Does any one have any ways to remove disks from nodes that would limit down time of Elasticsearch? I would like to know what the best method is.
You can remove shards from a node using shard allocation filtering, reconfigure the node the way you need and bring shards back in.