[Ray 2.9.0 Release] Update Ray versions from 2.8.0 to 2.9.0 #1770
Conversation
Signed-off-by: Archit Kulkarni <architkulkarni@users.noreply.github.com>
…ray-2.9.0-update Signed-off-by: Archit Kulkarni <architkulkarni@users.noreply.github.com>
Signed-off-by: Archit Kulkarni <architkulkarni@users.noreply.github.com>
| @@ -151,9 +151,9 @@ def test_ray_serve(self): | |||
| self.fail(f"Fail to execute test_ray_serve_2.py. The exit code is {exit_code}.") | |||
|
|
|||
| @unittest.skipIf( | |||
| ray_version == '2.8.0' or ray_version == 'nightly', | |||
| 'test_detached_actor is too flaky with Ray 2.8.0 and Ray nightly.' | |||
| 'Therefore, skip it until https://github.com/ray-project/ray/issues/41343 is solved.' | |||
There was a problem hiding this comment.
This is the test that is unskipped in this PR.
There was a problem hiding this comment.
Why does Compatibility Test 2.8.0 still fail with test_detached_actor? It should be skipped.
There was a problem hiding this comment.
Not sure how this is possible, still investigating. According to the code it looks like we are still skipping it.
There was a problem hiding this comment.
I suspect it's because the decorators are evaluated when the function is defined, not at the first the time the function is executed. So when it's evaluated, the global variable ray_version is still set to its initial value of 2.9.0, so it doesn't get skipped.
Before this PR the initial value was 2.8.0, so the test was actually being skipped for all compatibility tests (2.5, 2.6, 2.7, 2.8, nightly). I'll just fix it directly in this PR.
There was a problem hiding this comment.
Fixed! 3b0718d
| with: | ||
| ray_version: 2.5.0 | ||
|
|
||
| test-compatibility-2_6_3: |
There was a problem hiding this comment.
Previously we were testing 2.5.0, 2.6.3, 2.7.0, and 2.8.0.
In this PR, we only test 2.7.0, 2.8.0, and 2.9.0.
| @@ -151,9 +151,9 @@ def test_ray_serve(self): | |||
| self.fail(f"Fail to execute test_ray_serve_2.py. The exit code is {exit_code}.") | |||
|
|
|||
| @unittest.skipIf( | |||
| ray_version == '2.8.0' or ray_version == 'nightly', | |||
| 'test_detached_actor is too flaky with Ray 2.8.0 and Ray nightly.' | |||
| 'Therefore, skip it until https://github.com/ray-project/ray/issues/41343 is solved.' | |||
There was a problem hiding this comment.
Why does Compatibility Test 2.8.0 still fail with test_detached_actor? It should be skipped.
Signed-off-by: Archit Kulkarni <architkulkarni@users.noreply.github.com>
…ray-2.9.0-update Signed-off-by: Archit Kulkarni <architkulkarni@users.noreply.github.com>
Signed-off-by: Archit Kulkarni <architkulkarni@users.noreply.github.com>
| @@ -29,7 +29,7 @@ | |||
| ) | |||
|
|
|||
| # Default Ray version | |||
| ray_version = '2.8.0' | |||
| ray_version = '2.9.0' | |||
There was a problem hiding this comment.
I found that @unittest.skipIf uses ray_version with the 2.9.0. We may remove ray_version here, and only initialize the global variable in __main__.
There was a problem hiding this comment.
The interesting thing is that when I print ray_version in setUpClass, the value is 2.8.0.
|
I am not a big fan of checking whether to skip tests at runtime. I have opened an issue to track the progress: #1773. This is a good first issue for contributors. |
Signed-off-by: cindyz <cindyz@anyscale.com>
Why are these changes needed?
As part of the 2.9.0 release, this PR updates the latest Ray version used in the KubeRay repo to 2.9.0. Example PR: #1678
DAGDriver, which was removed completely in Ray 2.9.0 (Thanks @zcin !)Related issue number
Checks