I had a situation last night where a node left the cluster and I lost some primary shards. It rejoined, and the shards are showing up as UNASSIGNED with no documents:
If replica shard is in the cluster, elaticsearch will assign the shard as primary in a while. So there is no need to replace by youself. What GET /my-index/_recovery shows?
It had been going on for hours overnight, so I'm assuming that recovery couldn't help. I ended up deleting the indexes this morning. Thanks, though! I'll check that endpoint if it happens again and see why the replacement doesn't happen.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.