Join GitHub today
GitHub is home to over 40 million developers working together to host and review code, manage projects, and build software together.Sign up
kubernetes-csi release process #807
This gets added as a KEP because the proposed solution will have to run on CNCF resources (Prow) and publish to the same destination as core Kubernetes (gcr.io). However, this proposal is independent of any particular Kubernetes release and thus does not need to be tracked.
Producing a combined release of all components no longer is a goal. Instead, each component gets released independently and CSI driver developers put together a combined deployment, similar to the example in the csi-driver-host-path repo.
@msau42 I've updated the KEP - no more combined release, "rc" tags explicitly mentioned, etc.
Getting images published isn't supported well (or at all) in Prow at the moment, but it looks like there is a trick that is sufficient to test in a
We are in the process of enabling Prow testing for several repos in the kubernetes-csi organization (kubernetes/enhancements#807, kubernetes#11715). Instead of enabling the /test trigger individually for just the repos where it works and then remember to update it when adding new ones, it is simpler to just enable it for the entire org.
prow.sh is a generic script that does unit and E2E testing inside a Prow job. xref: kubernetes/enhancements#807 This separates which kind of tests to run (serial/parallel, stable/alpha) from specifying what those tests are. Regular expressions for choosing tests only need to be specified once now, the actual focus and skip options are derived from that. Merging all junit files into one while throwing away the tests that we are not interested in increases load times in Spyglass. The implementation in Go is not as fancy as a XLST style sheet or an XPath-based filtering tool, but has the advantage that it works without introducing additional dependencies. csi-sanity testing is useful for CSI drivers. There's no longer a need to explicitly provide a configuration and deployment for the Kubernetes version that is being used for testing. Instead, the "latest" settings and deployment are used as fallback.
Issues go stale after 90d of inactivity.
If this issue is safe to close now please do so with
Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
[APPROVALNOTIFIER] This PR is APPROVED
The full list of commands accepted by this bot can be found here.
The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing