Scraper Roadmap
The project is moving from a Prometheus metrics endpoint to an Oracle performance scraper backed by PostgreSQL and Grafana.
Near-term work is focused on useful troubleshooting data that is too high-cardinality for Prometheus:
- Improve Database Activity History dashboards and drilldowns.
- Expand SQL performance views around SQL ID, plan hash, executions, elapsed time, CPU, I/O, and wait behavior.
- Expand current session and blocking-session dashboards.
- Add Temp, PGA, and object hotspot collectors and dashboards.
- Improve PostgreSQL partitioning and retention operations.
- Add more operational documentation for real server installs.
- Improve alert log collection and dashboarding.
- Provide additional pre-built Grafana dashboards.
Bounded operational measurements can use optional files listed under
metrics.definitions and stored in oracle_metric_samples. Unbounded
diagnostic entities should use dedicated typed collectors and tables.