[core] telemetry for when memory monitor is enabled

[clarng]

Signed-off-by: Clarence Ng clarence.wyng@gmail.com

Why are these changes needed?

Add telemetry to record at worker init time whether the ray cluster is using memory monitor

Related issue number

Checks

• I've signed off every commit(by using the -s flag, i.e., git commit -s) in this PR.
• I've run scripts/format.sh to lint the changes in this PR.
• I've included any doc changes needed for https://docs.ray.io/en/master/.
• I've made sure the tests are passing. Note that there might be a few flaky tests, see the recent failures at https://flakey-tests.ray.io/
• Testing Strategy
  • Unit tests
  • Release tests
  • This PR is not tested :(

Manual test per https://docs.google.com/document/d/1XSWzx9ZatnlttsurCXzvV4vVsb2g1KxJ2T4zpf33bI8/edit

cat /tmp/ray/session_latest/usage_stats.json

{"usage_stats": {"ray_version": "3.0.0.dev0", "python_version": "3.7.7", "schema_version": "0.1", "source": "OSS", "session_id": "330fafc5-718b-48c5-92d0-dbe8dbc784a5", "git_commit": "{{RAY_COMMIT_SHA}}", "os": "linux", "collect_timestamp_ms": 1665116028212, "session_start_timestamp_ms": 1665115658078, "cloud_provider": null, "min_workers": null, "max_workers": null, "head_node_instance_type": null, "worker_node_instance_types": null, "total_num_cpus": 12, "total_num_gpus": null, "total_memory_gb": 10.672229005023837, "total_object_store_memory_gb": 5.336114501580596, "library_usages": ["dataset", "train", "tune"], "total_success": 302, "total_failed": 0, "seq_number": 302, "extra_usage_tags": {"gcs_storage": "memory", "memory_monitor_enabled": "false"}, "total_num_nodes": 1, "total_num_running_jobs": 1}, "success": true, "error": null}

[scv119]

need fix CI though

[rkooo567]

Can we update this from usage_lib.py? I think we can do this way;

1. Generate a method that gets all internal config from GetInternalConfigRequest.
2. When the dashboard first starts, we get internal config and update only relevant fields only once (in this case memory monitor config)

[rkooo567]

Hmm I think it is not going to work if you do ray start --head. Can you add a test with this case?

[rkooo567]

I'd love to talk in person to talk about how I'd approach this

[clarng]

I'd love to talk in person to talk about how I'd approach this

sounds good, should be back on this very soon

[clarng]

Fixed in #30472