Single-node restart results in inconsistency between primary and replica shard data

Search primary shard:

Search replica shard:

Please don't post images of text. They're had to read and search.

Repeating the ?preference parameter doesn't work as you expect. Elasticsearch ignores one or other repeat, so your searches are doing different things.

1 Like

ES version: 5.4.2

I just restarted a node. After the cluster state is green.

Sometimes the statement can look up the data, sometimes it can't. I use preference parameter to find that the data is on the primary shard, but the replica shard is not.

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.