Applying the Service Pack or Hotfix to Active Cluster
When updating Active Cluster, it is critical that clusterid remain unchanged. Changing it can adversely affect cached data.
These instructions are guidelines developed from testing with a two node Active Cluster environment.
To upgrade with scheduled system downtime
1. Determine and note the clusterid, so that it can remain unchanged.
2. Make sure that the cluster is in sync. That is, make sure both nodes are in the cluster and are functional.
3. Disable any cache refreshes on both of the nodes.
4. Configure the load balancer to stop sending traffic to node 1.
Shutting down node 1 would cause in-flight requests to fail.
5. Remove node 1 from the cluster.
6. If your TDV products are running, stop them, including all processes and databases used as repositories or caches.
7. Apply the service pack or hot fix to node1.
8. Start up node 1, but do not re-join the cluster.
9. Configure the load balancer to send traffic to node 1 instead of node 2.
10. Remove node 2 from the cluster.
11. Apply the service pack or hot fix to node 2.
12. Join node 1 and node 2 to the cluster.
13. Enable cache refreshes on both of the nodes.
14. Verify the cluster status.