-
Notifications
You must be signed in to change notification settings - Fork 9.9k
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
Create a nightly job for linearizability tests #14911
Conversation
make build | ||
mkdir -p /tmp/linearizability | ||
cat server/etcdserver/raft.fail.go | ||
EXPECT_DEBUG=true GO_TEST_FLAGS='-v --count 6000 --failfast --run TestLinearizability' RESULTS_DIR=/tmp/linearizability make test-linearizability |
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.
6000 might be an overkill :P.
60 runs take up to 20 minutes, so 500 should be enough.
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.
haha fair enough
Start with a simple job against main that runs for 3h by repeating it an order of magnitude more often than the PR job. Signed-off-by: Thomas Jungblut <tjungblu@redhat.com>
on: | ||
# schedules always run against the main branch | ||
schedule: | ||
- cron: '25 9 * * *' |
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.
Can you add a comment explaining the cron.
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.
will add on the next PR for testing the release branches
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.
Thanks @tjungblu . The only concern is the Linearizability is still under heavily development, so it may be flaky. I would suggest holding on enabling it or run it weekly instead of daily. WDYT? cc @ptabor @serathius @spzala |
That's why enabling periodics is super useful, it allow us to qualify both etcd and linearizability tests while not blocking any PRs.
Don't see much impact of running daily, benefit would be faster feedback loop. If we are worried about occupying github action workers we might decide to move the job to CNCF provided larger workers. |
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.
The main concern is maintainers' and some contributors' bandwidth. Hopefully it will NOT add too much additional overhead.
Start with a simple job against main that runs for 3h by repeating it more often than the PR job.
Signed-off-by: Thomas Jungblut tjungblu@redhat.com
fix for #14045
I didn't find a much better way to make it run for longer, I basically just incremented the --count by two orders of manitude. Let me know if there's a nicer way, happy to implement that.
The caveat with the schedule is that this is only for the main branch, we have to maybe split the jobs up to checkout the respective release branches in parallel and run the tests.