Re: [PR] [FLINK-38857][model] Add unified metrics support to AsyncPredictFunction and PredictFunction [flink]
featzhang commented on code in PR #27560:
URL: https://github.com/apache/flink/pull/27560#discussion_r2791988363
##
flink-table/flink-table-common/src/main/java/org/apache/flink/table/functions/PredictFunction.java:
##
@@ -29,28 +32,159 @@
* A wrapper class of {@link TableFunction} for synchronous model inference.
*
* The output type of this table function is fixed as {@link RowData}.
+ *
+ * This class provides built-in metrics for monitoring model inference
performance, including:
+ *
+ *
+ * inference_requests: Total number of inference requests
+ * inference_requests_success: Number of successful inference requests
+ * inference_requests_failure: Number of failed inference requests
+ * inference_latency: Histogram of inference latency in milliseconds
+ * inference_rows_output: Total number of output rows from inference
+ *
*/
@PublicEvolving
public abstract class PredictFunction extends TableFunction {
+//
Review Comment:
Thanks for the suggestion.
This PR only adds unified metrics around model invocation in this class and
does not modify the existing functional logic. I’ve added additional comments
to clarify the metric behavior and integration.
If you still think we should add explicit unit tests for these metric hooks,
I’m happy to extend the test coverage accordingly.
--
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]
Re: [PR] [FLINK-38857][model] Add unified metrics support to AsyncPredictFunction and PredictFunction [flink]
davidradl commented on code in PR #27560:
URL: https://github.com/apache/flink/pull/27560#discussion_r2788016818
##
flink-table/flink-table-common/src/main/java/org/apache/flink/table/functions/PredictFunction.java:
##
@@ -29,28 +32,159 @@
* A wrapper class of {@link TableFunction} for synchronous model inference.
*
* The output type of this table function is fixed as {@link RowData}.
+ *
+ * This class provides built-in metrics for monitoring model inference
performance, including:
+ *
+ *
+ * inference_requests: Total number of inference requests
+ * inference_requests_success: Number of successful inference requests
+ * inference_requests_failure: Number of failed inference requests
+ * inference_latency: Histogram of inference latency in milliseconds
+ * inference_rows_output: Total number of output rows from inference
+ *
*/
@PublicEvolving
public abstract class PredictFunction extends TableFunction {
+//
Review Comment:
could we have unit tests to implement and use these new classes to make sure
everything works as expected please.
--
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]
Re: [PR] [FLINK-38857][model] Add unified metrics support to AsyncPredictFunction and PredictFunction [flink]
flinkbot commented on PR #27560: URL: https://github.com/apache/flink/pull/27560#issuecomment-3874973141 ## CI report: * 60d8a8881f023790fb30d72c6cbb9cb8fb27e905 UNKNOWN Bot commands The @flinkbot bot supports the following commands: - `@flinkbot run azure` re-run the last Azure build -- 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]
