the Architecture is that there is 200+ K8s clusters (multi , single nodes, small and big resources), they are all running the deployments which is basically front + api, db inside etc...
All the clusters are on premised and spread all over the world.
The problem is within the CI/CD process I am trying to accomplish. Now I came up with the idea of deploying the helm chart during basic release of production docker image, which will trigger terraform cloud plan/apply to the self hosted runner inside every cluster , and send notification, I am talking to TF cloud via Lambda through its API, and Triggering Lambda with ECR push event. Authing to k8s APi through SA token passing to provider.
Why TF cloud? Easy and multi functional tool which can be used by Tech / Support teams, and convenient roll back mechanism.
The problem is that its not allowing me to onboard more then 10 clusters. Cuz for that I need to pay very big amount of money to use enterprise.
What do you recommend to use, change or modify in this setup
Crack the software or find an alternate vendor
Remind me of this tomorrow, I need to know better, maybe I can help
Chief, switch to Spacelift
I am a bit confused here. Your heom charts are being deployed by terraform?
going back to the point, how your cicd pipeline look like since you mentioned them?
Обсуждают сегодня