Aurynn Shaw

Good DevOps practice is to build your systems with failure in mind.

What you build *will* break. That's inevitable. So what you build should be built to be quickly recoverable.

You should have good disaster recovery plans.

You should have backups.

Your deployment should be automated, and use infrastructure-as-code extensively.

#DevOps #MastoAdmin

Nov 14, 2022 at 7:05pmWeb
Show threadAurynn ShawNov 14, 2022
If you do not schedule maintenance, your systems will schedule it for you.
Show threadSophie JaneNov 14, 2022
@aurynn And sometimes that’s the easiest way to get it done, speaking as someone who’s worked with a lot of decision-averse managers
Show threadAurynn ShawNov 14, 2022
@ghost_bird oof, I'm sorry to hear you've had that struggle
Show threadRich 🌈 🟥 🍁 🟥Nov 14, 2022
@aurynn guaranteed, the untested DR plans always fail.
Show threadAurynn ShawNov 14, 2022
@rich Untested plans are just hopes and dreams
Show threadLawler Hix Nov 14, 2022
@aurynn ominous
Show threadAurynn ShawNov 14, 2022
@lawlznet It comes from an older idea of factory machines that do not have scheduled maintenance will schedule it for you
Show threadLawler Hix Nov 14, 2022
@aurynn lol no worries, i got the joke. xP
Show threadAethylredNov 14, 2022

@aurynn I have reckons about automation...

okay it might almost be a thesis...

Show threadAnnamalNov 14, 2022
@aurynn and for your truly ancient legacy systems where spiders dwell, have some sort of break glass documentation for what to do when it does that thing it does once every 7 years because Terry (who knew how to soothe it) has left.
Show threadAurynn ShawNov 14, 2022
@annamal tested DR documentation is important
Show threadAnnamalNov 14, 2022
@aurynn especially in a disaster prone country :)
Show threadNicolas PettiauxNov 14, 2022
@aurynn indeed. The same for our personnal computers and smartphones
Show threadNealeNov 14, 2022
@aurynn Considering making this post my team's mission statement.
Show threadAurynn ShawNov 14, 2022
@neale It's good practice