[MAPREDUCE-6766] Concurrent local job failures due to uniqueNumberGenerator = new AtomicLong(System.currentTimeMillis())

• Type: Bug
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 2.3.0, 2.7.1

Environment

Druid 0.8.3

• Assignee: Unassigned
• Reporter: Mark S
• Created: 24/Aug/16 16:00
• Updated: 31/Oct/16 16:34

Description

I am seeing the following exception when attempting to execute multiple Hadoop Local Jobs with Druid.

java.io.IOException: Rename cannot overwrite non empty destination directory /tmp/hadoop-username/mapred/local/1472019105135

From a quick look at the Hadoop code base, it seems that the uniqueNumberGenerator for the LocalDistributedCacheManager is based on the System time, and this appears to cause problems for concurrent jobs.

// Generating unique numbers for FSDownload.
    AtomicLong uniqueNumberGenerator =
new AtomicLong(System.currentTimeMillis());

I am pretty sure the following line of code is responsible, and this seems to exist in latter versions of such as 2.7.1:

• Hadoop 2.3.0 - LocalDistributedCacheManager.java#L96
• Hadoop 2.7.1 - LocalDistributedCacheManager.java#L96

Full Stack Trace

java.lang.RuntimeException: java.lang.reflect.InvocationTargetException
	at com.google.common.base.Throwables.propagate(Throwables.java:160) ~[guava-16.0.1.jar:?]
	at io.druid.indexing.common.task.HadoopTask.invokeForeignLoader(HadoopTask.java:138) ~[druid-indexing-service-0.8.3.jar:0.8.3]
	at io.druid.indexing.common.task.HadoopIndexTask.run(HadoopIndexTask.java:206) ~[druid-indexing-service-0.8.3.jar:0.8.3]
	at io.druid.indexing.overlord.ThreadPoolTaskRunner$ThreadPoolTaskRunnerCallable.call(ThreadPoolTaskRunner.java:285) [druid-indexing-service-0.8.3.jar:0.8.3]
	at io.druid.indexing.overlord.ThreadPoolTaskRunner$ThreadPoolTaskRunnerCallable.call(ThreadPoolTaskRunner.java:265) [druid-indexing-service-0.8.3.jar:0.8.3]
	at java.util.concurrent.FutureTask.run(FutureTask.java:266) [?:1.8.0_71]
	at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142) [?:1.8.0_71]
	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617) [?:1.8.0_71]
	at java.lang.Thread.run(Thread.java:745) [?:1.8.0_71]
Caused by: java.lang.reflect.InvocationTargetException
	at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) ~[?:1.8.0_71]
	at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) ~[?:1.8.0_71]
	at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) ~[?:1.8.0_71]
	at java.lang.reflect.Method.invoke(Method.java:497) ~[?:1.8.0_71]
	at io.druid.indexing.common.task.HadoopTask.invokeForeignLoader(HadoopTask.java:135) ~[druid-indexing-service-0.8.3.jar:0.8.3]
	... 7 more
Caused by: java.lang.RuntimeException: java.io.IOException: java.util.concurrent.ExecutionException: java.io.IOException: Rename cannot overwrite non empty destination directory /tmp/hadoop-username/mapred/local/1472019105135
	at io.druid.indexer.IndexGeneratorJob.run(IndexGeneratorJob.java:211) ~[druid-indexing-hadoop-0.8.3.jar:0.8.3]
	at io.druid.indexer.JobHelper.runJobs(JobHelper.java:321) ~[druid-indexing-hadoop-0.8.3.jar:0.8.3]
	at io.druid.indexer.HadoopDruidIndexerJob.run(HadoopDruidIndexerJob.java:96) ~[druid-indexing-hadoop-0.8.3.jar:0.8.3]
	at io.druid.indexing.common.task.HadoopIndexTask$HadoopIndexGeneratorInnerProcessing.runTask(HadoopIndexTask.java:259) ~[druid-indexing-service-0.8.3.jar:0.8.3]
	at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) ~[?:1.8.0_71]
	at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) ~[?:1.8.0_71]
	at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) ~[?:1.8.0_71]
	at java.lang.reflect.Method.invoke(Method.java:497) ~[?:1.8.0_71]
	at io.druid.indexing.common.task.HadoopTask.invokeForeignLoader(HadoopTask.java:135) ~[druid-indexing-service-0.8.3.jar:0.8.3]
	... 7 more
Caused by: java.io.IOException: java.util.concurrent.ExecutionException: java.io.IOException: Rename cannot overwrite non empty destination directory /tmp/hadoop-username/mapred/local/1472019105135
	at org.apache.hadoop.mapred.LocalDistributedCacheManager.setup(LocalDistributedCacheManager.java:143) ~[?:?]
	at org.apache.hadoop.mapred.LocalJobRunner$Job.<init>(LocalJobRunner.java:163) ~[?:?]
	at org.apache.hadoop.mapred.LocalJobRunner.submitJob(LocalJobRunner.java:731) ~[?:?]
	at org.apache.hadoop.mapreduce.JobSubmitter.submitJobInternal(JobSubmitter.java:240) ~[?:?]
	at org.apache.hadoop.mapreduce.Job$10.run(Job.java:1290) ~[?:?]
	at org.apache.hadoop.mapreduce.Job$10.run(Job.java:1287) ~[?:?]
	at java.security.AccessController.doPrivileged(Native Method) ~[?:1.8.0_71]
	at javax.security.auth.Subject.doAs(Subject.java:422) ~[?:1.8.0_71]
	at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1657) ~[?:?]
	at org.apache.hadoop.mapreduce.Job.submit(Job.java:1287) ~[?:?]
	at io.druid.indexer.IndexGeneratorJob.run(IndexGeneratorJob.java:199) ~[druid-indexing-hadoop-0.8.3.jar:0.8.3]
	at io.druid.indexer.JobHelper.runJobs(JobHelper.java:321) ~[druid-indexing-hadoop-0.8.3.jar:0.8.3]
	at io.druid.indexer.HadoopDruidIndexerJob.run(HadoopDruidIndexerJob.java:96) ~[druid-indexing-hadoop-0.8.3.jar:0.8.3]
	at io.druid.indexing.common.task.HadoopIndexTask$HadoopIndexGeneratorInnerProcessing.runTask(HadoopIndexTask.java:259) ~[druid-indexing-service-0.8.3.jar:0.8.3]
	at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) ~[?:1.8.0_71]
	at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) ~[?:1.8.0_71]
	at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) ~[?:1.8.0_71]
	at java.lang.reflect.Method.invoke(Method.java:497) ~[?:1.8.0_71]
	at io.druid.indexing.common.task.HadoopTask.invokeForeignLoader(HadoopTask.java:135) ~[druid-indexing-service-0.8.3.jar:0.8.3]
	... 7 more
Caused by: java.util.concurrent.ExecutionException: java.io.IOException: Rename cannot overwrite non empty destination directory /tmp/hadoop-username/mapred/local/1472019105135
	at java.util.concurrent.FutureTask.report(FutureTask.java:122) ~[?:1.8.0_71]
	at java.util.concurrent.FutureTask.get(FutureTask.java:192) ~[?:1.8.0_71]
	at org.apache.hadoop.mapred.LocalDistributedCacheManager.setup(LocalDistributedCacheManager.java:139) ~[?:?]
	at org.apache.hadoop.mapred.LocalJobRunner$Job.<init>(LocalJobRunner.java:163) ~[?:?]
	at org.apache.hadoop.mapred.LocalJobRunner.submitJob(LocalJobRunner.java:731) ~[?:?]
	at org.apache.hadoop.mapreduce.JobSubmitter.submitJobInternal(JobSubmitter.java:240) ~[?:?]
	at org.apache.hadoop.mapreduce.Job$10.run(Job.java:1290) ~[?:?]
	at org.apache.hadoop.mapreduce.Job$10.run(Job.java:1287) ~[?:?]
	at java.security.AccessController.doPrivileged(Native Method) ~[?:1.8.0_71]
	at javax.security.auth.Subject.doAs(Subject.java:422) ~[?:1.8.0_71]
	at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1657) ~[?:?]
	at org.apache.hadoop.mapreduce.Job.submit(Job.java:1287) ~[?:?]
	at io.druid.indexer.IndexGeneratorJob.run(IndexGeneratorJob.java:199) ~[druid-indexing-hadoop-0.8.3.jar:0.8.3]
	at io.druid.indexer.JobHelper.runJobs(JobHelper.java:321) ~[druid-indexing-hadoop-0.8.3.jar:0.8.3]
	at io.druid.indexer.HadoopDruidIndexerJob.run(HadoopDruidIndexerJob.java:96) ~[druid-indexing-hadoop-0.8.3.jar:0.8.3]
	at io.druid.indexing.common.task.HadoopIndexTask$HadoopIndexGeneratorInnerProcessing.runTask(HadoopIndexTask.java:259) ~[druid-indexing-service-0.8.3.jar:0.8.3]
	at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) ~[?:1.8.0_71]
	at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) ~[?:1.8.0_71]
	at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) ~[?:1.8.0_71]
	at java.lang.reflect.Method.invoke(Method.java:497) ~[?:1.8.0_71]
	at io.druid.indexing.common.task.HadoopTask.invokeForeignLoader(HadoopTask.java:135) ~[druid-indexing-service-0.8.3.jar:0.8.3]
	... 7 more
Caused by: java.io.IOException: Rename cannot overwrite non empty destination directory /tmp/hadoop-username/mapred/local/1472019105135
	at org.apache.hadoop.fs.AbstractFileSystem.renameInternal(AbstractFileSystem.java:735) ~[?:?]
	at org.apache.hadoop.fs.FilterFs.renameInternal(FilterFs.java:236) ~[?:?]
	at org.apache.hadoop.fs.AbstractFileSystem.rename(AbstractFileSystem.java:678) ~[?:?]
	at org.apache.hadoop.fs.FileContext.rename(FileContext.java:958) ~[?:?]
	at org.apache.hadoop.yarn.util.FSDownload.call(FSDownload.java:366) ~[?:?]
	at org.apache.hadoop.yarn.util.FSDownload.call(FSDownload.java:62) ~[?:?]
	... 4 more

Issue Links

• is related to: MAPREDUCE-6441 Improve temporary directory name generation in LocalDistributedCacheManager for concurrent processes
• is related to: MAPREDUCE-6685 LocalDistributedCacheManager can have overlapping filenames