[Core] (Resource Isolation 1/n) Unify resource isolation config construction into a single step by Kunchd · Pull Request #59372 · ray-project/ray

Kunchd · 2025-12-11T02:51:07Z

Description

The existing resource_isolation_config is constructed in two steps.

The class is first instantiated using the constructor.
The construction is complete when add_object_store_memory is called.
This is undesirable, as forgetting to complete the second step in creating the resource_isolation_config can result in critical system misbehavior.

This PR addresses the TODO to unify the construction of resource isolation config.

Related issues

#57653

Additional information

N/A

gemini-code-assist

Code Review

This pull request successfully unifies the construction of ResourceIsolationConfig into a single, more robust step. The refactoring of object store memory calculation into a utility function is a good improvement. The changes are well-implemented and the tests are updated accordingly. I have a couple of minor suggestions to improve the code further.

python/ray/tests/resource_isolation/test_resource_isolation_integration.py

python/ray/_private/resource_isolation_config.py

python/ray/tests/resource_isolation/test_resource_isolation_integration.py

python/ray/_private/resource_isolation_config.py

python/ray/_private/utils.py

israbbani

Changes look good. A few comments around code organization. Let's clean up the get_configured_object_store_memory method in a follow up.

python/ray/_private/resource_isolation_config.py

python/ray/_private/utils.py

python/ray/_private/resource_and_label_spec.py

Signed-off-by: davik <davik@anyscale.com>

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com> Signed-off-by: Kunchen (David) Dai <54918178+Kunchd@users.noreply.github.com>

…s to make sure it is only called once Signed-off-by: davik <davik@anyscale.com>

python/ray/_private/utils.py

python/ray/_private/resource_isolation_config.py

github-actions · 2025-12-31T12:25:39Z

This pull request has been automatically marked as stale because it has not had
any activity for 14 days. It will be closed in another 14 days if no further activity occurs.
Thank you for your contributions.

You can always ask for help on our discussion forum or Ray's public slack channel.

If you'd like to keep this open, just leave any comment, and the stale label will be removed.

israbbani

LGTM. Thanks for cleaning this up.

We've still got work to do here so can you convert your JIRA into an issue and link the issue to relevant parts of the codebase with TODOs so we don't lose context?

python/ray/_private/resource_and_label_spec.py

python/ray/_private/resource_isolation_config.py

python/ray/_private/utils.py

Kunchd · 2026-02-02T23:13:59Z

Issue to track follow up tech debts exposed by this PR: #60683

Signed-off-by: davik <davik@anyscale.com>

python/ray/_private/resource_and_label_spec.py

python/ray/_private/resource_isolation_config.py

python/ray/_private/utils.py

python/ray/tests/test_utils.py

Signed-off-by: davik <davik@anyscale.com>

israbbani

🚢

python/ray/_private/worker.py

Signed-off-by: davik <davik@anyscale.com>

Signed-off-by: Kunchen (David) Dai <54918178+Kunchd@users.noreply.github.com>

cursor

Cursor Bugbot has reviewed your changes and found 1 potential issue.

cursor · 2026-02-04T19:21:04Z

python/ray/_private/worker.py

+
+        resource_isolation_config = ResourceIsolationConfig(
+            enable_resource_isolation=enable_resource_isolation,
+            cgroup_path=_cgroup_path,


Undefined variable _cgroup_path causes NameError

High Severity

The code references _cgroup_path but this variable is never defined. The function parameter is cgroup_path (without underscore, defined at line 1435), but the refactored code incorrectly uses _cgroup_path. This will cause a NameError at runtime when ray.init() is called on the code path that starts a new local cluster.

Signed-off-by: davik <davik@anyscale.com>

Kunchd · 2026-02-04T22:33:38Z

@edoakes Could you help me merge this when you get the chance? Thanks!

…ruction into a single step (ray-project#59372) ## Description The existing `resource_isolation_config` is constructed in two steps. 1. The class is first instantiated using the constructor. 2. The construction is complete when `add_object_store_memory` is called. This is undesirable, as forgetting to complete the second step in creating the `resource_isolation_config` can result in critical system misbehavior. This PR addresses the [TODO](https://github.com/ray-project/ray/blob/b13361a8045632a7c2308ce606b0487f096fd927/python/ray/_private/resource_isolation_config.py#L54-L56) to unify the construction of resource isolation config. ## Related issues ray-project#57653 ## Additional information N/A --------- Signed-off-by: davik <davik@anyscale.com> Signed-off-by: Kunchen (David) Dai <54918178+Kunchd@users.noreply.github.com> Co-authored-by: davik <davik@anyscale.com> Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com> Signed-off-by: tiennguyentony <46289799+tiennguyentony@users.noreply.github.com>

…ruction into a single step (ray-project#59372) ## Description The existing `resource_isolation_config` is constructed in two steps. 1. The class is first instantiated using the constructor. 2. The construction is complete when `add_object_store_memory` is called. This is undesirable, as forgetting to complete the second step in creating the `resource_isolation_config` can result in critical system misbehavior. This PR addresses the [TODO](https://github.com/ray-project/ray/blob/b13361a8045632a7c2308ce606b0487f096fd927/python/ray/_private/resource_isolation_config.py#L54-L56) to unify the construction of resource isolation config. ## Related issues ray-project#57653 ## Additional information N/A --------- Signed-off-by: davik <davik@anyscale.com> Signed-off-by: Kunchen (David) Dai <54918178+Kunchd@users.noreply.github.com> Co-authored-by: davik <davik@anyscale.com> Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

…ruction into a single step (#59372) ## Description The existing `resource_isolation_config` is constructed in two steps. 1. The class is first instantiated using the constructor. 2. The construction is complete when `add_object_store_memory` is called. This is undesirable, as forgetting to complete the second step in creating the `resource_isolation_config` can result in critical system misbehavior. This PR addresses the [TODO](https://github.com/ray-project/ray/blob/b13361a8045632a7c2308ce606b0487f096fd927/python/ray/_private/resource_isolation_config.py#L54-L56) to unify the construction of resource isolation config. ## Related issues #57653 ## Additional information N/A --------- Signed-off-by: davik <davik@anyscale.com> Signed-off-by: Kunchen (David) Dai <54918178+Kunchd@users.noreply.github.com> Co-authored-by: davik <davik@anyscale.com> Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com> Signed-off-by: elliot-barn <elliot.barnwell@anyscale.com>

…ruction into a single step (#59372) ## Description The existing `resource_isolation_config` is constructed in two steps. 1. The class is first instantiated using the constructor. 2. The construction is complete when `add_object_store_memory` is called. This is undesirable, as forgetting to complete the second step in creating the `resource_isolation_config` can result in critical system misbehavior. This PR addresses the [TODO](https://github.com/ray-project/ray/blob/b13361a8045632a7c2308ce606b0487f096fd927/python/ray/_private/resource_isolation_config.py#L54-L56) to unify the construction of resource isolation config. ## Related issues #57653 ## Additional information N/A --------- Signed-off-by: davik <davik@anyscale.com> Signed-off-by: Kunchen (David) Dai <54918178+Kunchd@users.noreply.github.com> Co-authored-by: davik <davik@anyscale.com> Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

…ruction into a single step (ray-project#59372) ## Description The existing `resource_isolation_config` is constructed in two steps. 1. The class is first instantiated using the constructor. 2. The construction is complete when `add_object_store_memory` is called. This is undesirable, as forgetting to complete the second step in creating the `resource_isolation_config` can result in critical system misbehavior. This PR addresses the [TODO](https://github.com/ray-project/ray/blob/b13361a8045632a7c2308ce606b0487f096fd927/python/ray/_private/resource_isolation_config.py#L54-L56) to unify the construction of resource isolation config. ## Related issues ray-project#57653 ## Additional information N/A --------- Signed-off-by: davik <davik@anyscale.com> Signed-off-by: Kunchen (David) Dai <54918178+Kunchd@users.noreply.github.com> Co-authored-by: davik <davik@anyscale.com> Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

…ruction into a single step (ray-project#59372) ## Description The existing `resource_isolation_config` is constructed in two steps. 1. The class is first instantiated using the constructor. 2. The construction is complete when `add_object_store_memory` is called. This is undesirable, as forgetting to complete the second step in creating the `resource_isolation_config` can result in critical system misbehavior. This PR addresses the [TODO](https://github.com/ray-project/ray/blob/b13361a8045632a7c2308ce606b0487f096fd927/python/ray/_private/resource_isolation_config.py#L54-L56) to unify the construction of resource isolation config. ## Related issues ray-project#57653 ## Additional information N/A --------- Signed-off-by: davik <davik@anyscale.com> Signed-off-by: Kunchen (David) Dai <54918178+Kunchd@users.noreply.github.com> Co-authored-by: davik <davik@anyscale.com> Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com> Signed-off-by: Adel Nour <ans9868@nyu.edu>

…ruction into a single step (ray-project#59372) ## Description The existing `resource_isolation_config` is constructed in two steps. 1. The class is first instantiated using the constructor. 2. The construction is complete when `add_object_store_memory` is called. This is undesirable, as forgetting to complete the second step in creating the `resource_isolation_config` can result in critical system misbehavior. This PR addresses the [TODO](https://github.com/ray-project/ray/blob/b13361a8045632a7c2308ce606b0487f096fd927/python/ray/_private/resource_isolation_config.py#L54-L56) to unify the construction of resource isolation config. ## Related issues ray-project#57653 ## Additional information N/A --------- Signed-off-by: davik <davik@anyscale.com> Signed-off-by: Kunchen (David) Dai <54918178+Kunchd@users.noreply.github.com> Co-authored-by: davik <davik@anyscale.com> Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

…ruction into a single step (ray-project#59372) ## Description The existing `resource_isolation_config` is constructed in two steps. 1. The class is first instantiated using the constructor. 2. The construction is complete when `add_object_store_memory` is called. This is undesirable, as forgetting to complete the second step in creating the `resource_isolation_config` can result in critical system misbehavior. This PR addresses the [TODO](https://github.com/ray-project/ray/blob/b13361a8045632a7c2308ce606b0487f096fd927/python/ray/_private/resource_isolation_config.py#L54-L56) to unify the construction of resource isolation config. ## Related issues ray-project#57653 ## Additional information N/A --------- Signed-off-by: davik <davik@anyscale.com> Signed-off-by: Kunchen (David) Dai <54918178+Kunchd@users.noreply.github.com> Co-authored-by: davik <davik@anyscale.com> Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com> Signed-off-by: peterxcli <peterxcli@gmail.com>

Kunchd requested a review from a team as a code owner December 11, 2025 02:51

gemini-code-assist bot reviewed Dec 11, 2025

View reviewed changes

python/ray/tests/resource_isolation/test_resource_isolation_integration.py Outdated Show resolved Hide resolved

python/ray/_private/resource_isolation_config.py Outdated Show resolved Hide resolved

cursor bot reviewed Dec 11, 2025

View reviewed changes

python/ray/tests/resource_isolation/test_resource_isolation_integration.py Outdated Show resolved Hide resolved

ray-gardener bot added the core Issues that should be addressed in Ray Core label Dec 11, 2025

Kunchd added the go add ONLY when ready to merge, run all tests label Dec 11, 2025

Kunchd force-pushed the unify_ri_config branch from 22bf1c1 to 0b94939 Compare December 11, 2025 21:24

Kunchd changed the title ~~Unify resource isolation config construction into a single step~~ [Core] Unify resource isolation config construction into a single step Dec 11, 2025

dayshah reviewed Dec 12, 2025

View reviewed changes

python/ray/_private/resource_isolation_config.py Show resolved Hide resolved

python/ray/_private/utils.py Outdated Show resolved Hide resolved

israbbani reviewed Dec 15, 2025

View reviewed changes

davik and others added 7 commits December 17, 2025 00:40

Unify resource isolation config construction into a single step

d0d8e71

Signed-off-by: davik <davik@anyscale.com>

ensure integration test runs in the right environment

496236c

Signed-off-by: davik <davik@anyscale.com>

Move attributes back

455fe3e

Signed-off-by: davik <davik@anyscale.com>

Unify resource isolation config construction into a single step

7660c69

Signed-off-by: davik <davik@anyscale.com>

ensure integration test runs in the right environment

1e695cf

Signed-off-by: davik <davik@anyscale.com>

Update python/ray/_private/resource_isolation_config.py

39a8083

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com> Signed-off-by: Kunchen (David) Dai <54918178+Kunchd@users.noreply.github.com>

Move available memory estimation to start of initialization code path…

84682b1

…s to make sure it is only called once Signed-off-by: davik <davik@anyscale.com>

Kunchd force-pushed the unify_ri_config branch from 39ed8d3 to 84682b1 Compare December 17, 2025 00:40

cursor bot reviewed Dec 17, 2025

View reviewed changes

python/ray/_private/utils.py Outdated Show resolved Hide resolved

python/ray/_private/resource_isolation_config.py Show resolved Hide resolved

github-actions bot added the stale The issue is stale. It will be closed within 7 days unless there are further conversation label Dec 31, 2025

Kunchd added unstale A PR that has been marked unstale. It will not get marked stale again if this label is on it. and removed stale The issue is stale. It will be closed within 7 days unless there are further conversation labels Jan 2, 2026

israbbani reviewed Jan 5, 2026

View reviewed changes

Merge branch 'master' into unify_ri_config

b4ae719

Kunchd mentioned this pull request Feb 2, 2026

[Core] Prune out-of-date logic to resolve default object_store_memory size #60683

Open

7 tasks

Add resolve_object_store_memory tests and improve comment clarity

caeea5a

Signed-off-by: davik <davik@anyscale.com>

israbbani reviewed Feb 3, 2026

View reviewed changes

Improve test name, comment, variable clarity

6cefd4d

Signed-off-by: davik <davik@anyscale.com>

israbbani approved these changes Feb 3, 2026

View reviewed changes

cursor bot reviewed Feb 3, 2026

View reviewed changes

python/ray/_private/worker.py Outdated Show resolved Hide resolved

davik and others added 2 commits February 3, 2026 01:08

Fix estimate_available_memory typo

7f979eb

Signed-off-by: davik <davik@anyscale.com>

Merge branch 'master' into unify_ri_config

4afdb14

Signed-off-by: Kunchen (David) Dai <54918178+Kunchd@users.noreply.github.com>

cursor bot reviewed Feb 4, 2026

View reviewed changes

Kunchd changed the title ~~[Core] Unify resource isolation config construction into a single step~~ [Core] (Resource Isolation 1/n) Unify resource isolation config construction into a single step Feb 4, 2026

Fix merge typo

a1aa625

Signed-off-by: davik <davik@anyscale.com>

edoakes merged commit 7bf368c into ray-project:master Feb 5, 2026
6 checks passed

Conversation

Kunchd commented Dec 11, 2025

Description

Related issues

Additional information

Uh oh!

gemini-code-assist bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

israbbani left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

github-actions bot commented Dec 31, 2025

Uh oh!

israbbani left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Kunchd commented Feb 2, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

israbbani left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

cursor bot left a comment

Choose a reason for hiding this comment

Uh oh!

cursor bot Feb 4, 2026

Choose a reason for hiding this comment

Undefined variable _cgroup_path causes NameError

Uh oh!

Kunchd commented Feb 4, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

Undefined variable `_cgroup_path` causes NameError