-
Notifications
You must be signed in to change notification settings - Fork 24
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
Explicitly restrict to amd64 hosts #326
Conversation
Thanks for the patches! One quick comment: we use a commit message style similar to the kernel where the "topic" (as in |
Sure! Can this done at the squash & merge point? |
That's not a workflow that I usually follow, but if you're not comfortable doing it yourself, I can look into it, for sure. |
Sorry I may be misunderstanding something here. This seems to be specific to the commit names in my fork, which don't necessarily carry onto the main repo in a PR, depending on the merge method. Would you prefer that I change the commit messages in my fork to align with the main repo? |
No worries, I am used to working on projects (on github) that resolve a PR with either "Rebase and merge" or "Create a merge commit". On this project we rebase. With these approaches the submitter's commit messages are preserved. I am aware of the squash option but I have not used it myself. I'm used to simply asking the submitter to update the commit messages. However, I don't require it as you enabled 'Maintainers are allowed to edit this pull request' - so if all else fails I can just fix up the commits myself prior to merge. On the topic of commit messages, we also prefer to have the "Signed-off-by" line present. This is not something I would add/change for you: https://stackoverflow.com/questions/1962094/what-is-the-sign-off-feature-in-git-for In case you haven't seen it yet, the CI's yaml linter has detected issues: |
Aah okay I understand, sorry I actually just was not aware of the use of this convention. I am very happy to do this I just don't have all the details. Would it be helpful if I change the commit(s) above into one with the message, for example but feel free to edit, 'config: restrict to amd64 nodes through affinity'? I can certainly do something like this! Additionally I will fix the indentation! |
Signed-off-by: Tiago Ferreira <me@tiferrei.com> feat: use samba fork config: fix indentation in affinity block Signed-off-by: Tiago Ferreira <me@tiferrei.com>
The CI is failing due to golang related issues that ought to be unrelated to this PR. I wonder if it's triggered by the recent go release?
Regardless of the cause for this, I pull the PR locally and ran the YAML lint checks again and it complains about the indent for two lines. These indents can be fixed by the following: diff --git a/config/manager-full/auth_proxy_patch.yaml b/config/manager-full/auth_proxy_patch.yaml
index 3b39da3..154e40a 100644
--- a/config/manager-full/auth_proxy_patch.yaml
+++ b/config/manager-full/auth_proxy_patch.yaml
@@ -13,7 +13,7 @@ spec:
requiredDuringSchedulingIgnoredDuringExecution:
nodeSelectorTerms:
- matchExpressions:
- - key: beta.kubernetes.io/arch
+ - key: beta.kubernetes.io/arch
operator: In
values:
- amd64
diff --git a/config/manager/manager.yaml b/config/manager/manager.yaml
index 593028d..819880d 100644
--- a/config/manager/manager.yaml
+++ b/config/manager/manager.yaml
@@ -36,7 +36,7 @@ spec:
requiredDuringSchedulingIgnoredDuringExecution:
nodeSelectorTerms:
- matchExpressions:
- - key: beta.kubernetes.io/arch
+ - key: beta.kubernetes.io/arch
operator: In
values:
- amd64 If you make these updates I think I'd be OK with bypassing the failing CI checks for the Go code. |
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.
See previous comment.
Should be good now! |
Changes look OK but the new patch lacks a sign off by, etc. If you prefer I squash them together I can do so later after. Is that what you'd prefer? |
Signed-off-by: Tiago Ferreira <me@tiferrei.com>
Sorry, now amended properly. |
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.
Those two commits should be squashed into one. See how I did it in my (untested) branch: synarete@fa85754
requiredDuringSchedulingIgnoredDuringExecution: | ||
nodeSelectorTerms: | ||
- matchExpressions: | ||
- key: beta.kubernetes.io/arch |
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.
key
should be kubernetes.io/arch
. See k8s labels-annotations-taints docs as well kubernetes/kubernetes#73333
Note that we already use those values for samba pods' selector: https://github.com/samba-in-kubernetes/samba-operator/blob/master/internal/planner/selector.go#L14
requiredDuringSchedulingIgnoredDuringExecution: | ||
nodeSelectorTerms: | ||
- matchExpressions: | ||
- key: beta.kubernetes.io/arch |
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.
Same comment as above: should be kubernetes.io/arch
Sorry, I have given more time to this PR than I can reasonably dedicate. I use my fork on my cluster, I have no practical motivation to keep investing time in these changes, or squashing and renaming commits that make no difference to me. I love the project, if you'd like to have the changes upstreamed you are welcome to do any edits you'd like, or close the PR if it is not of value. |
This pull request now has conflicts with the target branch. Please resolve these conflicts and force push the updated branch. |
Until ARM support is ready, I have been running this version on my mixed cluster to make sure that only amd64 nodes serve the deployments.