NLS Cluster backups issue

[mcapra]

You could also re-run the curator command to force a snapshot attempt:

Code: Select all

curator debug snapshot --repository my_repository indices --older-than 1 --time-unit days --timestring %Y.%m.%d

At any rate, let us know if the fresh repository doesn't fix things. We may need to remove the corrupt snapshots/indices by hand.

[gsl_ops_practice]

I created a new repository and the backup completed successfully, so at least that's working.

I would like to figure out which old backup is corrupt and fix/remove it, so that we restore our ability to restore indicies older than 120 days. Can you please advise how we would need to proceed?

Thanks,
Alex

[mcapra]

A remote assistance session might be best to tackle that particular problem. We can move this to an email ticket and set that up if you would like. Assuming your sending address has been whitelisted by our sales department ([email protected]), feel free to submit a ticket to [email protected] with reference to this thread.

Otherwise, can you share the output of this replacing old_repository with the old/broken repository's name:

Code: Select all

curl -XGET 'http://localhost:9200/_snapshot/old_repository/_all'

[gsl_ops_practice]

Hi,

Looking at our timelines with expanding the storage in order to do that, it is very likely I'll be reaching out for a remote support session beginning of January. I'm not able to run that command right now on the entire repository.

[mcapra]

Not a problem We will keep this thread open in the interim.