Snapshot repository

Elastic Stack Elasticsearch
1

hi again,
i have a problem with my snapshot repository

I'm using the data tiers in the elastic (the data from my winbeats is going from my winlogbeats to the logstash and from there to the elastic hot node, warm node and the frozen in the NFS server)

i have 2 repo's and one of them at 100% and the other is at 30%

I want to try and find a way to load balance them so it will work better

The repos are on a FTP and I'm not using the cloud version

I also tried to delete some of the data in the full repo and it and it failed every time, when i tried in the dev tools I got 502 error

If you have any idea on how i can make it more balanced I would love to get some help

2

Can you share your configuration for the repositories from your elasticsearch.yaml and the result of GET _snapshot/ on Kibana Dev Tools?

How you tried to delete? You get the 502 error immediatelly or after some time? If it is after some time this may happen as deleting snapshots can take a long time and the request will timeout.

3

the results of GET _snapshot/ gave me:

"repo": {

"type":"fs",

"uuid":"pXx....GhA"

"settings":{

"location":"/mnt/repo"

} and the same thing again but with /mnt/repo2 and a different uuid

in the elasticsearch.yaml I have path.repo: ["/mnt/repo", "/mnt/repo2"]

I tried to delete using the command: DELETE /_snapshot/my_repository/my_snapshot_name
it took some time to think and then it gave me the error

4

But with a different repository name, right? Can you share the full output? There is nothing sensitive in the response, but you can remove the uuid value if you want.

It is not possible to do this automatically, this depends on your snapshot policies configurations, if one of the repositories is being used more, this could mean that you have more indices using it or you have larger indices using it, you need to balance this in your snapshot policies based on your indices size.

Yeah, this can happen, the timeout for the request was reached before the snapshot was fully deleted, but the delete task will still run until it finishes, you can check it by listing the snapshots and validate if they were already deleted or not.

5

This is all true, but it also somewhat depends on what you mean by "the error". Please always quote the full error, including stack trace, in any questions like this. Otherwise we can only guess what it might be, and those guesses can land far of the mark sometimes.