Skip to content

[Observability] Ray Dashboard and Metrics aren't listing Driver by default #50097

@alexeykudinkin

Description

@alexeykudinkin

What happened + What you expected to happen

Recently, i've run into issue where my Driver was going was running out of heap memory, and there's surprisingly no direct way to observe it on any of Ray's observability surfaces

  • Dashboard doesn't list driver as the process running on the head node
  • Metrics don't list driver as standalone component (though it lists dashboard, agent, etc)

This creates a gap where driver isn't really covered by any of the existing surfaces.

Versions / Dependencies

2.41

Reproduction script

Make driver script allocate and retain memory indefinitely
Try to find the source of the problem from Ray Dashboard or Metrics

Issue Severity

None

Metadata

Metadata

Assignees

Labels

P1Issue that should be fixed within a few weeksbugSomething that is supposed to be working; but isn'tcommunity-backlogcore-observabilityobservabilityIssues related to the Ray Dashboard, Logging, Metrics, Tracing, and/or Profiling

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions