I would like to say that I have also had this issue. In two situations, one using Accumulo to store information and also when running multiple streaming jobs within the same streaming context (e.g. multiple save to hdfs). In my case the situation worsens when one of the jobs, which has a long slideduration executes. After this, all other jobs take longer to execute. The situation continues until the batches are too delayed and the system is unstable.
From: Gerard Maas [mailto:[email protected]] Sent: 07 October 2015 11:19 To: Tathagata Das Cc: Cody Koeninger; Adrian Tanase; spark users Subject: Re: Weird performance pattern of Spark Streaming (1.4.1) + direct Kafka *** WARNING *** This message originates from outside our organisation, either from an external partner or the internet. Consider carefully whether you should click on any links, open any attachments or reply. For information regarding Red Flags that you can look out for in emails you receive, click here<http://intranet.ent.baesystems.com/howwework/security/spotlights/Documents/Red%20Flags.pdf>. If you feel the email is suspicious, please follow this process<http://intranet.ent.baesystems.com/howwework/security/spotlights/Documents/Dealing%20With%20Suspicious%20Emails.pdf>. Thanks for the feedback. Cassandra does not seem to be the issue. The time for writing to Cassandra is in the same order of magnitude (see below) The code structure is roughly as follows: dstream.filter(pred).foreachRDD{rdd => val sparkT0 = currentTimeMs val metrics = rdd.mapPartitions{partition => val partitionT0 = currentTimeMs partition.foreach{ transform andThen storeInCassandra _} val partitionT1 = currentTimeMs Seq(Metric( "local time", executor, partitionT1 - partitionT0, records)).iterator } //materialize the rdd val allMetrics = metrics.collect() val sparkT1 = currentTimeMs val totalizedMetrics = // group by and reduce with sum val sparkT2 = currentTimeMs totalizedMetrics.foreach{ metric => gmetric.report(metric)} } Relating this code with the time table presented before (time in ms): How measured? Slow Task Fast Task executor local totalizedMetrics 347.6 281.53 spark computation sparkT1 - sparkT0 6930 263 metric collection sparkT2 - sparkT1 70 138 wall clock process sparkT2 - sparkT0 7000 401 total records processed totalizedMetrics 4297 5002 What we observe is that the largest difference comes from the materialization of the RDD. This pattern repeats cyclically one on, one off. Any ideas where to further look? kr, Gerard. On Wed, Oct 7, 2015 at 1:33 AM, Tathagata Das <[email protected]<mailto:[email protected]>> wrote: Good point! On Tue, Oct 6, 2015 at 4:23 PM, Cody Koeninger <[email protected]<mailto:[email protected]>> wrote: I agree getting cassandra out of the picture is a good first step. But if you just do foreachRDD { _.count } recent versions of direct stream shouldn't do any work at all on the executor (since the number of messages in the rdd is known already) do a foreachPartition and println or count the iterator manually. On Tue, Oct 6, 2015 at 6:02 PM, Tathagata Das <[email protected]<mailto:[email protected]>> wrote: Are sure that this is not related to Cassandra inserts? Could you just do foreachRDD { _.count } instead to keep Cassandra out of the picture and then test this agian. On Tue, Oct 6, 2015 at 12:33 PM, Adrian Tanase <[email protected]<mailto:[email protected]>> wrote: Also check if the Kafka cluster is still balanced. Maybe one of the brokers manages too many partitions, all the work will stay on that executor unless you repartition right after kakfka (and I'm not saying you should). Sent from my iPhone On 06 Oct 2015, at 22:17, Cody Koeninger <[email protected]<mailto:[email protected]>> wrote: I'm not clear on what you're measuring. Can you post relevant code snippets including the measurement code? As far as kafka metrics, nothing currently. There is an info-level log message every time a kafka rdd iterator is instantiated, log.info<http://log.info>(s"Computing topic ${part.topic}, partition ${part.partition} " + s"offsets ${part.fromOffset} -> ${part.untilOffset}") If you log once you're done with an iterator you should be able to see the delta. The other thing to try is reduce the number of parts involved in the job to isolate it ... first thing I'd do there is take cassandra out of the equation. On Tue, Oct 6, 2015 at 2:00 PM, Gerard Maas <[email protected]<mailto:[email protected]>> wrote: Hi Cody, The job is doing ETL from Kafka records to Cassandra. After a single filtering stage on Spark, the 'TL' part is done using the dstream.foreachRDD{rdd.foreachPartition{...TL ...}} pattern. We have metrics on the executor work which we collect and add together, indicated here by 'local computation'. As you can see, we also measure how much it cost us to measure :-) See how 'local work' times are comparable. What's not visible is the task scheduling and consuming the data from Kafka which becomes part of the 'spark computation' part. The pattern we see is 1 fast, 1 slow, 1 fast,... zig...zag... Are there metrics available somehow on the Kafka reading time? Slow Task Fast Task local computation 347.6 281.53 spark computation 6930 263 metric collection 70 138 wall clock process 7000 401 total records processed 4297 5002 (time in ms) kr, Gerard. On Tue, Oct 6, 2015 at 8:01 PM, Cody Koeninger <[email protected]<mailto:[email protected]>> wrote: Can you say anything more about what the job is doing? First thing I'd do is try to get some metrics on the time taken by your code on the executors (e.g. when processing the iterator) to see if it's consistent between the two situations. On Tue, Oct 6, 2015 at 11:45 AM, Gerard Maas <[email protected]<mailto:[email protected]>> wrote: Hi, We recently migrated our streaming jobs to the direct kafka receiver. Our initial migration went quite fine but now we are seeing a weird zig-zag performance pattern we cannot explain. In alternating fashion, one task takes about 1 second to finish and the next takes 7sec for a stable streaming rate. Here are comparable metrics for two successive tasks: Slow: [cid:[email protected]] Executor ID Address Task Time Total Tasks Failed Tasks Succeeded Tasks 20151006-044141-2408867082-5050-21047-S0 dnode-3.hdfs.private:36863 22 s 3 0 3 20151006-044141-2408867082-5050-21047-S1 dnode-0.hdfs.private:43812 40 s 11 0 11 20151006-044141-2408867082-5050-21047-S4 dnode-5.hdfs.private:59945 49 s 10 0 10 Fast: [cid:[email protected]] Executor ID Address Task Time Total Tasks Failed Tasks Succeeded Tasks 20151006-044141-2408867082-5050-21047-S0 dnode-3.hdfs.private:36863 0.6 s 4 0 4 20151006-044141-2408867082-5050-21047-S1 dnode-0.hdfs.private:43812 1 s 9 0 9 20151006-044141-2408867082-5050-21047-S4 dnode-5.hdfs.private:59945 1 s 11 0 11 We have some custom metrics that measure wall-clock time of execution of certain blocks of the job, like the time it takes to do the local computations (RDD.foreachPartition closure) vs total time. The difference between the slow and fast executing task is on the 'spark computation time' which is wall-clock for the task scheduling (DStream.foreachRDD closure) e.g. Slow task: local computation time: 347.60968499999996, spark computation time: 6930, metric collection: 70, total process: 7000, total_records: 4297 Fast task: local computation time: 281.539042, spark computation time: 263, metric collection: 138, total process: 401, total_records: 5002 We are currently running Spark 1.4.1. The load and the work to be done is stable -this is on a dev env with that stuff under control. Any ideas what this behavior could be? thanks in advance, Gerard. ******************************************************************** This email and any attachments are confidential to the intended recipient and may also be privileged. If you are not the intended recipient please delete it from your system and notify the sender. You should not copy it or use it for any purpose nor disclose or distribute its contents to any other person. ********************************************************************
