Skip to content

chore(docs): update release process for Cloud Build images and make release-new #1316

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.

Sign up for GitHub

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

Open
hakman wants to merge 1 commit into
base: master
Choose a base branch
from
+48 −75
Open

chore(docs): update release process for Cloud Build images and make release-new #1316

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
123 changes: 48 additions & 75 deletions docs/release_process.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,116 +1,89 @@
# Release Process

These are notes to help follow a consistent release process. See something
important missing? Please submit a pull request to add anything else that would
be useful!
These are notes to help follow a consistent release process. See something important missing? Please submit a pull request to add anything else that would be useful!

## Prerequisites

Ensure access to the container image [staging registry](https://console.cloud.google.com/gcr/images/k8s-staging-npd/global/node-problem-detector).
Add email to `k8s-infra-staging-npd` group in sig-node [groups.yaml](https://github.com/kubernetes/k8s.io/blob/main/groups/sig-node/groups.yaml).
See example https://github.com/kubernetes/k8s.io/pull/1599.

The steps below also require the following tools:

- [crane](https://github.com/google/go-containerregistry/blob/main/cmd/crane/README.md), to read image digests from the registries.
- Docker and a Go toolchain, to build the release tarballs with `make release-new`.

## Preparing for a release

There are a few steps that should be taken prior to creating the actual release
itself.
There are a few steps that should be taken prior to creating the actual release itself.

1. Collect changes since last release. This can be done by looking directly at
merged commit messages (``git log [last_release_tag]...HEAD``), or by
viewing the changes on GitHub (example: https://github.com/kubernetes/node-problem-detector/compare/v0.8.15...master).
1. Collect changes since last release. This can be done by looking directly at merged commit messages (``git log [last_release_tag]...HEAD``), or by viewing the changes on GitHub (example: https://github.com/kubernetes/node-problem-detector/compare/v1.35.0...master).

2. Based on the changes to be included in the release, determine what the next
release number should be. We strive to follow [SemVer](https://semver.org/)
as much as possible.
2. Based on the changes to be included in the release, determine what the next release number should be. We strive to follow [SemVer](https://semver.org/) as much as possible.

3. Update [CHANGELOG](https://github.com/kubernetes/node-problem-detector/blob/master/CHANGELOG.md)
with all significant changes.
3. Update [CHANGELOG](https://github.com/kubernetes/node-problem-detector/blob/master/CHANGELOG.md) with all significant changes.

## Create release

### Create the new version tag

#### Option 1
```
# Use v0.8.17 as an example.
git clone git@github.com:kubernetes/node-problem-detector.git
cd node-problem-detector/
git tag v0.8.17
git push origin v0.8.17
```
Update [version.txt](https://github.com/kubernetes/node-problem-detector/blob/master/version.txt) (example https://github.com/kubernetes/node-problem-detector/pull/1312).

#### Option 2
Update [version.txt](https://github.com/kubernetes/node-problem-detector/blob/master/version.txt)
(example https://github.com/kubernetes/node-problem-detector/pull/869).
### Container images are built automatically

### Build and push artifacts
This step builds the NPD into container files and tar files.
- The container file is pushed to the [staging registry](https://console.cloud.google.com/gcr/images/k8s-staging-npd/global/node-problem-detector).
You will promote the new image to registry.k8s.io later.
- The tar files are generated locally. You will upload those to github in the
release note later.
Pushing the tag triggers a Cloud Build job ([cloudbuild.yaml](../cloudbuild.yaml)) that runs `make push-container` and `make push-container-windows`, which build and push both multi-arch images to the staging registry automatically:

**Note: You need the access mentioned in the [prerequisites](#prerequisites)
section to perform steps in this section.**
- `gcr.io/k8s-staging-npd/node-problem-detector:v1.36.0` (linux/amd64, linux/arm64)
- `gcr.io/k8s-staging-npd/node-problem-detector-windows:v1.36.0` (windows/amd64)

```
# One-time setup
sudo apt-get install libsystemd-dev gcc-aarch64-linux-gnu
No local container build or push (`make release`) is needed. Verify the images in the [staging registry](https://console.cloud.google.com/gcr/images/k8s-staging-npd/global/node-problem-detector) before continuing.

cd node-problem-detector
make release
### Promote the NPD images to registry.k8s.io

# Get SHA256 of the tar files. For example
sha256sum node-problem-detector-v0.8.17-linux_amd64.tar.gz
sha256sum node-problem-detector-v0.8.17-linux_arm64.tar.gz
sha256sum node-problem-detector-v0.8.17-windows_amd64.tar.gz
1. Get the digests of the new NPD images from the staging registry:
```
crane digest gcr.io/k8s-staging-npd/node-problem-detector:v1.36.0
crane digest gcr.io/k8s-staging-npd/node-problem-detector-windows:v1.36.0
```
2. Promote both images to registry.k8s.io by adding the digests to [images.yaml](https://github.com/kubernetes/k8s.io/blob/main/registry.k8s.io/images/k8s-staging-npd/images.yaml) in the kubernetes/k8s.io repo: the linux digest under the `node-problem-detector` entry and the windows digest under the `node-problem-detector-windows` entry (example https://github.com/kubernetes/k8s.io/pull/9707).
3. After the promotion PR merges, verify that the promoted digests match staging:
```
crane digest registry.k8s.io/node-problem-detector/node-problem-detector:v1.36.0
crane digest registry.k8s.io/node-problem-detector/node-problem-detector-windows:v1.36.0
```

# Get MD5 of the tar files. For example
md5sum node-problem-detector-v0.8.17-linux_amd64.tar.gz
md5sum node-problem-detector-v0.8.17-linux_arm64.tar.gz
md5sum node-problem-detector-v0.8.17-windows_amd64.tar.gz
### Build the release artifacts

# Verify container image in staging registry and get SHA256.
docker pull gcr.io/k8s-staging-npd/node-problem-detector:v0.8.17
docker image ls gcr.io/k8s-staging-npd/node-problem-detector --digests
```
This step runs **after** the image promotion, because the release binaries are extracted from the promoted registry.k8s.io images.

### Promote new NPD image to registry.k8s.io
1. Get the SHA256 from the new NPD image from the [staging registry](https://console.cloud.google.com/gcr/images/k8s-staging-npd/global/node-problem-detector)
or previous step.
2. Promote the NPD image to registry.k8s.io ([images.yaml](https://github.com/kubernetes/k8s.io/blob/main/registry.k8s.io/images/k8s-staging-npd/images.yaml), example https://github.com/kubernetes/k8s.io/pull/6523).
3. Verify the container image.
```
docker pull registry.k8s.io/node-problem-detector/node-problem-detector:v0.8.17
docker image ls registry.k8s.io/node-problem-detector/node-problem-detector:v0.8.17
# One-time setup on Linux hosts (journald build tags for the test binaries).
sudo apt-get install libsystemd-dev gcc-aarch64-linux-gnu

cd node-problem-detector
make release-new VERSION=v1.36.0
```

Only `VERSION` needs to be set (`TAG` and `NPD_NAME_VERSION` are derived from it), so there is no need to check out the release tag. `make release-new`:

- Pulls the promoted `registry.k8s.io/node-problem-detector/node-problem-detector:v1.36.0` and `.../node-problem-detector-windows:v1.36.0` images and extracts the release binaries from them (`docker create` + `docker cp`).
- Packages `node-problem-detector-v1.36.0-linux_amd64.tar.gz`, `node-problem-detector-v1.36.0-linux_arm64.tar.gz` and `node-problem-detector-v1.36.0-windows_amd64.tar.gz`, each with a `.sha512` file next to it.
- Prints the SHA256 and MD5 of the tarballs ([hack/print-tar-sha-md5.sh](../hack/print-tar-sha-md5.sh)), ready to paste into the release note.

### Create the release note

Go to https://github.com/kubernetes/node-problem-detector/releases, draft a new
release note and publish. Make sure to include the following in the body of the
release note:
Go to https://github.com/kubernetes/node-problem-detector/releases, draft a new release note and publish. Make sure to include the following in the body of the release note:

1. For convenience, add a link to easily view the changes since the last
release (e.g.
[https://github.com/kubernetes/node-problem-detector/compare/v0.8.15...v0.8.17](https://github.com/kubernetes/node-problem-detector/compare/v0.8.15...v0.8.17)).
1. For convenience, add a link to easily view the changes since the last release (e.g. [https://github.com/kubernetes/node-problem-detector/compare/v1.35.0...v1.36.0](https://github.com/kubernetes/node-problem-detector/compare/v1.35.0...v1.36.0)).

2. There is no need to duplicate everything from the CHANGELOG, but include the
most significant things so someone just viewing the release entry will have
an idea of what it includes.
2. There is no need to duplicate everything from the CHANGELOG, but include the most significant things so someone just viewing the release entry will have an idea of what it includes.

3. Provide a link to the new image release (e.g. `Image:
registry.k8s.io/node-problem-detector/node-problem-detector:v0.8.17`)
3. Provide a link to the new image release (e.g. `Image: registry.k8s.io/node-problem-detector/node-problem-detector:v1.36.0`)

4. Upload the tar files built from [pevious step](#build-and-push-artifacts),
and include the SHA and MD5.
4. Upload the tar files built in the [previous step](#build-the-release-artifacts), and include the SHA and MD5 printed by `make release-new`.

## Post release steps

1. Update image version in node-problem-detector repo, so anyone deploying
directly from the repo deployment file will get the newest image deployed.
Example https://github.com/kubernetes/node-problem-detector/pull/897.
1. Update image version in node-problem-detector repo, so anyone deploying directly from the repo deployment file will get the newest image deployed. Example https://github.com/kubernetes/node-problem-detector/pull/897.

2. Update the NPD version in [kubernetes/kubernetes](https://github.com/kubernetes/kubernetes)
repo, so that kubernetes clusters use the new NPD version. Example
https://github.com/kubernetes/kubernetes/pull/123740.
2. Update the NPD version in [kubernetes/kubernetes](https://github.com/kubernetes/kubernetes) repo, so that kubernetes clusters use the new NPD version. Example https://github.com/kubernetes/kubernetes/pull/123740.