Deploying to a cluster with no downtime
I was wondering if it's possible to deploy to a cluster without taking the application down for a maintenance window. In our case we will likely have a hardware load balancer fronting the application so we could use 2 stand alone instances with the same application deployed where software deployments would go like this.
1. stop traffic to 1 instance from the LB
2. deploy the new application code to that instance and test
3. swap the instances in the LB and repeat
This of course does not take advantage of the Glassfish cluster features of config sharing and in-memory session replication. Can someone describe how you could achieve something like this using the cluster profile.