hack: Add cluster-push-*.sh scripts

[cgwalters]

I just got this working but I am feeling it's going
to be a dramatic improvement in my development workflow.

To set up, run hack/cluster-push-prep.sh once after doing
an install. Then run hack/cluster-push.sh to deploy a new
MCD. Use env WHAT=machine-config-controller hack/cluster-push.sh
to update the controller.

(Yes, these don't integrate with the Makefile yet...)

[ashcrow]

I assume this is to remind people to podman login 😄. NM, later on it occurs and this is a logging line 👍

[cgwalters]

Just to link explicitly, here is a key bit of sorcery required. For me I ran the setcap on /usr/bin/oc since kubectl is a symlink to it.

[ashcrow]

Looks good!! Only thing missing is a reference to it's use in one of the in-repo docs.

[cgwalters]

I sort of wonder though whether the real correct thing is to build a local release payload. There's some hints on that in the release payload doc, but I started this from the perpective of "automate the existing HACKING.md".

[ashcrow]

level=fatal msg="Error executing openshift-install: failed to fetch Cluster: failed to generate asset \"Cluster\": failed to create cluster: failed to apply using terraform"

openshift/installer#882 may fix this e2e issue.

/test e2e-aws

[cgwalters]

OK, I updated the HACKING.md. Can someone give this a try and report how it went?

[ashcrow]

/test e2e-aws

[jlebon]

Can someone give this a try and report how it went?

Sure, will give it a try.

[ashcrow]

/test e2e-aws

[cgwalters]

OK updated again with a few bugs fixed; there was a nasty one that took be a bit to track down that the hack/cluster-push-prep.sh had assumed an oc project openshift-machine-config-operator which I had locally. I also integrated with Makefile.

[jlebon]

OK, I'm stuck on trying to get a libvirt cluster up: openshift/installer#889.

[cgwalters]

OK, fixed one more bug. This now WFM from a clean install.

The next bug though I'm hitting is I think there's a race between me patching the images configmap and then the deployment; the deployment seems to get reset back right away. But there's still high latency between the configmap change and the deployment change.

[cgwalters]

/hold

But nope it's deploying an ancient image for some reason; debugging...

[cgwalters]

/hold cancel

Well...I dunno how that happened. Seems to work in another try. If this looks useful to people I'd say we merge, we can always do followups.

[jlebon]

Giving this a try locally!

[cgwalters]

Updated ⬆️

[jlebon]

OK, this works locally for me now. Good stuff!

/lgtm

[openshift-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: cgwalters, jlebon

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Details

Needs approval from an approver in each of these files:

• OWNERS [jlebon]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment