Friday morning. Coffee. Deploy routine update.
kubectl get pods
ImagePullBackOff.
What.
Bitnami moved MongoDB images overnight.
Old location: docker.io/bitnami/mongodb
New location: docker.io/bitnamilegacy/mongodb
My Helm chart: “Hey Bitnami, where’s mongo?”
Bitnami: “We moved. Check the other repo.”
My chart: “Nobody told me.”
Zero warning. Zero redirect. Just gone.
Best part?
Happened to THOUSANDS of teams simultaneously.
Some discovered during routine restarts.
Others during actual outages.
GitHub issues flooded with “WHY IS EVERYTHING BROKEN”
Bitnami’s answer: “Oh yeah, we announced that. Didn’t you see?”
(Narrator: They did not see.)
The real lesson:
Free services change the rules.
You can’t sue Docker Hub.
You can’t demand uptime SLAs.
You just… adapt.
Smart move: Mirror critical images to YOUR registry.
Your AWS ECR. Your GCP Artifact Registry. YOUR control.
Costs $5/month. Saves your weekend.
Trust vendor promises? Sure.
But when breaking them costs YOU money and THEM nothing…
Maybe keep a backup plan.
Ever had a “wait, where’d the image go?” moment?
Same pattern as with goo.gl shutdown in 2018 — when a free service changes the rules, you either have a backup plan or you scramble.
#Kubernetes #DevOps #Bitnami