Chen Shen comments

Results 61 comments of


                                            Chen Shen

[Core][WIP] allow worker lazily bind to job_id

ah this run_on_all_workers is annoying

[Core][WIP] allow worker lazily bind to job_id

most of test failures are around logging. we need to merge https://github.com/ray-project/ray/pull/31772 first

[core][autoscaler][v2] node providers - kuberay cloud provider[3/x]

better leave this to @kevin85421 and @DmitriGekhtman to review

[Core] Apparent object store memory leak

Thanks for reporting. I think there is some race condition between Ray GC and spill worker, so when the object has been GCed the spill worker already grab the object...

[RFC][Core] Add `get_actor(actor_id)` API to simplify applications.

sounds reasonable to me.

[serve] Add initial health check before marking a replica as RUNNING

@zcin this breaks bk://:java and bk://:mac: :apple: Ray C++ and Java

[Core][deprecate run_on_all_workers 1/n] set worker's sys.path through JobConfig._py_driver_sys_path

> Btw, could you also delete import thread which is not useful anymore with this work. Deletion will be done in https://github.com/ray-project/ray/pull/30895, since @rkooo567 has some concern on deprecation logistic.

[Core][deprecate run_on_all_workers 1/n] set worker's sys.path through JobConfig._py_driver_sys_path

> Thanks for fixing this. As synced offline, my only concern is that we make the code_search_path and the new introduced ones doing similar works. Maybe we should try to...

[Object Store] race condition: Pull Manager will hang in certain timings

Thanks for the detailed explaination. For my understanding, it looks like a bug that we still accepting chunks after step 9. Since at this point we should know the object...

[Object Store] race condition: Pull Manager will hang in certain timings

hmm i see. yeah let's implement proposal 1 given how simple it is?