Elastic managing snapshots

Elastic Stack Elasticsearch
1

Hi

When I am doing regular snapshots, let's say once or twice per day. I have understood that first snapshot that I take, it makes full backup, and after that all the backups are incremental, that are based on full snapshot.

Ok, but now antoher question pops up. Is it necessary to make full backups also from time to time..?
Or doing incremental snapshots, for years, is enough..?

Is there a command that deletes all snapshots from specific repository..? (let's say I am doing snapshots twice per day, so 14 snapshots within a week, and I wan't to make full backup from scratch, after that I would like to delete the oldest snapshots)

And final question. Am I right that when doing snapshots as follows: snap_1, snap_2, snap_3, then basically snap_1 and snap_2 can be deleted..? (so that all the data is included in the snap_3)
(Ofcourse I know it is not wise to leve just 1 backup behind, but just wanted to know if I have understood correctly)

2

Personally, I would create a weekly/monthly repo and then take snapshots for that week/month. Then you can restore to that specific point a little simpler.

A delete should take wildcards.

Yes. It'll likely keep around most of the data in snap_1 though, but we abstract it out so all you see is snap_3.

3

WIldcards actually are not working for me. Elasticsearch version 1.7.2

GET /_snapshot/paetlive_idx_data/ss_20170214_135*

{
"error": "InvalidSnapshotNameException[[paetlive_idx_data:ss_20170214_135*] Invalid snapshot name [ss_20170214_135*], must not contain the following characters [\, /, *, ?, ", <, >, |, , ,]]",
"status": 400
}

repository paetlive_idx_data has snapshot "ss_20170214_1350"

So, this works:
GET /_snapshot/paetlive_idx_data/ss_20170214_1350

4

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.