12 KiB
OWNERS files
Overview
OWNERS files are used to designate responsibility over different parts of the Kubernetes codebase. Today, we use them to assign the reviewer and approver roles used in our two-phase code review process. Our OWNERS files were inspired by Chromium OWNERS files, which in turn inspired GitHub's CODEOWNERS files.
The velocity of a project that uses code review is limited by the number of people capable of reviewing code. The quality of a person's code review is limited by their familiarity with the code under review. Our goal is to address both of these concerns through the prudent use and maintenance of OWNERS files
OWNERS spec
The mungegithub gitrepos feature is the main consumer of OWNERS files. If this page is out of date, look there.
Each directory that contains a unit of independent code or content may also contain an OWNERS file. This file applies to everything within the directory, including the OWNERS file itself, sibling files, and child directories.
OWNERS files are in YAML format and support the following keys:
approvers: a list of GitHub usernames or aliases that can/approvea PR- [DEPRECATED]
assignees: do not use, equivalent toapprovers(kubernetes/test-infra#3851) labels: a list of GitHub labels to automatically apply to a PRreviewers: a list of GitHub usernames or aliases that are good candidates to/lgtma PR
All users are expected to be assignable. In GitHub terms, this means they are either collaborators of the repo, or members of the organization to which the repo belongs.
A typical OWNERS file looks like:
approvers:
- alice
- bob # this is a comment
reviewers:
- alice
- carol # this is another comment
- sig-foo # this is an alias
Each repo may contain at its root an OWNERS_ALIAS file.
OWNERS_ALIAS files are in YAML format and support the following keys:
aliases: a mapping of alias name to a list of GitHub usernames
We use aliases for groups instead of GitHub Teams, because changes to GitHub Teams are not publicly auditable.
A sample OWNERS_ALISES file looks like:
aliases:
sig-foo:
- david
- erin
sig-bar:
- bob
- frank
GitHub usernames and aliases listed in OWNERS files are case-insensitive.
Code Review Process
This is a simplified description of our full PR testing and merge workflow that conveniently forgets about the existence of tests, to focus solely on the roles driven by OWNERS files.
- The author submits a PR
- Phase 0: Automation determines reviewers and approvers for the PR
- Determine the set of OWNERS files nearest to the code being changed
- Choose two reviewers, and assign them to the PR (choose more than one to avoid assigning to an inactive / unavailable reviewer)
- Choose suggested approvers, one from each OWNERS file, and list them in a comment on the PR
- Phase 1: Humans review the PR
- Reviewers look for general code quality, correctness, sane software engineering, style, etc.
- The PR author cannot be their own reviewer
- If the code changes look good to them, a reviewer types
/lgtmin a PR comment or review; if they change their mind, they/lgtm cancel - Once a reviewer has
/lgtm'ed,prow(@k8s-ci-robot) applies anlgtmlabel to the PR
- Phase 2: Humans approve the PR
- The PR author
/assign's all suggested approvers to the PR, and optionally notifies them (eg: "pinging @foo for approval") - The PR author can be their own approver (assuming they are listed in the appropriate OWNERS files)
- Approvers look for holistic acceptance criteria, including dependencies with other features, forwards/backwards compatibility, API and flag definitions, etc
- If the code changes look good to them, an approver types
/approvein a PR comment or review; if they change their mind, they/approve cancel mungegithub(@k8s-merge-robot) updates its comment in the PR to indicate which approvers still need to approve- Once all approvers (one from each of the previously identified OWNERS files) have approved,
mungegithub(@k8s-merge-robot) applies anapprovedlabel
- The PR author
- Phase 3: Automation merges the PR
- All required labels are present (eg:
lgtm,approved) - All required status checks for the PR are verified as green
- The PR is merged
- All required labels are present (eg:
Quirks of the Process
There are a number of behaviors we've observed that while possible are discouraged, as they go against the intent of this review process. Some of these could be prevented in the future, but this is the state of today.
- An approver's
/lgtmis simultaneously interpreted as an/approve- While a convenient shortcut for some, it can be surprising that the same command is interpreted in one of two ways depending on who the commenter is
- Instead, explicitly write out
/lgtmand/approveto help observers, or save the/lgtmfor a reviewer - This goes against the idea of having at least two sets of eyes on a PR, and may be a sign that there are too few reviewers (who aren't also **approves)
- An approver can
/approve no-issueto bypass the requirement that PR's must have linked issues- There is disagreement within the community over whether requiring every PR to have a linked issue provides value
- Protest is being expressed in the form of overuse of
/approve no-issue - Instead, suggest to the PR author that they edit the PR description to include a linked issue
- This is a sign that we need to actually deliver value with linked issues, or be able to define what a "trivial" PR is in a machine-enforceable way to be able to automatically waive the linked issue requirement
- Technically, anyone who is a member of the kubernetes GitHub organization can drive-by
/lgtma PR- Drive-by reviews from non-members are encouraged as a way of demonstrating experience and intent to become a collaborator or reviewer
- Drive-by
/lgtm's from members may be a sign that our OWNERS files are too small, or that the existing reviewers are too unresponsive - This goes against the idea of specifying reviewers in the first place, to ensure that author is getting actionable feedback from people knowledgeable with the code
- Reviewers, and approvers are unresponsive
- This causes a lot of frustration for authors who often have little visibility into why their PR is being ignored
- Many reviewers and approvers are so overloaded by GitHub notifications that @mention'ing is unlikely to get a quick response
- If an author
/assign's a PR, reviewers and approvers will be made aware of it on their PR dashboard - An author can work around this by manually reading the relevant OWNERS files,
/unassign'ing unresponsive individuals, and/assign'ing others - This is a sign that our OWNERS files are stale; pruning the reviewers and approvers lists would help with this
- Authors are unresponsive
- This costs a tremendous amount of attention as context for an individual PR is lost over time
- This hurts the project in general as its general noise level increases over time
- Instead, close PR's that are untouched after too long (we currently have a bot do this after 90 days)
Implementation
mungegithub
Mungegithub polls GitHub, and "munges" things it finds, including issues and pull requests. It is stateful, in that restarting it means it loses track of which things it has munged at what time.
- feature:
gitrepos
- responsible for parsing OWNERS and OWNERS_ALIAS files
- if its
use-reviewersflag is set to false, approvers will also be reviewers - if its
enable-md-yamlflag is set,.mdfiles will also be parsed to see if they have embedded OWNERS content (this is only used by kubernetes.github.io) - used by other mungers to get the set of reviewers or approvers for a given path
- munger:
blunderbuss
- responsible for determining reviewers and assigning to them
- chooses from people in the deepest/closest OWNERS files to the code being changed
- weights its choice based on the magnitude of lines changed for each file
- randomly chooses to ensure the same people aren't chosen every time
- if its
blunderbuss-number-assigneesflag is unset, it will default to 2 assignees
- munger:
approval-handler
- responsible for adding the
approvedlabel once an approver for each of the required OWNERS files has/approve'd - responsible for commenting as required OWNERS files are satisfied
- responsible for removing outdated approval status comments
- full description of the algorithm
- responsible for adding the
- munger:
submit-queue
- responsible for merging PR's
- responsible for updating a GitHub status check explaining why a PR can't be merged (eg: a
missing
lgtmorapprovedlabel)
prow
Prow receives events from GitHub, and reacts to them. It is effectively stateless.
- plugin: lgtm
- responsible for adding the
lgtmlabel when a reviewer comments/lgtmon a PR - the PR author may not
/lgtmtheir own PR
- responsible for adding the
- plugin: assign
- responsible for assigning GitHub users in response to
/assigncomments on a PR - responsible for unassigning GitHub users in response to
/unassigncomments on a PR
- responsible for assigning GitHub users in response to
GitHub
GitHub provides a few integration points for our automation:
- Status Checks: our tests and submit queue use these to communicate whether a commit is OK
- Protected Branches: ensure that a branch cannot be merged unless all status checks are green
- Merge Button: the submit queue effectively uses an API driven click of this button
Updating OWNERS
OWNERS files should be regularly maintained.
We should strive to:
- grow the number of OWNERS files
- add new people to OWNERS files
- ensure OWNERS files only contain org members and repo collaborators
- ensure OWNERS files only contain people are actively contributing to or reviewing the code they own
- remove inactive people from OWNERS files
Bad examples of OWNERS usage:
- directories that lack OWNERS files, resulting in too many hitting root OWNERS
- OWNERS files that have a single person as both approver and reviewer
- OWNERS files that haven't been touched in over 6 months
- OWNERS files that have non-collaborators present
Good examples of OWNERS usage:
- team aliases are used that correspond to sigs
- there are more
reviewersthanapprovers - the
approversare not in thereviewerssection - OWNERS files that are regularly updated (at least once per release)