Search not working after adding a second node to a Connections 5.5 cluster

connections cnx

In the past weeks I had several Connections environments which had big problems with the search indexing. Only environments where I added a second node after the installation are affected. The environments where the Installation Manager setup installed both nodes on the same time are working as requested.

The search index on the second node will not be updated. You will find error messages like this one:

SCHD0134I: Scheduler LotusConnectionsScheduler (scheduler/search) has lost its lease and will no longer run tasks on this application server.  Tasks will now run on server Cell01\Node01\InfraCluster_server1.

A ‘SearchService.listIndexingNodes()’ wsadmin call will only list the first search node.

To fix this issue you need to be sure you have created your second search node based on those steps:

  • Create/configure second node as described in the documentation
  • Copy search conversion tools and customize WebSphere variables and environment entries like described here
  • Creating the search work managers listed here

After those steps you need to configure two more settings:

  • Create a work manager called ‘wm/search’ for the second node using ISC (Resources – Asynchronous beans – Work managers). Be sure that the scope is set to the second node
  • Enable bean service on startup for the second node search application server (Medium deployment: InfraCluster_server2) using ISC (Application servers -InfraCluster_server2 – Container – Startup beans service)

Finally you need to copy your local search directory from the first to the second node and restart your environment.

Leave a Reply

Your email address will not be published. Required fields are marked *