Operator restructure#8
Conversation
21832e9 to
53f0343
Compare
|
@enj fixed the roles, and disabled authn/authz delegation in the default config. I also reverted my changes to the boilerplate and used controller.New like you suggested.. |
| - apiGroups: | ||
| - "" | ||
| resources: | ||
| - pods |
There was a problem hiding this comment.
Are all these required because of the controller command's default behavior?
|
Going to tag because I have a feeling that changing #8 (comment) will require messing with the controller command bits. /lgtm |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
8 similar comments
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
| func NewMachineApproverOperatorCommand() *cobra.Command { | ||
| cmd := &cobra.Command{ | ||
| Use: "machine-approver", | ||
| Short: "OpenShift osin OAuth server operator", |
There was a problem hiding this comment.
Looks like you didn't fix this.
|
/retest Please review the full test history for this PR and help us cut down flakes. |
8 similar comments
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/retest |
|
/retest Please review the full test history for this PR and help us cut down flakes. |
|
/lgtm |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: enj, mrogers950 The full list of commands accepted by this bot can be found here. The pull request process is described here DetailsNeeds approval from an approver in each of these files:
Approvers can indicate their approval by writing |
|
Since this merged I've seen two failures where the machine-approver-controller is crash looping: https://openshift-gce-devel.appspot.com/build/origin-ci-test/logs/release-openshift-origin-installer-e2e-aws-4.0/2708 which I had never seen prior to to this. |
|
I would have liked to see a lot more green runs before this merged, given that it was a clean controller. It looks like you got exactly 1 pass after failing 29 times in a row - that alone is an eyebrow raiser. In general, whenever we rewrite / refactor a controller, please ensure you see multiple green e2e-aws runs before merging. If I see continued flakes in the next few days, and it looks like this controller is at fault, I will be reverting this (unless you can point to a different cause). |
| @@ -0,0 +1,131 @@ | |||
| #!/bin/bash | |||
There was a problem hiding this comment.
Why do you have a build-rpms script?
| @@ -0,0 +1,47 @@ | |||
| # junitreport | |||
There was a problem hiding this comment.
You are no longer supposed to vendor this
| @@ -0,0 +1,28 @@ | |||
| package operator | |||
There was a problem hiding this comment.
Why do you have a package called boilerplate?
There was a problem hiding this comment.
It is my personal variation on the controller / operator sync loop. It is used today by the following operators:
- Alfred
- Console
- Authentication
- Here
My goal was to add it to library-go once I felt it was "good enough."
| FROM registry.svc.ci.openshift.org/openshift/release:golang-1.10 AS builder | ||
| COPY . /go/src/github.com/openshift/cluster-machine-appover | ||
| RUN cd /go/src/github.com/openshift/cluster-machine-appover && go build -o machine-approver . | ||
| # |
There was a problem hiding this comment.
This less consistent with our default pattern than the previous (which was also not correct)
See https://github.com/openshift/cluster-version-operator/blob/master/Dockerfile for the standard format, you should only have to change the RUN line in the builder and insert your custom steps after the COPY line in the runtime image.
After the restart it looks like you are renewing your lease incredibly often: Please fix that (and fix the library-go default if that's at fault). This controller doesn't even need really need a lock. |
I still do not understand the
Yeah, |
|
Ah nvm, this runs with |
Use the cluster-osin-operator repo layout and boilerplate (with minor modifications).
@openshift/sig-auth