pulsar-helm-chart icon indicating copy to clipboard operation
pulsar-helm-chart copied to clipboard

replace handmade release script with official action

Open tisonkun opened this issue 1 year ago • 1 comments

This refers to #290.

The helm/chart-releaser-action update index.yaml in an incremental manner, so we need to push the existing index.yaml file to gh-pages later.

tisonkun avatar Oct 09 '22 05:10 tisonkun

cc @michaeljmarshall @ericsyh @mattisonchao @yaalsn

tisonkun avatar Oct 09 '22 06:10 tisonkun

@michaeljmarshall Thanks for your follow-up. Let me check your investigation today and feedback. My first object is to decouple the Chart index file from the website :)

tisonkun avatar Oct 19 '22 01:10 tisonkun

@tisonkun - where do you think we should host it? One concern with moving it is that helm repo update will break.

michaeljmarshall avatar Oct 19 '22 01:10 michaeljmarshall

will break

Yep. As mentioned in #290, we can set a server-side redirection in the website .htaccess. The reason is that, website build process doesn't handle Chart yaml index so that it's possible we accidentally remove the file - they're totally different topics.

Setting up a redirection is in the domain of a website maintenance and avoid further development accidentally modifies it.

tisonkun avatar Oct 19 '22 02:10 tisonkun

Sure, I agree that the current workflow is a bit of a hack, and the current index.yaml is not very well protected. I added a comment to #290, too.

michaeljmarshall avatar Oct 19 '22 03:10 michaeljmarshall

@michaeljmarshall I've narrow the scope to replace the lint scripts with chart-testing-action. PTAL.

I'm going through your proposal on release process now.

tisonkun avatar Oct 19 '22 04:10 tisonkun

Weird. Is it an unstable test?

tisonkun avatar Oct 19 '22 10:10 tisonkun

@michaeljmarshall I don't think the failures are related to this patch. Could you take a looks and rerun the tests?

tisonkun avatar Oct 19 '22 12:10 tisonkun

@tisonkun - yes, the TLS Installation and ZK & BK TLS Only tests are unstable. I have spent some time looking into them this week, but I haven't found the root cause. It seems like there may be some instability with the environment because some of the errors are about refused connections to ports that appear to be working for some pods but not others.

michaeljmarshall avatar Oct 19 '22 15:10 michaeljmarshall

Here is an issue for the flaky tests: #311

michaeljmarshall avatar Oct 19 '22 20:10 michaeljmarshall