-
Notifications
You must be signed in to change notification settings - Fork 286
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
Check, take action or notify and auto-update Entrance criteria for releases #5332
Comments
Tagging @opensearch-project/engineering-effectiveness to get some inputs or feedback if they have any. |
As of current, roadmap can be auto-updated with a Each meta should also have corresponding label for version so that they can automatically categorized on roadmap. Thanks. |
See the above linked PRs for tasks completed. There is some ambiguity around the code as the data to take actions is mainly retrieved from the OpenSearch metrics cluster and we need to form queries, write tests for those queries as well as libraries for robustness. The data in metrics cluster needs to be up to date in order to take action accordingly. With 3.0.0 going on we are facing one of the edge case with the qualifier. See opensearch-project/opensearch-metrics#121 |
Is your feature request related to a problem? Please describe
Coming from #5171
Just focusing on entrance criteria for the scope of this issue
Currently, the release manager must manually verify the entrance criteria before or during the release window. This process is prone to human error and allows for inconsistent execution of releases, as there is no standardized approach. To streamline the process and eliminate these challenges, it would be ideal for these criteria to be automatically checked, updated, and made actionable.
Describe the solution you'd like
The current entrance criterias as as below:
* Each component release issue has an assigned owner
Check:
Check if all the release component issues have release owners assigned using data in the metrics cluster.
Action:
If not, below are the AI:
* Documentation draft PRs are up and in tech review for all component changes.
Check:
Check if the PRs are up for all raised issues. Sample query:
Action:
* Sanity testing is done for all components.
Check:
No direct way of checking it. We reply on component teams to do the needful. From distribution side, we can run smoke tests and validation for sanity.
* Code coverage has not decreased (all new code has tests).
Check:
We are indexing the code coverage data in the metrics cluster. For given release (eg: 2.19) we can check the difference against the last minor (2.18 in this case).
Action:
Notify the teams by commenting on the GH issue as well as main release issue asking to add tests for missing ones as the code coverage as decreased.
Heads-up: Looks like code coverage reporting is a problem for many components. We can use this opportunity to fix that as well.
* Release notes are ready and available for all components.
Check:
Check using metrics data.
Action:
Tag and notify by commenting on component Release issue asking to fulfill the missing criteria.
* Roadmap is up-to-date (information is available to create release highlights).
We rely on component owners and maintainers to update the roadmap. Wondering if making sure the issues are closed as part of the check makes sense @Pallavi-AWS @getsaurabh02
* Release ticket is cut, and there's a forum post announcing the start of the window.
Check:
Check using metrics data and compare against the release input manifest to match the number of components and issues.
Action:
If some are missing, build/infra repo will create the component release issue accordingly.
* Any necessary security reviews are complete.
We rely on component owners and maintainers to follow through.
The text was updated successfully, but these errors were encountered: