For clustered nodes, where datasources are deployed to the farm directory of one instance, the datasources, when viewed in the GUI are flagged as being unavailable in the second node even though the server logs from both instances indicate the datasources were successfully deployed.
The problem also occurs if the datasources are deployed to the farm directory after the instances have already been started.
Test Case 1:
- Make two copies of the 'all' config (named node1 and node2)
- Copy multiple datasources to node2/farm
- Using the service binding manager, startup both instances
./run.sh -c node1 -Djboss.service.binding.set=ports-01
./run.sh -c node2 -Djboss.service.binding.set=ports-02
- Import the resources into the repository ( the datasource will appear as unavailable in one of the nodes)
Test Case 2 :
Add new datasources to node2/farm (the Server GUI wouldn't reflect the availability of the new datasource in node1)