[jira] [Comment Edited] (SPARK-30460) Spark checkpoint failing after some run with S3 path
[ https://issues.apache.org/jira/browse/SPARK-30460?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17012784#comment-17012784 ] Sachin Pasalkar edited comment on SPARK-30460 at 1/10/20 12:19 PM: --- [~gsomogyi] Yes may be or may be not. I was able to run this on my production for 4-6 hours without any other issues for 4-5 times. It always failed with this issue. If this fix the some part of problem we should fix it. I understand spark 3.0 has new committer but as you said it is not deeply tested. Soon I am going to run my Production with this fix in place, I will update ticket around next EOW. If I was able to run system smoothly or not was (Author: sachin): Yes may be or may be not. I was able to run this on my production for 4-6 hours without any other issues for 4-5 times. It always failed with this issue. If this fix the some part of problem we should fix it. I understand spark 3.0 has new committer but as you said it is not deeply tested. Soon I am going to run my Production with this fix in place, I will update ticket around next EOW. If I was able to run system smoothly or not > Spark checkpoint failing after some run with S3 path > - > > Key: SPARK-30460 > URL: https://issues.apache.org/jira/browse/SPARK-30460 > Project: Spark > Issue Type: Bug > Components: DStreams >Affects Versions: 2.4.4 >Reporter: Sachin Pasalkar >Priority: Major > > We are using EMR with the SQS as source of stream. However it is failing, > after 4-6 hours of run, with below exception. Application shows its running > but stops the processing the messages > {code:java} > 2020-01-06 13:04:10,548 WARN [BatchedWriteAheadLog Writer] > org.apache.spark.streaming.util.BatchedWriteAheadLog:BatchedWriteAheadLog > Writer failed to write ArrayBuffer(Record(java.nio.HeapByteBuffer[pos=0 > lim=1226 cap=1226],1578315850302,Future())) > java.lang.UnsupportedOperationException > at > com.amazon.ws.emr.hadoop.fs.s3n2.S3NativeFileSystem2.append(S3NativeFileSystem2.java:150) > at org.apache.hadoop.fs.FileSystem.append(FileSystem.java:1181) > at > com.amazon.ws.emr.hadoop.fs.EmrFileSystem.append(EmrFileSystem.java:295) > at > org.apache.spark.streaming.util.HdfsUtils$.getOutputStream(HdfsUtils.scala:35) > at > org.apache.spark.streaming.util.FileBasedWriteAheadLogWriter.stream$lzycompute(FileBasedWriteAheadLogWriter.scala:32) > at > org.apache.spark.streaming.util.FileBasedWriteAheadLogWriter.stream(FileBasedWriteAheadLogWriter.scala:32) > at > org.apache.spark.streaming.util.FileBasedWriteAheadLogWriter.(FileBasedWriteAheadLogWriter.scala:35) > at > org.apache.spark.streaming.util.FileBasedWriteAheadLog.getLogWriter(FileBasedWriteAheadLog.scala:229) > at > org.apache.spark.streaming.util.FileBasedWriteAheadLog.write(FileBasedWriteAheadLog.scala:94) > at > org.apache.spark.streaming.util.FileBasedWriteAheadLog.write(FileBasedWriteAheadLog.scala:50) > at > org.apache.spark.streaming.util.BatchedWriteAheadLog.org$apache$spark$streaming$util$BatchedWriteAheadLog$$flushRecords(BatchedWriteAheadLog.scala:175) > at > org.apache.spark.streaming.util.BatchedWriteAheadLog$$anon$1.run(BatchedWriteAheadLog.scala:142) > at java.lang.Thread.run(Thread.java:748) > 2020-01-06 13:04:10,554 WARN [wal-batching-thread-pool-0] > org.apache.spark.streaming.scheduler.ReceivedBlockTracker:Exception thrown > while writing record: > BlockAdditionEvent(ReceivedBlockInfo(0,Some(3),None,WriteAheadLogBasedStoreResult(input-0-1578315849800,Some(3),FileBasedWriteAheadLogSegment(s3://mss-prod-us-east-1-ueba-bucket/streaming/checkpoint/receivedData/0/log-1578315850001-1578315910001,0,5175 > to the WriteAheadLog. > org.apache.spark.SparkException: Exception thrown in awaitResult: > at org.apache.spark.util.ThreadUtils$.awaitResult(ThreadUtils.scala:226) > at > org.apache.spark.streaming.util.BatchedWriteAheadLog.write(BatchedWriteAheadLog.scala:84) > at > org.apache.spark.streaming.scheduler.ReceivedBlockTracker.writeToLog(ReceivedBlockTracker.scala:242) > at > org.apache.spark.streaming.scheduler.ReceivedBlockTracker.addBlock(ReceivedBlockTracker.scala:89) > at > org.apache.spark.streaming.scheduler.ReceiverTracker.org$apache$spark$streaming$scheduler$ReceiverTracker$$addBlock(ReceiverTracker.scala:347) > at > org.apache.spark.streaming.scheduler.ReceiverTracker$ReceiverTrackerEndpoint$$anonfun$receiveAndReply$1$$anon$1$$anonfun$run$1.apply$mcV$sp(ReceiverTracker.scala:522) > at org.apache.spark.util.Utils$.tryLogNonFatalError(Utils.scala:1340) > at >
[jira] [Comment Edited] (SPARK-30460) Spark checkpoint failing after some run with S3 path
[ https://issues.apache.org/jira/browse/SPARK-30460?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17012756#comment-17012756 ] Gabor Somogyi edited comment on SPARK-30460 at 1/10/20 11:40 AM: - [~Sachin] even if somebody hunt down this specific issue S3 checkpoint makes streaming jobs dead many other ways. was (Author: gsomogyi): [~Sachin] even if somebody hunt down this specific issue S3 checkpoint makes streaming jobs dead many other different ways. > Spark checkpoint failing after some run with S3 path > - > > Key: SPARK-30460 > URL: https://issues.apache.org/jira/browse/SPARK-30460 > Project: Spark > Issue Type: Bug > Components: DStreams >Affects Versions: 2.4.4 >Reporter: Sachin Pasalkar >Priority: Major > > We are using EMR with the SQS as source of stream. However it is failing, > after 4-6 hours of run, with below exception. Application shows its running > but stops the processing the messages > {code:java} > 2020-01-06 13:04:10,548 WARN [BatchedWriteAheadLog Writer] > org.apache.spark.streaming.util.BatchedWriteAheadLog:BatchedWriteAheadLog > Writer failed to write ArrayBuffer(Record(java.nio.HeapByteBuffer[pos=0 > lim=1226 cap=1226],1578315850302,Future())) > java.lang.UnsupportedOperationException > at > com.amazon.ws.emr.hadoop.fs.s3n2.S3NativeFileSystem2.append(S3NativeFileSystem2.java:150) > at org.apache.hadoop.fs.FileSystem.append(FileSystem.java:1181) > at > com.amazon.ws.emr.hadoop.fs.EmrFileSystem.append(EmrFileSystem.java:295) > at > org.apache.spark.streaming.util.HdfsUtils$.getOutputStream(HdfsUtils.scala:35) > at > org.apache.spark.streaming.util.FileBasedWriteAheadLogWriter.stream$lzycompute(FileBasedWriteAheadLogWriter.scala:32) > at > org.apache.spark.streaming.util.FileBasedWriteAheadLogWriter.stream(FileBasedWriteAheadLogWriter.scala:32) > at > org.apache.spark.streaming.util.FileBasedWriteAheadLogWriter.(FileBasedWriteAheadLogWriter.scala:35) > at > org.apache.spark.streaming.util.FileBasedWriteAheadLog.getLogWriter(FileBasedWriteAheadLog.scala:229) > at > org.apache.spark.streaming.util.FileBasedWriteAheadLog.write(FileBasedWriteAheadLog.scala:94) > at > org.apache.spark.streaming.util.FileBasedWriteAheadLog.write(FileBasedWriteAheadLog.scala:50) > at > org.apache.spark.streaming.util.BatchedWriteAheadLog.org$apache$spark$streaming$util$BatchedWriteAheadLog$$flushRecords(BatchedWriteAheadLog.scala:175) > at > org.apache.spark.streaming.util.BatchedWriteAheadLog$$anon$1.run(BatchedWriteAheadLog.scala:142) > at java.lang.Thread.run(Thread.java:748) > 2020-01-06 13:04:10,554 WARN [wal-batching-thread-pool-0] > org.apache.spark.streaming.scheduler.ReceivedBlockTracker:Exception thrown > while writing record: > BlockAdditionEvent(ReceivedBlockInfo(0,Some(3),None,WriteAheadLogBasedStoreResult(input-0-1578315849800,Some(3),FileBasedWriteAheadLogSegment(s3://mss-prod-us-east-1-ueba-bucket/streaming/checkpoint/receivedData/0/log-1578315850001-1578315910001,0,5175 > to the WriteAheadLog. > org.apache.spark.SparkException: Exception thrown in awaitResult: > at org.apache.spark.util.ThreadUtils$.awaitResult(ThreadUtils.scala:226) > at > org.apache.spark.streaming.util.BatchedWriteAheadLog.write(BatchedWriteAheadLog.scala:84) > at > org.apache.spark.streaming.scheduler.ReceivedBlockTracker.writeToLog(ReceivedBlockTracker.scala:242) > at > org.apache.spark.streaming.scheduler.ReceivedBlockTracker.addBlock(ReceivedBlockTracker.scala:89) > at > org.apache.spark.streaming.scheduler.ReceiverTracker.org$apache$spark$streaming$scheduler$ReceiverTracker$$addBlock(ReceiverTracker.scala:347) > at > org.apache.spark.streaming.scheduler.ReceiverTracker$ReceiverTrackerEndpoint$$anonfun$receiveAndReply$1$$anon$1$$anonfun$run$1.apply$mcV$sp(ReceiverTracker.scala:522) > at org.apache.spark.util.Utils$.tryLogNonFatalError(Utils.scala:1340) > at > org.apache.spark.streaming.scheduler.ReceiverTracker$ReceiverTrackerEndpoint$$anonfun$receiveAndReply$1$$anon$1.run(ReceiverTracker.scala:520) > at > java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) > at > java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) > at java.lang.Thread.run(Thread.java:748) > Caused by: java.lang.UnsupportedOperationException > at > com.amazon.ws.emr.hadoop.fs.s3n2.S3NativeFileSystem2.append(S3NativeFileSystem2.java:150) > at org.apache.hadoop.fs.FileSystem.append(FileSystem.java:1181) > at > com.amazon.ws.emr.hadoop.fs.EmrFileSystem.append(EmrFileSystem.java:295) > at >