"master not discovered exception" after having 999 of 1000 max. shards open

Hello there,
I'm migrating my live data from our live elasticsearch server to our archive elasticsearch server. Both are single nodes. While doing this I missed to close indices on the archive server. Using curator I got the error "Validation Failed: 1: this action would add [2] total shards, but this cluster currently has [999]/[1000] maximum shards open;" somewhere while executing the task. I tried to close indices on the archive server, but I always get "master_not_discovered_exception", regardless of what I'm trying to do, e. g. quering some statistics from the archive:

root@elk-archive-01 ~ # curl -X GET "10.0.0.210:9200/_cluster/health?wait_for_status=yellow&timeout=50s&pretty"
{
  "error" : {
    "root_cause" : [
      {
        "type" : "master_not_discovered_exception",
        "reason" : null
      }
    ],
    "type" : "master_not_discovered_exception",
    "reason" : null
  },
  "status" : 503
}

In the logs the following lines are suspicous:

failing [elected-as-master ([1] nodes joined)[{elk-archive-01-n1}{QGFmDMAITRu4JNov0SwH1Q}{RbaWF3t7QKyXKWDa3fEbYw}{10.0.0.210}{10.0.0.210:9300}{dilm}{ml.machine_memory=33449811968, xpack.installed=true, ml.max_open_jobs=20} elect leader, _BECOME_MASTER_TASK_, _FINISH_ELECTION_]]: failed to commit cluster state version [19845]
org.elasticsearch.cluster.coordination.FailedToCommitClusterStateException: publication failed
[...]
Caused by: org.elasticsearch.ElasticsearchException: publication cancelled before committing: timed out after 30s

Unfortunately I don't know where to upload the whole logfile :frowning:

How can I close the indices to get the server back up again? Thanks a lot for your help!

Ok, following "Can not elect master when restarting cluster from 7.3 upgrade" fixed the problem.

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