I see a behavior in elasticsearch that all the primary shards (we have 5 shards per index) added to only one of the three nodes in cluster all the time. Why is this happening?
Currently we have replicas disabled. When we add replica to that index, then the replicas are distributed among other two nodes.
Even before the cluster.routing was there this issue raised.
But I added that setting for rolling restarting of nodes yesterday.
But this issue of all primary shards into one node is happening for last one week.
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.