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
Next-gen (3.0) chart layout #977
Comments
I would like to do a GA release of the chart in #976 sooner. The scope there Is there some reason we believe a GA release would prevent us from taking Keeping it unreleased requires continuing to deal with the problems associated That said, some preliminary reviews of the options: Single chart, separate values.yamlsThis seeks to allow installing chart-releaser's package step appears to just be a wrapper around the Actually modifying these tools to provide some sort of values.yaml injection The simplicity of the packager does allow us to just copy a chart directory Injecting a complete values.yaml rather than overlaying overrides requires some Separate charts with shared templatesWe'd ideally have some clearer expectation around what we cannot do now (or Our topologies generally all involve installing a Kong Deployment, with the The operator is an exception because its chart would presumably not deploy The one example I have for using a library dependency (Bitnami's chart library) Dependency charts load the same way regardless of whether they're rendered in Symlinking a library template file may be a way to circumvent the need for a |
I think that the first option @rainest describes ( Picking the second option has its issues but I think it is cleaner and closer to the standard approach (e.g. to how bitnami does that). I’d see that with the following steps to be executed:
We won’t be able to predict up-front all templates that are really to be extracted to a common chart so I think that trying to create new charts would be the best method to find out what we can extract from Regarding the issue with multi-step releases in case of modifying something in |
Problem Statement
We've decided that the 3.0 version of Helm charts will come with a complete revamp of the user experience: instead of starting from the "implementation" (1-1 relationship with the deployment that is either Kong or KIC or both) we will start from the "use case" (KIC, no KIC, Konnect DP, Operator, etc.)
Proposed Solution
First step: merge #976 (3.0.0-alpha.1)
Second step: reach agreement on the desired UX. We have two options on the table
kong/ingress
,kong/konnect-dp
,kong/operator
- perhaps more) where default values have a reasonable "batteries included" capabilityvalues.yaml
examples for different use casesWe currently believe that Option 1 is more promising but more research needs to go into it.
Third step: 3.0.0-alpha.2 is implemented in the chosen option out of the above
Fourth step: 3.0.0-alpha.N promoted to 3.0.0
Acceptance Criteria
The text was updated successfully, but these errors were encountered: