goutamadwant commented on issue #10588: URL: https://github.com/apache/seatunnel/issues/10588#issuecomment-5159448906
@davidzollo @DanielCarter-stack I have #11629 for this as the first documentation focused part of this issue. The PR adds English and Chinese guidance for collecting Zeta logs, isolating a job with `ST-JID`, preserving the first failure and exception chain, redacting sensitive values, and validating a diagnosis against metrics and the external system. I kept runtime and tooling changes outside this PR. Some useful follow-up areas I see are: - add engine specific collection guidance for Flink and Spark - provide a sanitized support bundle containing job context, logs, checkpoint state, metrics and Kubernetes events - add structured error extraction so the earliest actionable exception is separated from retries and wrapper exceptions - expose read-only diagnostic tools through SeaTunnel Skill and MCP for job logs, status, checkpoints and realtime metrics - correlate logs with realtime backpressure metrics and StainTrace or OpenTelemetry data for performance problems - maintain a small regression corpus of known failures so diagnosis and repair changes can be tested consistently These would be better handled as separate issues after the documentation workflow is reviewed. The support bundle and structured error extraction look like the most useful next steps because they would improve the Web UI, CLI, Skill and MCP paths without making diagnosis dependent on one model or client. Let me know your thoughts. -- 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]
