[
https://issues.apache.org/jira/browse/SPARK-37147?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Apache Spark reassigned SPARK-37147:
------------------------------------
Assignee: (was: Apache Spark)
> MetricsReporter producing NullPointerException when element
> 'triggerExecution' not present in Map[]
> ---------------------------------------------------------------------------------------------------
>
> Key: SPARK-37147
> URL: https://issues.apache.org/jira/browse/SPARK-37147
> Project: Spark
> Issue Type: Bug
> Components: Structured Streaming
> Affects Versions: 3.2.0
> Reporter: Radoslaw Busz
> Priority: Major
>
> The exception occurs in MetricsReporter when it tries to register gauges
> using lastProgress of each stream. This problem was partially fixed in
> https://issues.apache.org/jira/browse/SPARK-22975 but in introduced a NPE
> when 'triggerExecution' element is not present in durationMS map:
>
> {code:java}
> registerGauge("latency", _.durationMs.get("triggerExecution").longValue(), 0L)
> {code}
>
> I find it difficult to reproduce this every time but it happens every few
> restarts when structured streaming uses a slow event source (very little or
> no events). In my case it breaks metric reporting via Codehale/Dropwizard and
> generates multiple stacktraces such as:
> {code:java}
> 21/09/16 09:51:36 ERROR ScheduledReporter: Exception thrown from
> GangliaReporter#report. Exception was suppressed.
> java.lang.NullPointerException
> at
> org.apache.spark.sql.execution.streaming.MetricsReporter.$anonfun$new$3(MetricsReporter.scala:43)
> at
> org.apache.spark.sql.execution.streaming.MetricsReporter.$anonfun$new$3$adapted(MetricsReporter.scala:43)
> at scala.Option.map(Option.scala:230)
> at
> org.apache.spark.sql.execution.streaming.MetricsReporter$$anon$1.getValue(MetricsReporter.scala:68)
> at
> com.codahale.metrics.ganglia.GangliaReporter.reportGauge(GangliaReporter.java:353)
> at
> com.codahale.metrics.ganglia.GangliaReporter.report(GangliaReporter.java:240)
> at
> com.codahale.metrics.ScheduledReporter.report(ScheduledReporter.java:237)
> at
> com.codahale.metrics.ScheduledReporter.lambda$start$0(ScheduledReporter.java:177)
> at
> java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
> at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:308)
> at
> java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:180)
> at
> java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:294)
> at
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
> at
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
> at java.lang.Thread.run(Thread.java:748)
> {code}
> I'm happy to implement a fix.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
