-
Notifications
You must be signed in to change notification settings - Fork 1.9k
modules/understanding-upgrade-channels: Recommend clearing channel #22252
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
Conversation
9afff16 to
635018c
Compare
|
/assign @kalexand-rh |
kalexand-rh
left a comment
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 have a question and some suggestions.
eb93d29 to
e8552a5
Compare
|
@jianlinliu, will you PTAL? I think this change should go back to 4.2. Do you agree? |
|
LGTM. @jiajliu Can you help review this from your side? |
|
There is a note in official doc https://docs.openshift.com/container-platform/4.4/updating/updating-cluster-between-minor.html. But from web console, we don't provide null channel in "Channel" option currently. So my suggestion here is that, we'd better to recommend users to clear channel trough CLI and provide the command to clear channel. |
|
@LalatenduMohanty, @wking, do you have the command to clear the upgrade channel? |
|
This is only true in CI since openshift/release#8631. User installs should have default channels, see openshift/installer#3848. |
|
Is this blocked on docs with an |
|
I've floated openshift/oc#576 for convenient, |
|
Issues go stale after 90d of inactivity. Mark the issue as fresh by commenting If this issue is safe to close now please do so with /lifecycle stale |
|
/lifecycle frozen Still trying to get review on openshift/oc#576. Now that 4.7 is in feature freeze, I'll push again once 4.7 forks off master and 4.8 opens up for new features. |
|
@wking is this still valid? |
Yeah, we still want to recommend clearing the channel as the way to get the CVO to stop trying to poll the configured |
e8552a5 to
bf742cc
Compare
|
✔️ Deploy Preview for osdocs ready! 🔨 Explore the source changes: 4244510 🔍 Inspect the deploy log: https://app.netlify.com/sites/osdocs/deploys/6100461cd998620007195424 😎 Browse the preview: https://deploy-preview-22252--osdocs.netlify.app/openshift-enterprise/latest/updating/updating-cluster-between-minor |
|
Rebased onto master with e8552a5732 -> bf742cc4e5. |
bf742cc to
e494b86
Compare
|
I've pushed bf742cc4e -> e494b8661, rebasing onto master and updating to mention |
e494b86 to
a0f4278
Compare
|
And I've pushed e494b8661 -> a0f42786e to update the:
language mentioned earlier, and instead mention |
|
LGTM. |
|
@wking @LalatenduMohanty Since we introduce |
LalatenduMohanty
left a comment
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.
/lgtm
OTA-416 is our tracker for this change, if that's useful for QE to hang some test-cases on. |
QE will follow up asap. |
jeana-redhat
left a comment
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.
Just a couple minor suggestions.
a0f4278 to
4085e6d
Compare
|
New changes are detected. LGTM label has been removed. |
I would prefer to configure this by clearing the 'upstream' setting, which seems more intuitive for "there is no upstream" to me. But sadly, it seems that the CVO has been falling back to a default URI when the ClusterVersion upstream is empty since way back [1,2], and that this behavior is enshrined in the API [3]. Although the channel docs also talk about defaults [4], the only channel defaulting in the CVO is when creating a ClusterVersion object after the in-cluster copy was (accidentally?) deleted [5]. So maybe we could talk folks into adjusting the CVO logic to return NoUpstream in the empty-upstream case, but at the moment, clearing the channel is the best approach for "the CVO keeps complaining that it can't hit the upstream, and I want ot quiet it down [6]". The 'oc adm upgrade channel' command just landed for 4.9 in [7]. [1]: https://github.com/openshift/cluster-version-operator/blame/2c4931dc283c551938be1a00fee290de0b79d99c/pkg/cvo/availableupdates.go#L27-L31 [2]: openshift/cluster-version-operator@ab4d84a#diff-78f2af341fa49292dd6930f378018867R24 [3]: https://github.com/openshift/api/blame/0422dc17083e9e8df18d029f3f34322e96e9c326/config/v1/types_cluster_version.go#L56-L57 [4]: https://github.com/openshift/api/blame/0422dc17083e9e8df18d029f3f34322e96e9c326/config/v1/types_cluster_version.go#L62-L63 [5]: https://github.com/openshift/cluster-version-operator/blob/2c4931dc283c551938be1a00fee290de0b79d99c/pkg/cvo/cvo.go#L602 [6]: https://bugzilla.redhat.com/show_bug.cgi?id=1827378 [7]: openshift/oc#576
4085e6d to
4244510
Compare
|
/cherrypick enterprise-4.9 |
|
@jeana-redhat: new pull request created: #34900 DetailsIn response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
I would prefer to configure this by clearing the
upstreamsetting, which seems more intuitive for "there is no upstream" to me. But sadly, it seems that the CVO has been falling back to a default URI when the ClusterVersion upstream is empty since way back, and that this behavior is enshrined in the API. Although the channel docs also talk about defaults, the only channel defaulting in the CVO is when creating a ClusterVersion object after the in-cluster copy was (accidentally?) deleted. So maybe we could talk folks into adjusting the CVO logic to returnNoUpstreamin the empty-upstream case, but at the moment, clearing the channel is the best approach for "the CVO keeps complaining that it can't hit the upstream, and I want to quiet it down" (rhbz#1827378)5.