[ https://issues.apache.org/jira/browse/SPARK-18838?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16024690#comment-16024690 ]
Antoine PRANG commented on SPARK-18838: --------------------------------------- [~joshrosen] I uploaded the new timings and new version of my commits. I think that the bottleneck in the StorageListener execution time is the call to the method "storageStatusList" in the "StorageStatusListener". This call is done for each BlockUpdate (very frequent) and in my case (I have 1000 executors) will re-create a huge Seq for each call. I try to do something for that. But it is not so conclusive (see the last timings that I uploaded). So I did not put this commit in my pull request. But it is in my metrics branch in my fork if you want to take a look). Can I have a review on my pull request ? We should definitely transform the EventLoggingListener in an asynchronous fashion. No events are dropped anymore in my test case, the global execution time of the messages is significantly decreased. Thanks > High latency of event processing for large jobs > ----------------------------------------------- > > Key: SPARK-18838 > URL: https://issues.apache.org/jira/browse/SPARK-18838 > Project: Spark > Issue Type: Improvement > Affects Versions: 2.0.0 > Reporter: Sital Kedia > Attachments: perfResults.pdf, SparkListernerComputeTime.xlsx > > > Currently we are observing the issue of very high event processing delay in > driver's `ListenerBus` for large jobs with many tasks. Many critical > component of the scheduler like `ExecutorAllocationManager`, > `HeartbeatReceiver` depend on the `ListenerBus` events and this delay might > hurt the job performance significantly or even fail the job. For example, a > significant delay in receiving the `SparkListenerTaskStart` might cause > `ExecutorAllocationManager` manager to mistakenly remove an executor which is > not idle. > The problem is that the event processor in `ListenerBus` is a single thread > which loops through all the Listeners for each event and processes each event > synchronously > https://github.com/apache/spark/blob/master/core/src/main/scala/org/apache/spark/scheduler/LiveListenerBus.scala#L94. > This single threaded processor often becomes the bottleneck for large jobs. > Also, if one of the Listener is very slow, all the listeners will pay the > price of delay incurred by the slow listener. In addition to that a slow > listener can cause events to be dropped from the event queue which might be > fatal to the job. > To solve the above problems, we propose to get rid of the event queue and the > single threaded event processor. Instead each listener will have its own > dedicate single threaded executor service . When ever an event is posted, it > will be submitted to executor service of all the listeners. The Single > threaded executor service will guarantee in order processing of the events > per listener. The queue used for the executor service will be bounded to > guarantee we do not grow the memory indefinitely. The downside of this > approach is separate event queue per listener will increase the driver memory > footprint. -- This message was sent by Atlassian JIRA (v6.3.15#6346) --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org