Snapshot backup to AWS s3


Could I create a snapshot backup on AWS s3 from some specific index pattern and later create another one without saving the index that already exist on the previous snapshot? I know it could be done by day just considering the current date on the index pattern, but could it be done by hour for example? (elasticsearch index is assigned by day)

Sorry if the explanation is a bit confusing. I would appreciate any help.

It's not really clear what you're asking. Could you spell it out in more detail, maybe with an example?

Perhaps you are asking whether snapshots are incremental on an index-by-index basis? If so, yes, they are, you don't need to include every index in every snapshot to get the benefits of incremental snapshots. But it might be confusing to keep track of which snapshots have which indices in them.

Thanks for the reply David. Sorry for the messy topic, I'm new to Elasticsearch so I can't explain myself properly. So if i understood you correctly, if snapshots are incremental only new segments would be copied when I create a snapshot because the others already have a backup, right?

Also, what happen if I make a new snapshot request and, at the same time, some new files are indexed on elasticsearch? Would only the existing segments before I make the request be copied? or maybe, while the snapshot is being created, the files indexed after the request would be copied?

Exactly right, yes.

Yes, that's right.

1 Like

Thank David for all your assistance! I really appreciate your help.

1 Like

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.