tdcmeehan opened a new pull request, #59111: URL: https://github.com/apache/spark/pull/59111
### What changes were proposed in this pull request? Create SQL metrics for the portable external UDF worker counters, byte totals, and nanosecond timings. When a worker session closes, aggregate terminal metrics returned by both FinishResponse and acknowledged CancelResponse. Unknown metric names and malformed, negative, or overflowing values are ignored so workers can evolve without breaking query completion. ### Why are the changes needed? Worker terminal responses already preserve per-execution metrics, but ExternalUDFExec discarded the returned Termination. This connects that existing client boundary to Spark SQL metrics. ### Does this PR introduce any user-facing change? Yes. External UDF physical operators expose worker-provided execution metrics in the SQL UI and event data. ### How was this patch tested? Added ExternalUDFExecSuite coverage for finish and cancellation aggregation and for unknown, malformed, negative, and overflowing values. The focused suite passed with 2 tests, and the 19-module reactor compiled successfully. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
