-
Notifications
You must be signed in to change notification settings - Fork 641
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.
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
ci: stabilize workflows #10457
ci: stabilize workflows #10457
Conversation
b0289c5
to
8376f28
Compare
- name: Create build output log file | ||
run: echo "BUILD_OUTPUT_FILE_PATH=$(mktemp)" >> $GITHUB_ENV | ||
- name: Maven Test Build | ||
run: > | ||
mvn -B --no-snapshot-updates | ||
-D skipITs -D skipChecks | ||
-pl ${{ matrix.project }} | ||
-D junitThreadCount=16 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
assuming there are maven modules that can run in parallel would adding -T1C
or less eager a static value of 4/8 be worth it for speed-up?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Maybe, yeah 👍 I've tried to max out CPU via JUnit first but RandomizedRaftTest
is taking 5 minutes and blocking progress for all others.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Done, it removed around 3 minutes
20a7a67
to
0640208
Compare
0640208
to
cfffca1
Compare
Can we have a short chat in the TR about solutions? Generally some of these are good ideas, but I have some ideas/suggestions (e.g. split unit tests across a fixed number of groups), and questions (e.g. why is Testcontainers Cloud a source of flakiness?). |
Some potential alternative:
|
87c6954
to
efda33a
Compare
Modifies the concurrency of the code quality, test, and deploy workflows. The deploy workflow is sequenced such that every deploy is ran only after the previous push's deploy run is finished. This will avoid having concurrent deployments which could overwrite each other or mess up the ordering, ensuring the latest SNAPSHOT is from the latest successful commit. For other workflows, these will run and cancel previous runs, such that only one run at the same time is running. Hopefully this cuts down on resource usage and perhaps auto-scaling issues.
efda33a
to
f766cf4
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
yolo
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
looks good to me let's see it in action 🤞
bors r+ |
bors r- already merged |
Canceled. |
This PR combines multiple improvements and has two goals:
To achieve the first goal of stabilizing CI:
To improve maintainability:
This has a couple of drawbacks: