Implement Metric Discovery Endpoints
- Dominant language
- Python
- Stars
- 670
- Forks
- 183
- Avg merge
- 17h 7m
- Merged PRs (30d)
- 358
Description
## User Story
As a Backend.AI client developer, I want discovery APIs to find available metrics and presets so that I can build dynamic UIs without hardcoding metric names.
## Implementation Details
- Implement GraphQL query to list available metrics
- Implement GraphQL query to list available presets
- Implement REST endpoint for metric metadata
- Implement REST endpoint for preset catalog
- Include metric descriptions and units
- Include preset descriptions and required parameters
- Add filtering options for discovery queries
- Support versioning information in responses
## Acceptance Criteria
- Clients can retrieve list of all available metrics
- Clients can retrieve list of all available presets
- Metric metadata includes name, description, type, unit, labels
- Preset metadata includes name, description, category, parameters
- Discovery responses are cacheable
- API documentation includes discovery endpoint usage
- Integration tests verify discovery functionality
## Technical Notes
- Based on approved BEP specification
- Consider dynamic discovery from Prometheus metadata
- Cache discovery responses appropriately
JIRA Issue: BA-4055
Contributor guide
Assessment
This issue has not been assessed yet.