HN Explorer

Ask HN: How do you track what's backed up?

2 points by atomicnature ↗ HN
My team recently lost some production data - not because backups failed, but because some critical things weren’t tracked at all.

Postgres dumps -> backed up

GitLab repos -> backed up

Ghost configs -> backed up

Listmonk database -> NOT backed up

Snapshots exist, but they’re heavy and don’t make it clear whether everything important is included. Also, there's the overhead of managing them, policies, etc - which just feels like overkill. Backing up the entire OS and temp files and so on doesn't sit well with me.

Ad-hoc scripts give an illusion of safety -- but practically it’s easy to miss something (we had a bunch of scripts/cron for this stuff).

The hard part isn’t running backups -- it’s visibility: knowing what needs backup, what’s actually covered, and where it lives.

How do you handle this in your organization? What’s the hardest part of tracking backups in practice?

3 comments

[ 2.6 ms ] story [ 17.7 ms ] thread
toomuchtodo
A data inventory should be the source of truth, which should flow into backup strategy and technical implementation details. You cannot backup (or manage the lifecycle of) what you do not know you have or aren't tracking.
pestatije
backup is half the job...the other half is testing the backup (ie. get to a working environment with just the backup)
cybersecmele
It is great you implemented backups but as someone in the comments has already said make sure they are tested. Also, remember about recovery capabilities. I mainly use GitProtect.io as it covers my entire stack, allows me to oversee all the processes, works in accordance with the 3-2-1 backup rule and is automated too. So yeah, in terms of tracking backups, that tool resolves my issues as there is a central management console where you manage everything from one place.