We are wanting to use Elastic Search Snapshots, but we are trying to keep costs down. It appears from what I am reading that the Snapshot feature wants to backup all indices that match a pattern, however we are only wanting to backup the indice just before it is deleted (so save on disk space used). Is there a way to accomplish this?
I think that if you are using ILM, there's this policy which might help.
But what triggers the Delete operation on your cluster?