Phoenix has made recent improvements to dashboard and project page performance, including optimizations to reduce data fetched, lazy loading, and cache handling. Upgrading to the latest Phoenix version (11.8.0 or newer) is recommended, as it includes major speed improvements for tracing and dashboards, persistent column selection, and better database usage visibility. Additionally, ensure your deployment uses batch processing and gRPC transport for efficient data transfer, and monitor memory/disk usage to adjust resources as needed.
If you still experience slow loads with large datasets, consider further tuning by reducing the amount of data loaded initially (e.g., smaller page sizes, lazy loading details), and check for any ongoing improvements in the release notes. For advanced monitoring, Phoenix 11.5+ adds a database disk usage monitor to help identify bottlenecks. Let me know if you want more detail on specific configuration or tuning steps.
Sources:
- <https://arize.com/docs/phoenix/production-guide|Production Guide - Phoenix>
- <https://github.com/Arize-ai/phoenix/issues/8420|Slow database queries in spans/traces overview>
- <https://github.com/Arize-ai/phoenix/issues/6201|Performance degradation when opening a project in Phoenix as the number of traces grows>
