Skip to content

Latest commit

 

History

History
343 lines (261 loc) · 15.5 KB

releasing.md

File metadata and controls

343 lines (261 loc) · 15.5 KB

WARNING WARNING WARNING WARNING WARNING

PLEASE NOTE: This document applies to the HEAD of the source tree

If you are using a released version of Kubernetes, you should refer to the docs that go with that version.

The latest 1.0.x release of this document can be found [here](http://releases.k8s.io/release-1.0/docs/devel/releasing.md).

Documentation for other releases can be found at releases.k8s.io.

Releasing Kubernetes

This document explains how to cut a release, and the theory behind it. If you just want to cut a release and move on with your life, you can stop reading after the first section.

How to cut a Kubernetes release

Regardless of whether you are cutting a major or minor version, cutting a release breaks down into four pieces:

  1. Selecting release components.
  2. Tagging and merging the release in Git.
  3. Building and pushing the binaries.
  4. Writing release notes.

You should progress in this strict order.

Building a New Major/Minor Version (vX.Y.0)

Selecting Release Components

When cutting a major/minor release, your first job is to find the branch point. We cut vX.Y.0 releases directly from master, which is also the branch that we have most continuous validation on. Go first to the main GCE Jenkins end-to-end job and next to the Critical Builds page and hopefully find a recent Git hash that looks stable across at least kubernetes-e2e-gce and kubernetes-e2e-gke-ci. First glance through builds and look for nice solid rows of green builds, and then check temporally with the other Critical Builds to make sure they're solid around then as well. Once you find some greens, you can find the Git hash for a build by looking at the "Console Log", then look for githash=. You should see a line line:

+ githash=v0.20.2-322-g974377b

Because Jenkins builds frequently, if you're looking between jobs (e.g. kubernetes-e2e-gke-ci and kubernetes-e2e-gce), there may be no single githash that's been run on both jobs. In that case, take the a green kubernetes-e2e-gce build (but please check that it corresponds to a temporally similar build that's green on kubernetes-e2e-gke-ci). Lastly, if you're having trouble understanding why the GKE continuous integration clusters are failing and you're trying to cut a release, don't hesitate to contact the GKE oncall.

Before proceeding to the next step:

export BRANCHPOINT=v0.20.2-322-g974377b

Where v0.20.2-322-g974377b is the git hash you decided on. This will become our (retroactive) branch point.

Branching, Tagging and Merging

Do the following:

  1. export VER=x.y (e.g. 0.20 for v0.20)
  2. cd to the base of the repo
  3. git fetch upstream && git checkout -b release-${VER} ${BRANCHPOINT} (you did set ${BRANCHPOINT}, right?)
  4. Make sure you don't have any files you care about littering your repo (they better be checked in or outside the repo, or the next step will delete them).
  5. make clean && git reset --hard HEAD && git clean -xdf
  6. make (TBD: you really shouldn't have to do this, but the swagger output step requires it right now)
  7. ./build/mark-new-version.sh v${VER}.0 to mark the new release and get further instructions. This creates a series of commits on the branch you're working on (release-${VER}), including forking our documentation for the release, the release version commit (which is then tagged), and the post-release version commit.
  8. Follow the instructions given to you by that script. They are canon for the remainder of the Git process. If you don't understand something in that process, please ask!

TODO: how to fix tags, etc., if you have to shift the release branchpoint.

Building and Pushing Binaries

In your git repo (you still have ${VER} set from above right?):

  1. git checkout upstream/master && build/build-official-release.sh v${VER}.0 (the build-official-release.sh script is version agnostic, so it's best to run it off master directly).
  2. Follow the instructions given to you by that script.
  3. At this point, you've done all the Git bits, you've got all the binary bits pushed, and you've got the template for the release started on GitHub.

Writing Release Notes

This helpful guide describes how to write release notes for a major/minor release. In the release template on GitHub, leave the last PR number that the tool finds for the .0 release, so the next releaser doesn't have to hunt.

Building a New Patch Release (vX.Y.Z for Z > 0)

Selecting Release Components

We cut vX.Y.Z releases from the release-vX.Y branch after all cherry picks to the branch have been resolved. You should ensure all outstanding cherry picks have been reviewed and merged and the branch validated on Jenkins (validation TBD). See the Cherry Picks for more information on how to manage cherry picks prior to cutting the release.

Tagging and Merging

  1. export VER=x.y (e.g. 0.20 for v0.20)
  2. export PATCH=Z where Z is the patch level of vX.Y.Z
  3. cd to the base of the repo
  4. git fetch upstream && git checkout -b upstream/release-${VER}
  5. Make sure you don't have any files you care about littering your repo (they better be checked in or outside the repo, or the next step will delete them).
  6. make clean && git reset --hard HEAD && git clean -xdf
  7. make (TBD: you really shouldn't have to do this, but the swagger output step requires it right now)
  8. ./build/mark-new-version.sh v${VER}.${PATCH} to mark the new release and get further instructions. This creates a series of commits on the branch you're working on (release-${VER}), including forking our documentation for the release, the release version commit (which is then tagged), and the post-release version commit.
  9. Follow the instructions given to you by that script. They are canon for the remainder of the Git process. If you don't understand something in that process, please ask! When proposing PRs, you can pre-fill the body with hack/cherry_pick_list.sh upstream/release-${VER} to inform people of what is already on the branch.

TODO: how to fix tags, etc., if the release is changed.

Building and Pushing Binaries

In your git repo (you still have ${VER} and ${PATCH} set from above right?):

  1. git checkout upstream/master && build/build-official-release.sh v${VER}.${PATCH} (the build-official-release.sh script is version agnostic, so it's best to run it off master directly).
  2. Follow the instructions given to you by that script. At this point, you've done all the Git bits, you've got all the binary bits pushed, and you've got the template for the release started on GitHub.

Writing Release Notes

Run hack/cherry_pick_list.sh ${VER}.${PATCH}~1 to get the release notes for the patch release you just created. Feel free to prune anything internal, like you would for a major release, but typically for patch releases we tend to include everything in the release notes.

Origin of the Sources

Kubernetes may be built from either a git tree (using hack/build-go.sh) or from a tarball (using either hack/build-go.sh or go install) or directly by the Go native build system (using go get).

When building from git, we want to be able to insert specific information about the build tree at build time. In particular, we want to use the output of git describe to generate the version of Kubernetes and the status of the build tree (add a -dirty prefix if the tree was modified.)

When building from a tarball or using the Go build system, we will not have access to the information about the git tree, but we still want to be able to tell whether this build corresponds to an exact release (e.g. v0.3) or is between releases (e.g. at some point in development between v0.3 and v0.4).

Version Number Format

In order to account for these use cases, there are some specific formats that may end up representing the Kubernetes version. Here are a few examples:

  • v0.5: This is official version 0.5 and this version will only be used when building from a clean git tree at the v0.5 git tag, or from a tree extracted from the tarball corresponding to that specific release.
  • v0.5-15-g0123abcd4567: This is the git describe output and it indicates that we are 15 commits past the v0.5 release and that the SHA1 of the commit where the binaries were built was 0123abcd4567. It is only possible to have this level of detail in the version information when building from git, not when building from a tarball.
  • v0.5-15-g0123abcd4567-dirty or v0.5-dirty: The extra -dirty prefix means that the tree had local modifications or untracked files at the time of the build, so there's no guarantee that the source code matches exactly the state of the tree at the 0123abcd4567 commit or at the v0.5 git tag (resp.)
  • v0.5-dev: This means we are building from a tarball or using go get or, if we have a git tree, we are using go install directly, so it is not possible to inject the git version into the build information. Additionally, this is not an official release, so the -dev prefix indicates that the version we are building is after v0.5 but before v0.6. (There is actually an exception where a commit with v0.5-dev is not present on v0.6, see later for details.)

Injecting Version into Binaries

In order to cover the different build cases, we start by providing information that can be used when using only Go build tools or when we do not have the git version information available.

To be able to provide a meaningful version in those cases, we set the contents of variables in a Go source file that will be used when no overrides are present.

We are using pkg/version/base.go as the source of versioning in absence of information from git. Here is a sample of that file's contents:

var (
    gitVersion   string = "v0.4-dev"  // version from git, output of $(git describe)
    gitCommit    string = ""          // sha1 from git, output of $(git rev-parse HEAD)
)

This means a build with go install or go get or a build from a tarball will yield binaries that will identify themselves as v0.4-dev and will not be able to provide you with a SHA1.

To add the extra versioning information when building from git, the hack/build-go.sh script will gather that information (using git describe and git rev-parse) and then create a -ldflags string to pass to go install and tell the Go linker to override the contents of those variables at build time. It can, for instance, tell it to override gitVersion and set it to v0.4-13-g4567bcdef6789-dirty and set gitCommit to 4567bcdef6789... which is the complete SHA1 of the (dirty) tree used at build time.

Handling Official Versions

Handling official versions from git is easy, as long as there is an annotated git tag pointing to a specific version then git describe will return that tag exactly which will match the idea of an official version (e.g. v0.5).

Handling it on tarballs is a bit harder since the exact version string must be present in pkg/version/base.go for it to get embedded into the binaries. But simply creating a commit with v0.5 on its own would mean that the commits coming after it would also get the v0.5 version when built from tarball or go get while in fact they do not match v0.5 (the one that was tagged) exactly.

To handle that case, creating a new release should involve creating two adjacent commits where the first of them will set the version to v0.5 and the second will set it to v0.5-dev. In that case, even in the presence of merges, there will be a single commit where the exact v0.5 version will be used and all others around it will either have v0.4-dev or v0.5-dev.

The diagram below illustrates it.

Diagram of git commits involved in the release

After working on v0.4-dev and merging PR 99 we decide it is time to release v0.5. So we start a new branch, create one commit to update pkg/version/base.go to include gitVersion = "v0.5" and git commit it.

We test it and make sure everything is working as expected.

Before sending a PR for it, we create a second commit on that same branch, updating pkg/version/base.go to include gitVersion = "v0.5-dev". That will ensure that further builds (from tarball or go install) on that tree will always include the -dev prefix and will not have a v0.5 version (since they do not match the official v0.5 exactly.)

We then send PR 100 with both commits in it.

Once the PR is accepted, we can use git tag -a to create an annotated tag pointing to the one commit that has v0.5 in pkg/version/base.go and push it to GitHub. (Unfortunately GitHub tags/releases are not annotated tags, so this needs to be done from a git client and pushed to GitHub using SSH or HTTPS.)

Parallel Commits

While we are working on releasing v0.5, other development takes place and other PRs get merged. For instance, in the example above, PRs 101 and 102 get merged to the master branch before the versioning PR gets merged.

This is not a problem, it is only slightly inaccurate that checking out the tree at commit 012abc or commit 345cde or at the commit of the merges of PR 101 or 102 will yield a version of v0.4-dev but those commits are not present in v0.5.

In that sense, there is a small window in which commits will get a v0.4-dev or v0.4-N-gXXX label and while they're indeed later than v0.4 but they are not really before v0.5 in that v0.5 does not contain those commits.

Unfortunately, there is not much we can do about it. On the other hand, other projects seem to live with that and it does not really become a large problem.

As an example, Docker commit a327d9b91edf has a v1.1.1-N-gXXX label but it is not present in Docker v1.2.0:

$ git describe a327d9b91edf
v1.1.1-822-ga327d9b91edf

$ git log --oneline v1.2.0..a327d9b91edf
a327d9b91edf Fix data space reporting from Kb/Mb to KB/MB

(Non-empty output here means the commit is not present on v1.2.0.)

Release Notes

No official release should be made final without properly matching release notes.

There should be made available, per release, a small summary, preamble, of the major changes, both in terms of feature improvements/bug fixes and notes about functional feature changes (if any) regarding the previous released version so that the BOM regarding updating to it gets as obvious and trouble free as possible.

After this summary, preamble, all the relevant PRs/issues that got in that version should be listed and linked together with a small summary understandable by plain mortals (in a perfect world PR/issue's title would be enough but often it is just too cryptic/geeky/domain-specific that it isn't).

Analytics