You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Wenchen Fan (Jira)" <ji...@apache.org> on 2021/12/09 09:25:00 UTC
[jira] [Commented] (SPARK-37578) DSV2 is not updating Output Metrics
[ https://issues.apache.org/jira/browse/SPARK-37578?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17456259#comment-17456259 ]
Wenchen Fan commented on SPARK-37578:
-------------------------------------
Maybe we can define some builtin metric names like `bytesWritten`, so that Spark can recognize it and update the task metrics accordingly.
cc [~viirya]
> DSV2 is not updating Output Metrics
> -----------------------------------
>
> Key: SPARK-37578
> URL: https://issues.apache.org/jira/browse/SPARK-37578
> Project: Spark
> Issue Type: Bug
> Components: Spark Core
> Affects Versions: 3.0.3, 3.1.2
> Reporter: Sandeep Katta
> Priority: Major
>
> Repro code
> ./bin/spark-shell --master local --jars /Users/jars/iceberg-spark3-runtime-0.12.1.jar
>
> {code:java}
> import scala.collection.mutable
> import org.apache.spark.scheduler._val bytesWritten = new mutable.ArrayBuffer[Long]()
> val recordsWritten = new mutable.ArrayBuffer[Long]()
> val bytesWrittenListener = new SparkListener() {
> override def onTaskEnd(taskEnd: SparkListenerTaskEnd): Unit = {
> bytesWritten += taskEnd.taskMetrics.outputMetrics.bytesWritten
> recordsWritten += taskEnd.taskMetrics.outputMetrics.recordsWritten
> }
> }
> spark.sparkContext.addSparkListener(bytesWrittenListener)
> try {
> val df = spark.range(1000).toDF("id")
> df.write.format("iceberg").save("Users/data/dsv2_test")
>
> assert(bytesWritten.sum > 0)
> assert(recordsWritten.sum > 0)
> } finally {
> spark.sparkContext.removeSparkListener(bytesWrittenListener)
> } {code}
>
>
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org