Jialing He

@rkooo567 - windows://python/ray/tests:test_reference_counting: It seems to be related. After this PR, this test was slowed by about 80s, it cause timeout, I increased the timeout and it seems to work:(The...

@rkooo567 I compared the windows tests before and after this PR, and it seems that the time-consuming of many tests has increased significantly, as well as these several reverts, I...

@rkooo567 The difference is that both `AgentInfo` and `NodeInfo` exist on `GCSNodeInfo`, which is better than `internalKV`: 1. InternalKV data will leak, there is no place to delete it after...

> Some CI jobs tests failed, e.g. [buildkite.com/ray-project/oss-ci-build-pr/builds/431#01837a6f-fb4c-4e09-97e4-c9a57fdb10c9/3394-4825](https://buildkite.com/ray-project/oss-ci-build-pr/builds/431#01837a6f-fb4c-4e09-97e4-c9a57fdb10c9/3394-4825) , could you please take a look? I tried to fix it, and now all three tests can finish on my machine

> @Catch-Bull Possibly relevant failure? [buildkite.com/ray-project/oss-ci-build-pr/builds/566#01837fbf-2e34-449b-8125-1834439b9e51/3305-3805](https://buildkite.com/ray-project/oss-ci-build-pr/builds/566#01837fbf-2e34-449b-8125-1834439b9e51/3305-3805) @architkulkarni I fixed the UT again and it looks fine now.

> 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...

> A few high-level comments: > > 1. I think this doc is lacking key applications. The original idea behind ownership was that reference holders would often fate-share with the...

@ericl @iycheng @stephanie-wang @scv119 We discussed it internally: # Options to implement object HA with checkpoint ## Option 1: Ownerless objects Every checkpointed object will be ownerless. When `ray.get` on...

> What are the plans for implementing this feature? we will give the plans after this REP gets merged.

Object High Availability prototype in here: https://github.com/ray-project/ray/pull/27541