fix(chart): cilium-configmap with unended condition#42916
fix(chart): cilium-configmap with unended condition#42916joestringer merged 3 commits intocilium:mainfrom
Conversation
|
Commits 033807c, 12be7fc, 6b629df do not match "(?m)^Signed-off-by:". Please follow instructions provided in https://docs.cilium.io/en/stable/contributing/development/contributing_guide/#developer-s-certificate-of-origin |
6b629df to
5e175c4
Compare
|
Commits 033807c, 12be7fc do not match "(?m)^Signed-off-by:". Please follow instructions provided in https://docs.cilium.io/en/stable/contributing/development/contributing_guide/#developer-s-certificate-of-origin |
5e175c4 to
93949e9
Compare
|
Commit 44b1c6d does not match "(?m)^Signed-off-by:". Please follow instructions provided in https://docs.cilium.io/en/stable/contributing/development/contributing_guide/#developer-s-certificate-of-origin |
|
Commits 44b1c6d, 2a27ba7 do not match "(?m)^Signed-off-by:". Please follow instructions provided in https://docs.cilium.io/en/stable/contributing/development/contributing_guide/#developer-s-certificate-of-origin |
|
Thanks for this @mliner, you'll need to make sure all your commits have DCO signoff at the bottom of the commit message to fix that particularly linting error. |
2a27ba7 to
6c6540b
Compare
|
Commit 44b1c6d does not match "(?m)^Signed-off-by:". Please follow instructions provided in https://docs.cilium.io/en/stable/contributing/development/contributing_guide/#developer-s-certificate-of-origin |
6c6540b to
699a04d
Compare
699a04d to
b808efc
Compare
4530e8b to
3b9a12f
Compare
|
I was going to fix up the style things locally and force push to get this merged, but it looks like maintainer pushes to the branch are disabled. Here's a suggested breakdown of the changes here that explain the changes in a way we typically document, and also addresses the current CI failures: I made this version also available here if you agree with the changes and want to pull/update the PR: https://github.com/joestringer/cilium/commits/pr/joe/serviceTopology-helm |
|
@joestringer I've reorganized and documented the commits based on your proposal. Please review. |
|
Thanks. It's still breaking the documentation workflow: The text you proposed looks more useful, but it will need to be in the comments in the |
|
Sorry, looks like the spellchecker is triggering a failure due to the config option: |
It would be great if a pre-commit hook was implemented to resolve these issues before they reach the pipeline. |
|
Yeah good point. I filed #44039 to open some discussion on that. If there's a way to do it in a way that either has minimal impact for the average commit command (or at least has an opt-out) then maybe we can integrate such a change into the tree. |
|
/test |
|
The tests failed after multiple tries, but I don't know if the failures are related. I can't rebase the PR for you but maybe you could try rebasing the PR against the tip of |
The 'loadBalancer' option conditional in the Helm template was missing its closing statement, creating a potential issue where omitting this configuration section could prevent half of the available option configurations from being applied to Cilium. While this issue was not encountered in practice due to default values always being present for loadBalancer configuration, it represents a logical error in the template structure that should be corrected to ensure a valid and robust Helm chart. Signed-off-by: Matej Líner <matej.liner@lablabs.io>
This option was previously hidden from user documentation due to being commented out in the Helm template. The default Helm value aligns with the default value in the Cilium binary, ensuring no functional changes when enabled. Exposing this configuration improves discoverability for users who need to modify this setting. Signed-off-by: Matej Líner <matej.liner@lablabs.io>
Added Labyrinth Labs to USERS.md Signed-off-by: Matej Líner <matej.liner@lablabs.io>
|
/test |
1 similar comment
|
/test |
|
@cilium/sig-datapath @cilium/sig-clustermesh the clustermesh suite seems like it's emitting failures but the failure seems unrelated to the changes proposed in the PR. Could you take a look and follow up? |
@joestringer Marco and I looking into it here: #44313 |
|
Thanks! In that case I'll retrigger the failing test and queue to merge. |
|
@cilium/community PTAL |
|
I bypassed the review request for @cilium/community and will follow up separately how this got stuck for so long. Thanks for the contribution @mliner! |
Description of change
Even though the helm chart seems to be valid (tested locally), I believe that there is a logical issue with the missing
{{- end}}statement (previously commented in commit mentioned below). It can prevent the future problems in case helm will be more strict with template validation.{{- end}}for condition (currently commented)loadbalancer.serviceTopology: falsein order to make it visible in chart-docs so that Cilium users are aware if this option, plus adjusting description to make it more clear for first usersFixes: commit
Release note