Hi,
Would appreciate any pointers by experts reading this list. We have on our SolrCloud 4.6.1 cluster, it has stopped updating clusterstate.json in Zookeeper after restarted one node. Seems like the overseer process has stopped, because we cannot see a /overseer_elect/leader file in zookeeper. Because the clusterstate.json is not updating, active leader in clusterstate.json points to a dead node, hence no index update can be performed on the cluster. Is there a way to force a restart of the overseer process, or force a re-election? Maybe some special logs we can switch on to help us troubleshoot this issue? This is production deployment so we would like to avoid a full stop and restart of all nodes if possible. Any insights greatly appreciated. Regards, Edward
