fsk119 commented on code in PR #26641:
URL: https://github.com/apache/flink/pull/26641#discussion_r2136762747
##########
flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/plan/nodes/exec/stream/StreamExecMLPredictTableFunction.java:
##########
@@ -301,9 +337,9 @@ private Transformation<RowData> createAsyncModelPredict(
createTransformationMeta(ML_PREDICT_TRANSFORMATION, config),
new AsyncWaitOperatorFactory<>(
asyncFunc,
- asyncLookupOptions.asyncTimeout,
- asyncLookupOptions.asyncBufferCapacity,
- asyncLookupOptions.asyncOutputMode),
+ asyncOptions.asyncTimeout,
Review Comment:
Nope.
1. If upsert keys are subset of the the lookup keys, we don't need to
introduce shuffle here. Because planner promises the data with same upsert keys
is located at the same subtask.
2. If the downstream operator doesn't use output of the predict function as
upsert keys, we don't need deterministic here.
--
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: issues-unsubscr...@flink.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
