mercredi 9 septembre 2009 10:17
I have a Windows 2003 clustering running for Exchange 2003 enterprise with 2 nodes. Recently, there has been problem move the cluster group from node 1 to 2. One of the resources will fail and this makes the mail service unavailable till i hard-reboot the servers.
The usual steps will be as follows:
1) Move all cluster groups from node 1 (active) to node 2 (passive)
2) restart node 1
3) Move all cluster groups back from node 2 to node 1
4) restart node 2
It often fails at step 2 or 4 with smtp virtual instance 1 unable to take offline (Event ID: 1004) , fails to stop service (Event ID: 1024) and fails to take offline because of timeout (Event ID: 1020)
I like to check if i can do another way of restarting the servers by:
1) shut down the node 2 (passive)
2) restart the node 1 (active)
3) start up the node 2 (passive)
Will there be any known issues arising with either Exchange or Cluster using this method.
The objective is just to free up the server resources while i am trying to figure out the cluster failover issue.