Skip to content

Conversation

@pzhan9
Copy link
Contributor

@pzhan9 pzhan9 commented Nov 14, 2025

Summary:

  1. Use name = "SupervisionEvent" so the supervision logs can be searched with this value;
  2. Change log levels: warn for error and debug for others.

Differential Revision: D87094538

@meta-codesync
Copy link

meta-codesync bot commented Nov 14, 2025

@pzhan9 has exported this pull request. If you are a Meta employee, you can view the originating Diff in D87094538.

@meta-cla meta-cla bot added the CLA Signed This label is managed by the Meta Open Source bot. label Nov 14, 2025
pzhan9 added a commit to pzhan9/monarch-1 that referenced this pull request Nov 17, 2025
Summary:

1. Use `name = "SupervisionEvent"` so the supervision logs can be searched with this value;
2. Change log levels: `warn` for error and `debug` for others.

Reviewed By: mariusae

Differential Revision: D87094538
Summary:

There is a bug in `Instance`'s current Drop impl:

https://www.internalfb.com/code/fbsource/[5b70b5b30638d5db5dfef1cbf929dea6df7d3f56]/fbcode/monarch/hyperactor/src/proc.rs?lines=1594-1596

since we clone a `Instance` object on the python side:

https://www.internalfb.com/code/fbsource/[5b70b5b30638d5db5dfef1cbf929dea6df7d3f56]/fbcode/monarch/hyperactor/src/proc.rs?lines=1566

It means the Instance's actor status will be changed to `Stopped` when a clone is dropped, although the actor is still alive.

This diff adds a `InstanceState` struct, so the instance's lifetime can tracked by `Arc`'s ref counter. Then we move the `Drop` impl to this  `InstanceState`, so it will only be called when the ref count goes to 0.

Reviewed By: dulinriley

Differential Revision: D87075455
Summary:

1. Use `name = "SupervisionEvent"` so the supervision logs can be searched with this value;
2. Change log levels: `warn` for error and `debug` for others.

Reviewed By: mariusae

Differential Revision: D87094538
@meta-codesync
Copy link

meta-codesync bot commented Nov 17, 2025

This pull request has been merged in deeaee9.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

CLA Signed This label is managed by the Meta Open Source bot. fb-exported Merged meta-exported

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants