
Start the Windows Failover Cluster Manager (as a shortcut, press and release the Windows key, type “failover” to bring up the Failover Cluster Manager icon, then press ENTER).
Locate the new Sybase Companion server in the Services and Applications tree. The companion server is named primary_server_name_GRP.
Click on companion server name to view the Summary of primary_server_name_GRP display.
In the center panel under Other Resources, right-click the Sybase Companion server and select Properties.
Highlight the Policies tab of the Properties panel.
Change the Pending Timeout property to a value that is about two minutes larger than the longest time the server takes to recover.
Click OK.
