Burn down and recreate k8s-infra cluster

[spiffxp]

Last we talked, the original manually created cluster had not yet been burned down. We want to burn it down, and use a script to stand up a cluster.

umbrella ref: #152

/assign @thockin @justinsb @hh
/wg k8s-infra

TBD: what... is actually being stood back up?

• accounts / iam policies?
• nodes
• RBAC roles/policies
• infra running on the cluster

[spiffxp]

also certificates

[spiffxp]

Fair to start with cluster singular before figuring out how to manage N clusters

First place to start is the script itself. There are many knobs in the UI, how do we set them up in the script. Interested folks may use a test project and we'll iterate on turning up cluster there

[ameukam]

/assign

[cblecker]

When we talked about this 6 months ago, I had thought we could use terraform for this: https://github.com/cblecker/k8s-tf-cluster

Don't know if we want to go that way, or if we want bash+gcloud, or something else.

[spiffxp]

We may try to get together to hack on this at KubeCon EU (contributor summit?)

[ameukam]

+100 to do this. we'll need someone that access to the GCP project that owns kubernetes.io.

[spiffxp]

relevant PR's:

• Add k8s-infra-cluster-operators-test google group. #276
• Add a script to recreate the "main" GCP project #266
• Clean up GCP k8s-infra scripts #279

[spiffxp]

Use results of #250 to create a new cluster

Burn down / migrate contents of development and development2 clusters to this new cluster

Use this cluster to confirm per-namespace billing

[spiffxp]

Our goal is to move away from bash toward terraform.

Does terraform support everything we need? Ran into a beta GKE feature that wasn't yet supported by the terraform provider. We could probably tap the right people on the shoulder to motivate which features get prioritized.

Would we be able to drop down to gcloud if we had to? Yes, this is something we can do if necessary.

Opting to just go straight to terraform

[spiffxp]

Awaiting a directory in k8s.io from @cblecker that contains terraform prototype

[cblecker]

Opened #306

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

[cblecker]

/remove-lifecycle stale

[thockin]

We now have a cluster that is born of terraform. Mission accomplished (enough to close this)