Chaos Engineering is ridiculous

Hiya,

One thing that is IMHO is ridiculous: Chaos Engineering. How many engineering organizations actually need Chaos Engineering?

Let me ask it a different way. How many orgs are just getting by with minimal automation and staff, but somebody read something about Netflix Chaos Engineering and dropped what they were working on because "we gotta do that"?

Cargo culting at its finest.

I find it funny. I might be a bit ignorant on the subject, but I've just never once seen a need for Chaos Engineering. Why?

There are always higher priorities that need immediate action.

And yet I still see it randomly in folks "steps to become a DevOps engineer" or "your platform's roadmap" thought leadership posts.

No, don't do that. Stop.

At scale, I believe Chaos Engineering makes sense. For example, if you work for Netflix.

I just think I hear more folks talking about it when they're nowhere near the usage that make sense. These people aren't likely to work in places where that scale exists.

The number of orgs who need true Chaos Engineering feel to me like they're in a very small minority.

Now, if you are talking about testing business continuity or disaster recovery efforts, that’s a different beast. I don’t classify that as Chaos Engineering, but instead one of those higher priorities mentioned above.

Disaster recovery testing provides immediate business value. My team and I have done for this customers, and making sure your vital systems can handle discrete failures is critical. But that’s a far cry from Chaos Engineering.

Happy DR planning,

Matt @ Masterpoint

PS On a different note, this post dives deep into GitOps with Terraform Controller. Check it out.