Master list can be found here
-
Feature areas:
A specific portion of K8s functionality. Ex. sig-auth, sig-apps, sig-big-data, sig-cli, sig-ui...
-
Plumbing:
In charge of aspects that cross all other SIGs. Ex. sig-cluster-lifecycle, sig-api-machinery, sig-instrumentation
-
Cloud providers:
One for each cloud provider that is supported. Ex. sig-aws, sig-gcp, sig-openstack, ...
-
Meta:
Aspects of running the k8s project. Ex. sig-architecture, sif-release, sig-product-management, ...
-
Documentation & website: Contributing to other SIGs requires also contributing to documentation. sig-docs
-
Working groups => old way; Subprojects => new way
-
For specific tools (ex. Helm), goals (ex. Resource Management) or areas (ex. Machine Learning).
-
Examples: wg-app-def, wg-apply, wg-cluster-api, ...
-
Groups change around more frequently than SIGs, some might be temporary.
- Figure out which area you would like to contribute to
- Find out which SIG / WG / subproject covers that (tip: ask on #sig-contribex Slack channel)
- Join that SIG / WG / subproject (you should also join the main SIG when joining a WG / subproject)
Disclaimer: Currently cleaning up / refactoring repos. This might change soon!
-
Core repository (where the core code is).
kubernetes/kubernetes (also known as: k/k)
-
Project
Repos for running and organising the project k/Community, k/Features, k/Steering, ...
k/Community: has one folder for every SIG, and for most of WGs, CoC, ... k/Features: you file an issue when preparing a new feature k/Test-Infra and k/Perf-Tests: separate repos for performance tests and other tests
-
Docs / Website
k/website, k/kubernetes-docs-cn, k/kubernetes-docs-ko
-
Developer Tools
k/sample-controller, k/k8s.io, ...
-
Staging Repos
Mirror specific directories api, code-generator, client-go, metrics, ...
-
SIG repos
Some SIGs have their own dedicated repos release, federation, autoscaler
-
Cloud Providers
cloud-provider-azure, cloud-provider-gcp, ...
-
Products & Tools
kubeadm, kubectl, kops, helm, ...
Reporting bugs by filing a detailed bug report (including labels).
Issues are used as specs:
- Feature or API change proposals
- New tests
- Docs.
You start with creating a new issue, with all the relevant labels, and raise it on a SIG meeting or on the list, then when "lazy consensus" is achieved you submit a PR.
List of labels can be found here
We use a prefix with a -:
- sig/label defines which SIG the issue belongs to
- kind/label defines what type the issue is (bug, feature, documentation, ...)
- triage/label for issues that need to be closed (duplicate, needs-information, support, unreproduceable, unresolved, ...)
- priority/label are assigned by SIG leads, signal how critical the issue is (critical-urgent, important-soon, backlog, ...)
- area/label is an optional label, there is no cannonical list (kubectl, api, dns, platform/gce, ...)
- help-wanted does not signal that this is a good first issue (dedicated label for this coming soon for "good first issue")