OpenStack failed to find a real foothold.

First it was supposed to be the “AWS killer”. Nope. Then it was the “Private Cloud/VMware killer”. Nope. Then it’s last hurrah was for Telco NFV (Network Function Virtualization), but proved far too complex and unstable at scale. https://twitter.com/thecloudcastnet/status/1360971780870004742

Ask anyone how their OpenStack upgrade process went. Especially if it’s a telco running live traffic. Ask them how many one-offs and snowflakes they had to deploy to meet each vendors demands. It was an Operational nightmare. And that’s IF you were even able to get it into Prod.

I know Telcos that spent 5+ years and over a Billion dollars trying to get “free” OpenStack to work at scale and never did. I also know lots of people who lost their jobs over it.

And depending on the community for support proved untenable for a lot of companies, especially the telcos.

Competitors (like VMware) offered a turnkey alternative that was in Prod in 2 months instead of years of struggle.

It was a great idea, but just struggled to find a market

I’ve seen the details of probably every major telco OpenStack project on earth. Some quite intimately as we ripped them out for something that actually works. Others that limped along for years, never reaching anything near what was expected. They’ll all be killed off for k8s.