Not certain of the feasibility but wanted to open up the discussion since it seems like copying an entire shard from on index to another seems like it should be doable.
Again unsure of actual implementation but I would think it may be able to leverage the same mechanism as the shrink API?
I understand that is the current solution, but it is expensive and time consuming to do so, not to mention it can affect "live" index rate into the rest of the cluster.
I guess I'm trying to gain a deeper understanding of Elasticsearch and why this is NOT possible for indices but IS possible for shards.
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.