MergeQueue

Automated merge workflow
Introduction
MergeQueue is a configurable queue that manages the merge workflow for your GitHub repository. The Aviator bot uses GitHub labels to identify pull requests (PRs) that are ready to be merged and queues them. Once a PR has been queued, the Aviator bot pulls the latest base branch for each PR, runs and verifies the required status checks, and then merges the changes once the checks pass. 
Dashboard view
Why MergeQueue?
CI tools can run tests on every pull request when it's opened, as well as on every branch after it's pushed, but it may not be sufficient to avoid broken builds.
For example, if you have two pull requests that modify dependent code, the tests could pass on each pull request independently, but the build may still break after the merge.
You could configure GitHub to block pull requests that are not up-to-date with master/main to avoid this issue, but that is not a scalable solution as your team grows. This configuration means every individual engineer has to:
Update current branch with master.
Wait for the test to pass again.
Merge the pull request when it's done.
In case another pull request is merged before that, repeat the steps above.
How does it work?
Aviator provides several merge modes that you can optimize for your team's repositories. All modes follow the same basic flow:
1.Aviator monitors all pull requests on your GitHub repository.
2.Instead of manually merging pull requests, the engineers comment /aviator merge or add a GitHub label when they are ready to merge their PR.
3.Based on your configurations, Aviator performs the required operations on ready-to-merge PRs.
4.Aviator automatically merges the PR when all the merge criteria has been met.
5.Aviator reports errors and dequeues the PRs that fail the criteria.
MergeQueue automatically dequeues PRs and reports build failures.
Modes
Default - The default mode uses a simple FIFO queue. PRs will be merged in the order they are queued. Aviator bot pulls latest changes from mainline to validate CI before merging.
​Parallel mode - CI builds are run optimistically in parallel in order to decrease time-to-merge for high output teams.
​Fast Forwarding - This works similar to parallel mode, but keeps branch history linear and avoids creating extra commits.
​Affected targets - Use MergeQueue as a dynamic sub-queue based on the build targets.
​ChangeSets - PRs can be merged together in a user-defined set.
Status Checks
Aviator automatically captures both required and non-required the status checks from your repository. By default, Aviator will validate the required status checks configured in GitHub. To learn more about status checks, checkout customizing required checks section.
Introduction
Quick Setup
Last modified 1mo ago