DanielLeens commented on issue #11351: URL: https://github.com/apache/seatunnel/issues/11351#issuecomment-4888298643
This is a reasonable enhancement direction, and the problem statement is strong: the missing piece today is not raw metrics, but a live topology-centered diagnosis path. The main thing I would keep tight is the first-delivery boundary. A runtime graph becomes expensive very quickly if it tries to solve every observability need at once. For V1, I would suggest making these parts explicit: 1. reuse existing metrics/state signals as much as possible instead of creating a parallel data model 2. optimize first for small and medium DAGs, with a clear fallback for larger graphs 3. make node/edge drill-down the default way to reach details, rather than stuffing everything into the graph itself 4. keep refresh cost and UI noise under control so the graph stays operationally useful If the first design focuses on “where did the slowdown start, and how is it propagating through the DAG”, that is already a strong and reviewable first milestone. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
