Skip to content

[core][autoscalerV2] Strucutred cluster_status endpoint #41551

@rickyyx

Description

@rickyyx

What happened + What you expected to happen

for /api/cluster_status, we should not be returning (un)format strings to the client, but a structured output for maintainability.

async def get_cluster_status(self, req):
"""Returns status information about the cluster.
Currently contains two fields:
autoscaling_status (str)-- a status message from the autoscaler.
autoscaling_error (str)-- an error message from the autoscaler if
anything has gone wrong during autoscaling.
These fields are both read from the GCS, it's expected that the
autoscaler writes them there.
"""
# TODO(rickyx): We should be able to get the cluster status from the
# autoscaler directly with V2. And we should be able to return structured data
# rather than a string.

Versions / Dependencies

master

Reproduction script

NA

Issue Severity

None

Metadata

Metadata

Assignees

Labels

P2Important issue, but not time-criticalbugSomething that is supposed to be working; but isn'tcoreIssues that should be addressed in Ray CoredashboardIssues specific to the Ray DashboardobservabilityIssues 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