Create a meta-pipeline to update our EKS K8s Clusters via Pulumi

[feoh]

Description/Context

Since we are using Pulumi to manage our K8s clusters, and there is a great deal of code behind the scenes split over a few separate modules, using a pipeline to keep them all up to date makes sense.

Plan/Design

Two open questions are cluster enumeration and version tracking.

Do we try to enumerate all our clusters automagically so new ones are added with no operator intervention, or is that too risky?

Do we need to track reseource versions separately (ungh) or can we just say that all clusters must have the latest greatest code always? This is the approach I favor but I'm willing to entertain reasons why this isn't great.

[feoh]

An excellent example can be found here. This contains an example of both the meta pipeline and the actual single cluster (equivalent!) pipeline that's used to populate it.