Issue Summary

When triggering a site reindex on a clustered instance with a subset of options selected, the job fails to propagate across nodes, and the job hangs with no progress once the initial node is reindexed.

Steps to Reproduce

1. Run a clustered confluence instance with 2 or more nodes
2. Trigger a site reindex via the UI, selecting any 1-2 of the options (not all 3)

Expected Results

Site is reindexed successfully, and the updated index is propagated across the nodes in the cluster

Actual Results

The node that the job is triggered on is reindexed successfully, but the job hangs when waiting to propagate the changes across nodes. The logs don't show any job being sent to the other nodes. This job hangs forever until a node goes down and the cleanup job runs.

Workaround

use the "resetjob" endpoint to reset the status shown in the UI, however this will not propagate the changes, only remove the error from the UI so it's usable again.