-
Notifications
You must be signed in to change notification settings - Fork 3
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
SubmitRayJob enhancements #50
Conversation
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #50 +/- ##
==========================================
+ Coverage 97.12% 97.80% +0.68%
==========================================
Files 5 5
Lines 521 546 +25
==========================================
+ Hits 506 534 +28
+ Misses 15 12 -3 ☔ View full report in Codecov by Sentry. |
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.
@venkatajagannath thanks for advancing on this and addressing customer's feedback!
Could you highlight in the PR description if/ what are the breaking changes of this PR?
Also, assuming there are breaking changes, would it make sense to release an alpha with this change (e.g. 0.2.0a1), so we can iterate with the customers on the desired interface/feature quickly? We can have as many alpha releases as we want without impacting other people
@tatiana There are no breaking changes in the release. It just simplifies cluster spin up and down. So in a way its actually an enhancement to an existing operator. Sure, we can do an alpha release. Will update the PR description. |
If this is urgent you can also cut alpha from this branch |
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.
LGTM, There are minor breaking changes in theory we have documented them in the Changelog.
This PR contains the following changes -
Enhancements to the SubmitRayJob operator
Based on customer feedback, we learnt that it would be a much easier UX to spin up/down the cluster in the background of a task. The user would simply decorate their python function with @ray.task and the decorator would orchestrate the rest.
To enable this feature, we had to make changes to the code for SetupRayCluster and DeleteRayCluster operators. Making these changes helps us avoid duplication.
Add more more example DAGs
Earlier we had only 2 example dags. We now have 4. And we execute a different DAG for integration test.
Making the Decorator more robust
We made some changes to the decorator source code to make it more robust
Unit tests updated
Added unit tests where necessary and deleted where unnecessary. Updated where required.
Documentation improvements
Note: This PR does not contain any breaking changes.