According to this technote, failover only occurs when the user is “out” of a database and then re-clicks on the database causing it to fail over.
http://www-1.ibm.com/support/docview.wss?uid=swg21086484
Has anyone found this to be un-true? Last week one of our primary servers went down and my mail file failed over with no action.
Is there a more predictable way that anyone has found to trigger failover?
Subject: Cluster Failover
You should see a dialog that states the server is not available and would you like the client to try and connect to a cluster partner instead.
Did you not see this ?
Subject: RE: Cluster Failover
check the file cluster.ncf if it contains the correct entries, also create a replication connection between the two servers in addition to the cluster replication.In my situation, when the user clicks on a mail 2-3 minutes ago, he gets the message to connect to another server and that’s all.