Skip to content
Notifications
Clear all

Complete newbie here - where to start evaluating if we should switch IaC tools?

4 Posts
4 Users
0 Reactions
60 Views
(@baller_analytics)
Honorable Member
Joined: 5 months ago
Posts: 483
Topic starter   [#15204]

You're asking the wrong question first. "Should we switch" is a vanity metric. The real question is: what problem are you trying to solve that your current tool can't?

Start here:
* Define your failure criteria. What does your current IaC tool do poorly? Be specific. Is it state locking? Provider coverage? Module reuse? Team velocity?
* Quantify the pain. How many hours are lost to workarounds? What's the error rate in deployments? Track it for a month.
* List your non-negotiables. For example: must have deterministic plans, must support our existing cloud providers, must integrate with our CI/CD.

Then, and only then, look at tools. Terraform, Pulumi, CDK, etc. Test against your actual failure points, not their marketing slides. Migrating state is the easy part. Refactoring all your modules and retraining your team is the real cost. Prove the ROI before you write a single line of new config.


If it's not a retention curve, I don't care.


   
Quote
(@emilyw)
Reputable Member
Joined: 4 months ago
Posts: 188
 

Okay, this makes a lot of sense. You're saying we need to diagnose the actual illness before shopping for medicine. I get that.

But how do you quantify the pain if the team just grumbles "it's clunky"? Do you literally start a spreadsheet to log every hiccup? Feels a bit meta.



   
ReplyQuote
(@brian)
Reputable Member
Joined: 4 months ago
Posts: 282
 

Yeah, the spreadsheet is overkill. You'll get "data" but no insight.

Just ask "what's the last thing that blocked you for half a day?" Get concrete stories. The guy who spent Tuesday afternoon hacking a workaround for a missing resource? That's quantifiable. The "clunky" feeling is usually five of those stories a month.

If they can't name one recent, specific blocker, you don't have a tool problem. You have a team looking for shiny objects.


Trust but verify.


   
ReplyQuote
(@briana)
Reputable Member
Joined: 4 months ago
Posts: 319
 

Oh absolutely, you're spot on about concrete stories. The "half-day blocker" question is gold. But I'd add you've gotta ask "And how many times has that *specific* thing happened?" That separates a one-off quirk from a chronic issue.

From my own migration trenches, you often find the team has the same three stories on repeat, they just tell them differently each week. It's like that one missing provider resource - you'll hear about the RDS workaround, then the load balancer fix, then the storage class hack. It's all the same core problem manifesting.

So yeah, get the story. Then immediately dig for the pattern. If it's the same root cause three times, that's your quantifiable pain right there.


Backup first.


   
ReplyQuote