Closed kennethgillen closed 7 years ago
Oh, I'm glad this got caught earlier rather than later!
Yep, agree with those sentiments for sure. I'd dallied adding the backup file size/age monitoring - that's still not in place, I prioritised omero-user-sessions
(unique users and total sessions), but on the server upgrades running manual dumps I noticed the du
was off.
Can confirm:
[root@ome-dundeeomero ~]# date && du -sh /backups/*
Wed 5 Jul 09:35:50 BST 2017
1.7G /backups/nightly-pg_dump_omero.dir
Code looks fine, can't actually see it in action though.
There's a mysql-backup role: https://github.com/openmicroscopy/ansible-role-mysql-backup If you create something similar for postgres it'll be easier to deploy on other servers.
@manics - if you're happy with https://github.com/openmicroscopy/infrastructure/pull/279#issuecomment-313039103 showing it works, care to merge, or 👍 for merge?
Agreed that that looks like a good solution, but additionally :+1: for this being fixed and getting it in ASAP.
An update to address the fact that the nightly backup was failing, though testing at the shell was successful.
Check mode against
ome-dundeeomero
: ansible_run-backup-script-CHECK-ome-dundeeomero_20170704-120625Z.txtLive run against
ome-dundeeomero
: ansible_run-backup-script-UPDATE-ome-dundeeomero_20170704-120909Z.txtWithout this PR
With this PR