-
Notifications
You must be signed in to change notification settings - Fork 4.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Instantiate endpoint for deploymentconfigs #9349
Instantiate endpoint for deploymentconfigs #9349
Conversation
@openshift/api-review |
@@ -277,7 +277,7 @@ func GetBootstrapClusterRoles() []authorizationapi.ClusterRole { | |||
|
|||
authorizationapi.NewRule(read...).Groups(buildGroup).Resources("builds", "buildconfigs", "buildconfigs/webhooks").RuleOrDie(), | |||
authorizationapi.NewRule(read...).Groups(buildGroup).Resources("builds/log").RuleOrDie(), | |||
|
|||
// Why the View role has generatedeploymentconfigs, deploymentconfigrollbacks ? |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
oversight. If they aren't needed, remove them.
Fixed, I had to switch from POST to PUT... |
So I've got this working fine but the trigger controller needs image stream caches if it's going to use the endpoint and process the config without the need for the image change controller. I will work on a separate pull that adds those caches. |
|
||
// Tag references are already validated | ||
name, tag, _ := imageapi.SplitImageStreamTag(params.From.Name) | ||
istag, err := r.istn.ImageStreamTags(params.From.Namespace).Get(name, tag) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@smarterclayton @miminar @soltysh is there any reason not to use the image stream tag client instead of the image stream client?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I don't know of any, unless you already work with image stream client on some other place.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ok, switched back to an image stream client.
[test] |
@openshift/api-review I want to move forward with this. Can I get a review for the API? |
// Manual should always try to force a new deployment to run. This is meant to be used | ||
// by all clients. | ||
Manual bool | ||
} |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'd probably prefer "ForceDeployment" - "If true, a deployment is always created by incrementing the generation ..."
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
What generation has to do with this? We increment generation on any spec change (and on latestVersion changes)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yes. This is forcing a generation / latestVersion change, which is why we want to explicitly state that in the godoc and have the name reflect the change. manual doesn't mean anything and we try not to use it.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Regarding the name, how about Force
(that's what I was using before switching to Manual
) ? In a similar fashion to not making Name
(and any field name for that matter) verbose since they are namescoped by DeploymentRequest
.
API approved. |
Still can't get past protobuf generation - @smarterclayton do I need to use the same protoc version as the one used in Jenkins? Are there any other pitfalls I need to be aware of for generating protobuf (used to work in the past, not sure why this time is a special one) ? |
Run: OS_BUILD_ENV_PRESERVE=. hack/env hack/update-generated-protobuf.sh On Tue, Sep 27, 2016 at 6:53 PM, Michail Kargakis notifications@github.com
|
This was a bad idea anyway and now we can provide users with 'oc rollout latest' which will manually do the resolving for them.
Evaluated for origin test up to 2f86bfa |
continuous-integration/openshift-jenkins/test SUCCESS (https://ci.openshift.redhat.com/jenkins/job/test_pr_origin/9498/) |
@mfojtik @smarterclayton any final comments |
I will merge this later today if there are no other comments, need to leave room for merging the Conditions PR until Wednesday (devcut) |
LGTM |
[merge] |
Evaluated for origin merge up to 2f86bfa |
continuous-integration/openshift-jenkins/merge SUCCESS (https://ci.openshift.redhat.com/jenkins/job/test_pr_origin/9612/) (Image: devenv-rhel7_5128) |
Add an instantiate endpoint for processing deployment configs and plug it in the trigger controller. The trigger controller has been reconciling deployment configs on image stream updates since #9485 meaning the image change controller is obsolete now.
Partially fixes #6980.
Fixes #4359
@mfojtik @smarterclayton @ironcladlou