[GitHub] [spark] cloud-fan commented on a change in pull request #31451: [SPARK-34338][SQL] Report metrics from Datasource v2 scan

[GitBox <gi...@apache.org>]

cloud-fan commented on a change in pull request #31451:
URL: https://github.com/apache/spark/pull/31451#discussion_r605374302



##########
File path: sql/catalyst/src/main/java/org/apache/spark/sql/connector/read/PartitionReader.java
##########
@@ -51,7 +51,8 @@
   T get();
 
   /**
-   * Returns an array of custom task metrics. By default it returns empty array.
+   * Returns an array of custom task metrics. By default it returns empty array. Note that it is
+   * not recommended to put heavy logic in this method as it may affect reading performance.

Review comment:
       Since we can't control how frequently the underlying data source tracks the metrics, this is all about how frequently we want to update the SQL metrics.
   
   Since the actual metrics update is sent to the driver via heartbeat, it's not very useful if we update the metrics per-row. How about we do it like per 100 rows? Only update once may be problematic if the task runs for a long time.




-- 
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.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org