We’re trying to get Spinnaker to work with a simple Deploy (Manifest) pipeline on EKS. When we talk about a new deploy and point it at a server, is it trying to create a new EKS cluster, and how would we be able to tell, since we’re not getting errors in either AWS or (that I can see) clouddriver?
I’ve been looking through the spinnaker documentation and am having a hard time finding out where you deploy to a new cluster that you create in Spinnaker (which our install doesn’t seem to want to do, despite having an AWS account with sufficient permissions), or to an already-existing cluster (which it doesn’t seem to want to do, because we keep getting “waiting for manifest to stabilize” until the deploy task dies (I created another thread for this particular problem).
Specifically about accounts: I know it can’t create a new cluster with our kubernetes account defined in .kube/config and exposed to Spinnaker via .hal/config, and the names of the aws account/kubernetes account shouldn’t have an impact on each other, but I’m having trouble nailing down not only where the problem is, but what behavior I should be expecting.
Is there any documentation you could point us to that goes into this specific set of concepts in detail?
Thank you in advance.