Ray Train test_jax_trainer::test_minimal_multihost Flaky Test Fix #56548
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Signed-off-by: JasonLi1909 <jasli1909@gmail.com>
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Code Review
This pull request aims to fix a flaky test, test_jax_trainer.py::test_minimal_multihost, by moving the worker_runtime_env from the RunConfig of the JaxTrainer to the runtime_env of ray.init in the test fixtures. This change is sound, but I've found a critical issue in the implementation for the multi-host test fixture where runtime_env is incorrectly defined as a tuple instead of a dictionary. I've also included a suggestion to reduce code duplication for better maintainability.
Signed-off-by: JasonLi1909 <jasli1909@gmail.com>
![cursor[bot]](https://avatars.githubusercontent.com/in/1210556?s=60&v=4){kind=small}
liulehui
approved these changes
Sep 22, 2025
JasonLi1909
added
go
ZacAttack
pushed a commit
to ZacAttack/ray
that referenced
this pull request
Sep 24, 2025
…y-project#56548) A fix that addresses the failing flaky test `test_jax_trainer.py::test_minimal_multihost`. https://buildkite.com/ray-project/postmerge/builds/12941#01993f89-cc62-4e31-8de2-8b18f81ac177 Issue: The `test_minimal_multihost` introduces a race condition by attempting to initialize a virtualenv directory twice at the same directory path during worker runtime environment setup. This test would not fail in a true multi-host environment, but the tests simulate a multi-host environment on a singular device. This might be a ray core issue resulting in errors on runtime _env, but this PR will at least unblock the test so it is no longer flaky. Fix: Move `worker_runtime_env` to the job level so that the `pip install jax` only happens once --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: zac <zac@anyscale.com>
elliot-barn
pushed a commit
that referenced
this pull request
Sep 24, 2025
…6548) A fix that addresses the failing flaky test `test_jax_trainer.py::test_minimal_multihost`. https://buildkite.com/ray-project/postmerge/builds/12941#01993f89-cc62-4e31-8de2-8b18f81ac177 Issue: The `test_minimal_multihost` introduces a race condition by attempting to initialize a virtualenv directory twice at the same directory path during worker runtime environment setup. This test would not fail in a true multi-host environment, but the tests simulate a multi-host environment on a singular device. This might be a ray core issue resulting in errors on runtime _env, but this PR will at least unblock the test so it is no longer flaky. Fix: Move `worker_runtime_env` to the job level so that the `pip install jax` only happens once --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: elliot-barn <elliot.barnwell@anyscale.com>
marcostephan
pushed a commit
to marcostephan/ray
that referenced
this pull request
Sep 24, 2025
…y-project#56548) A fix that addresses the failing flaky test `test_jax_trainer.py::test_minimal_multihost`. https://buildkite.com/ray-project/postmerge/builds/12941#01993f89-cc62-4e31-8de2-8b18f81ac177 Issue: The `test_minimal_multihost` introduces a race condition by attempting to initialize a virtualenv directory twice at the same directory path during worker runtime environment setup. This test would not fail in a true multi-host environment, but the tests simulate a multi-host environment on a singular device. This might be a ray core issue resulting in errors on runtime _env, but this PR will at least unblock the test so it is no longer flaky. Fix: Move `worker_runtime_env` to the job level so that the `pip install jax` only happens once --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: Marco Stephan <marco@magic.dev>
elliot-barn
pushed a commit
that referenced
this pull request
Sep 27, 2025
…6548) A fix that addresses the failing flaky test `test_jax_trainer.py::test_minimal_multihost`. https://buildkite.com/ray-project/postmerge/builds/12941#01993f89-cc62-4e31-8de2-8b18f81ac177 Issue: The `test_minimal_multihost` introduces a race condition by attempting to initialize a virtualenv directory twice at the same directory path during worker runtime environment setup. This test would not fail in a true multi-host environment, but the tests simulate a multi-host environment on a singular device. This might be a ray core issue resulting in errors on runtime _env, but this PR will at least unblock the test so it is no longer flaky. Fix: Move `worker_runtime_env` to the job level so that the `pip install jax` only happens once --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: elliot-barn <elliot.barnwell@anyscale.com>
matthewdeng
pushed a commit
that referenced
this pull request
Oct 3, 2025
Revisiting #56548 as test continues to be flaky on CI **Solution**: The previous attempt to deflake this test still used a `pip install jax` via the `ray.init` runtime_env args. Hence, the pip install related error persisted. This PR instead adds `jax` and `jaxlib` as a dependency of CI train tests, avoiding the need to `pip install jax` via the runtime_env. --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: Jason Li <57246540+JasonLi1909@users.noreply.github.com>
dstrodtman
pushed a commit
to dstrodtman/ray
that referenced
this pull request
Oct 6, 2025
…y-project#56548) A fix that addresses the failing flaky test `test_jax_trainer.py::test_minimal_multihost`. https://buildkite.com/ray-project/postmerge/builds/12941#01993f89-cc62-4e31-8de2-8b18f81ac177 Issue: The `test_minimal_multihost` introduces a race condition by attempting to initialize a virtualenv directory twice at the same directory path during worker runtime environment setup. This test would not fail in a true multi-host environment, but the tests simulate a multi-host environment on a singular device. This might be a ray core issue resulting in errors on runtime _env, but this PR will at least unblock the test so it is no longer flaky. Fix: Move `worker_runtime_env` to the job level so that the `pip install jax` only happens once --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: Douglas Strodtman <douglas@anyscale.com>
dstrodtman
pushed a commit
to dstrodtman/ray
that referenced
this pull request
Oct 6, 2025
Revisiting ray-project#56548 as test continues to be flaky on CI **Solution**: The previous attempt to deflake this test still used a `pip install jax` via the `ray.init` runtime_env args. Hence, the pip install related error persisted. This PR instead adds `jax` and `jaxlib` as a dependency of CI train tests, avoiding the need to `pip install jax` via the runtime_env. --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: Jason Li <57246540+JasonLi1909@users.noreply.github.com> Signed-off-by: Douglas Strodtman <douglas@anyscale.com>
eicherseiji
pushed a commit
to eicherseiji/ray
that referenced
this pull request
Oct 6, 2025
Revisiting ray-project#56548 as test continues to be flaky on CI **Solution**: The previous attempt to deflake this test still used a `pip install jax` via the `ray.init` runtime_env args. Hence, the pip install related error persisted. This PR instead adds `jax` and `jaxlib` as a dependency of CI train tests, avoiding the need to `pip install jax` via the runtime_env. --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: Jason Li <57246540+JasonLi1909@users.noreply.github.com> Signed-off-by: Seiji Eicher <seiji@anyscale.com>
eicherseiji
pushed a commit
to eicherseiji/ray
that referenced
this pull request
Oct 6, 2025
Revisiting ray-project#56548 as test continues to be flaky on CI **Solution**: The previous attempt to deflake this test still used a `pip install jax` via the `ray.init` runtime_env args. Hence, the pip install related error persisted. This PR instead adds `jax` and `jaxlib` as a dependency of CI train tests, avoiding the need to `pip install jax` via the runtime_env. --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: Jason Li <57246540+JasonLi1909@users.noreply.github.com>
eicherseiji
pushed a commit
to eicherseiji/ray
that referenced
this pull request
Oct 6, 2025
Revisiting ray-project#56548 as test continues to be flaky on CI **Solution**: The previous attempt to deflake this test still used a `pip install jax` via the `ray.init` runtime_env args. Hence, the pip install related error persisted. This PR instead adds `jax` and `jaxlib` as a dependency of CI train tests, avoiding the need to `pip install jax` via the runtime_env. --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: Jason Li <57246540+JasonLi1909@users.noreply.github.com>
eicherseiji
pushed a commit
to eicherseiji/ray
that referenced
this pull request
Oct 6, 2025
Revisiting ray-project#56548 as test continues to be flaky on CI **Solution**: The previous attempt to deflake this test still used a `pip install jax` via the `ray.init` runtime_env args. Hence, the pip install related error persisted. This PR instead adds `jax` and `jaxlib` as a dependency of CI train tests, avoiding the need to `pip install jax` via the runtime_env. --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: Jason Li <57246540+JasonLi1909@users.noreply.github.com>
eicherseiji
pushed a commit
to eicherseiji/ray
that referenced
this pull request
Oct 6, 2025
Revisiting ray-project#56548 as test continues to be flaky on CI **Solution**: The previous attempt to deflake this test still used a `pip install jax` via the `ray.init` runtime_env args. Hence, the pip install related error persisted. This PR instead adds `jax` and `jaxlib` as a dependency of CI train tests, avoiding the need to `pip install jax` via the runtime_env. --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: Jason Li <57246540+JasonLi1909@users.noreply.github.com>
liulehui
pushed a commit
to liulehui/ray
that referenced
this pull request
Oct 9, 2025
Revisiting ray-project#56548 as test continues to be flaky on CI **Solution**: The previous attempt to deflake this test still used a `pip install jax` via the `ray.init` runtime_env args. Hence, the pip install related error persisted. This PR instead adds `jax` and `jaxlib` as a dependency of CI train tests, avoiding the need to `pip install jax` via the runtime_env. --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: Jason Li <57246540+JasonLi1909@users.noreply.github.com>
joshkodi
pushed a commit
to joshkodi/ray
that referenced
this pull request
Oct 13, 2025
Revisiting ray-project#56548 as test continues to be flaky on CI **Solution**: The previous attempt to deflake this test still used a `pip install jax` via the `ray.init` runtime_env args. Hence, the pip install related error persisted. This PR instead adds `jax` and `jaxlib` as a dependency of CI train tests, avoiding the need to `pip install jax` via the runtime_env. --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: Jason Li <57246540+JasonLi1909@users.noreply.github.com> Signed-off-by: Josh Kodi <joshkodi@gmail.com>
justinyeh1995
pushed a commit
to justinyeh1995/ray
that referenced
this pull request
Oct 20, 2025
…y-project#56548) A fix that addresses the failing flaky test `test_jax_trainer.py::test_minimal_multihost`. https://buildkite.com/ray-project/postmerge/builds/12941#01993f89-cc62-4e31-8de2-8b18f81ac177 Issue: The `test_minimal_multihost` introduces a race condition by attempting to initialize a virtualenv directory twice at the same directory path during worker runtime environment setup. This test would not fail in a true multi-host environment, but the tests simulate a multi-host environment on a singular device. This might be a ray core issue resulting in errors on runtime _env, but this PR will at least unblock the test so it is no longer flaky. Fix: Move `worker_runtime_env` to the job level so that the `pip install jax` only happens once --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com>
justinyeh1995
pushed a commit
to justinyeh1995/ray
that referenced
this pull request
Oct 20, 2025
Revisiting ray-project#56548 as test continues to be flaky on CI **Solution**: The previous attempt to deflake this test still used a `pip install jax` via the `ray.init` runtime_env args. Hence, the pip install related error persisted. This PR instead adds `jax` and `jaxlib` as a dependency of CI train tests, avoiding the need to `pip install jax` via the runtime_env. --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: Jason Li <57246540+JasonLi1909@users.noreply.github.com>
landscapepainter
pushed a commit
to landscapepainter/ray
that referenced
this pull request
Nov 17, 2025
…y-project#56548) A fix that addresses the failing flaky test `test_jax_trainer.py::test_minimal_multihost`. https://buildkite.com/ray-project/postmerge/builds/12941#01993f89-cc62-4e31-8de2-8b18f81ac177 Issue: The `test_minimal_multihost` introduces a race condition by attempting to initialize a virtualenv directory twice at the same directory path during worker runtime environment setup. This test would not fail in a true multi-host environment, but the tests simulate a multi-host environment on a singular device. This might be a ray core issue resulting in errors on runtime _env, but this PR will at least unblock the test so it is no longer flaky. Fix: Move `worker_runtime_env` to the job level so that the `pip install jax` only happens once --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com>
landscapepainter
pushed a commit
to landscapepainter/ray
that referenced
this pull request
Nov 17, 2025
Revisiting ray-project#56548 as test continues to be flaky on CI **Solution**: The previous attempt to deflake this test still used a `pip install jax` via the `ray.init` runtime_env args. Hence, the pip install related error persisted. This PR instead adds `jax` and `jaxlib` as a dependency of CI train tests, avoiding the need to `pip install jax` via the runtime_env. --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: Jason Li <57246540+JasonLi1909@users.noreply.github.com>
Aydin-ab
pushed a commit
to Aydin-ab/ray-aydin
that referenced
this pull request
Nov 19, 2025
Revisiting ray-project#56548 as test continues to be flaky on CI **Solution**: The previous attempt to deflake this test still used a `pip install jax` via the `ray.init` runtime_env args. Hence, the pip install related error persisted. This PR instead adds `jax` and `jaxlib` as a dependency of CI train tests, avoiding the need to `pip install jax` via the runtime_env. --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: Jason Li <57246540+JasonLi1909@users.noreply.github.com> Signed-off-by: Aydin Abiar <aydin@anyscale.com>
Future-Outlier
pushed a commit
to Future-Outlier/ray
that referenced
this pull request
Dec 7, 2025
…y-project#56548) A fix that addresses the failing flaky test `test_jax_trainer.py::test_minimal_multihost`. https://buildkite.com/ray-project/postmerge/builds/12941#01993f89-cc62-4e31-8de2-8b18f81ac177 Issue: The `test_minimal_multihost` introduces a race condition by attempting to initialize a virtualenv directory twice at the same directory path during worker runtime environment setup. This test would not fail in a true multi-host environment, but the tests simulate a multi-host environment on a singular device. This might be a ray core issue resulting in errors on runtime _env, but this PR will at least unblock the test so it is no longer flaky. Fix: Move `worker_runtime_env` to the job level so that the `pip install jax` only happens once --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: Future-Outlier <eric901201@gmail.com>
Future-Outlier
pushed a commit
to Future-Outlier/ray
that referenced
this pull request
Dec 7, 2025
Revisiting ray-project#56548 as test continues to be flaky on CI **Solution**: The previous attempt to deflake this test still used a `pip install jax` via the `ray.init` runtime_env args. Hence, the pip install related error persisted. This PR instead adds `jax` and `jaxlib` as a dependency of CI train tests, avoiding the need to `pip install jax` via the runtime_env. --------- Signed-off-by: JasonLi1909 <jasli1909@gmail.com> Signed-off-by: Jason Li <57246540+JasonLi1909@users.noreply.github.com> Signed-off-by: Future-Outlier <eric901201@gmail.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
A fix that addresses the failing flaky test
test_jax_trainer.py::test_minimal_multihost.https://buildkite.com/ray-project/postmerge/builds/12941#01993f89-cc62-4e31-8de2-8b18f81ac177
Issue:
The
test_minimal_multihostintroduces a race condition by attempting to initialize a virtualenv directory twice at the same directory path during worker runtime environment setup. This test would not fail in a true multi-host environment, but the tests simulate a multi-host environment on a singular device. This might be a ray core issue resulting in errors on runtime _env, but this PR will at least unblock the test so it is no longer flaky.Fix:
Move
worker_runtime_envto the job level so that thepip install jaxonly happens once