mridulm commented on a change in pull request #34749:
URL: https://github.com/apache/spark/pull/34749#discussion_r763317060
##########
File path:
core/src/main/scala/org/apache/spark/scheduler/EventLoggingListener.scala
##########
@@ -249,21 +249,18 @@ private[spark] class EventLoggingListener(
}
override def onExecutorMetricsUpdate(event:
SparkListenerExecutorMetricsUpdate): Unit = {
+ if (event.execId == SparkContext.DRIVER_IDENTIFIER) {
+ logEvent(event)
+ }
Review comment:
I am not sure I follow the comment thread.
The current behavior is, `SparkListenerExecutorMetricsUpdate` is logged when
`spark.eventLog.logStageExecutorMetrics` is enabled. I am not seeing good
reasons to change this behavior.
Given event files are consumed not just by spark, but external applications
and analysis tools, changing semantics needs to be looked into more carefully.
--
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]
