Open pulpbot opened 2 years ago
From: @bmbouter (bmbouter) Date: 2021-06-02T18:58:26Z
I'd like to see this improve so +1 to this effort.
I have these concerns about the proposal (although I like the idea sharing):
Can more info be shared about the problems with an LVM snapshot approach? I've always heard that's the answer, but maybe in practice it's not that simple.
From: @jlsherrill (jsherril@redhat.com) Date: 2021-06-08T19:00:30Z
Evgeni was the one that shared the issues, essentially what it sounded like was that once you have a snapshot created, write performance drops considerably.
I found a blog post that details this: https://www.percona.com/blog/2009/02/05/disaster-lvm-performance-in-snapshot-mode/
This issue has been marked 'stale' due to lack of recent activity. If there is no further activity, the issue will be closed in another 30 days. Thank you for your contribution!
Author: @jlsherrill (jsherril@redhat.com)
Redmine Issue: 8852, https://pulp.plan.io/issues/8852
It would be useful if we could:
1) perform an online backup of the pulp postgresql database, this is relatively quick 2) backup the /var/lib/pulp/ filesystem
This is nearly possible today as we could simply not run orphan cleanup. However publication files (such as repomd files for yum content), are cleaned up when a publication is deleted. It would be awesome to defer that until orphan cleanup time. This would allow /var/lib/pulp/ to be additive until orphan cleanup is called.