[jira] [Updated] (MAPREDUCE-5867) Possible NPE in KillAMPreemptionPolicy related to ProportionalCapacityPreemptionPolicy
[ https://issues.apache.org/jira/browse/MAPREDUCE-5867?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Sunil G updated MAPREDUCE-5867: --- Attachment: MapReduce-5867.2.patch Thank You Devaraj for the review. 1. I have updated patch as per the comment for local variable extraction. 2. As of now, there are no test classes available for testing the different AM Preemption policies. May be creating a set of test cases for that feature can be tracked with another Jira. Pls suggest. Possible NPE in KillAMPreemptionPolicy related to ProportionalCapacityPreemptionPolicy -- Key: MAPREDUCE-5867 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5867 Project: Hadoop Map/Reduce Issue Type: Bug Components: resourcemanager Affects Versions: 2.3.0 Reporter: Sunil G Assignee: Sunil G Attachments: MapReduce-5867.2.patch, Yarn-1980.1.patch I configured KillAMPreemptionPolicy for My Application Master and tried to check preemption of queues. In one scenario I have seen below NPE in my AM 014-04-24 15:11:08,860 ERROR [RMCommunicator Allocator] org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: ERROR IN CONTACTING RM. java.lang.NullPointerException at org.apache.hadoop.mapreduce.v2.app.rm.preemption.KillAMPreemptionPolicy.preempt(KillAMPreemptionPolicy.java:57) at org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator.getResources(RMContainerAllocator.java:662) at org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator.heartbeat(RMContainerAllocator.java:246) at org.apache.hadoop.mapreduce.v2.app.rm.RMCommunicator$1.run(RMCommunicator.java:267) at java.lang.Thread.run(Thread.java:662) I was using 2.2.0 and merged MAPREDUCE-5189 to see how AM preemption works. -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (MAPREDUCE-5638) Port Hadoop Archives document to trunk
[ https://issues.apache.org/jira/browse/MAPREDUCE-5638?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13985390#comment-13985390 ] Hudson commented on MAPREDUCE-5638: --- SUCCESS: Integrated in Hadoop-Yarn-trunk #556 (See [https://builds.apache.org/job/Hadoop-Yarn-trunk/556/]) MAPREDUCE-5638. Port Hadoop Archives document to trunk (Akira AJISAKA via jeagles) (jeagles: http://svn.apache.org/viewcvs.cgi/?root=Apache-SVNview=revrev=1591107) * /hadoop/common/trunk/hadoop-mapreduce-project/CHANGES.txt * /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-core/src/site/markdown/HadoopArchives.md.vm * /hadoop/common/trunk/hadoop-project/src/site/site.xml Port Hadoop Archives document to trunk -- Key: MAPREDUCE-5638 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5638 Project: Hadoop Map/Reduce Issue Type: Sub-task Components: documentation Reporter: Akira AJISAKA Assignee: Akira AJISAKA Fix For: 3.0.0, 2.5.0 Attachments: MAPREDUCE-5638-md.patch, MAPREDUCE-5638.patch Now Hadoop Archive document exists only in branch-1. Let's port Hadoop Archives document to trunk. -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Assigned] (MAPREDUCE-5103) Remove dead code QueueManager and JobEndNotifier
[ https://issues.apache.org/jira/browse/MAPREDUCE-5103?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] jhanver chand sharma reassigned MAPREDUCE-5103: --- Assignee: jhanver chand sharma Remove dead code QueueManager and JobEndNotifier Key: MAPREDUCE-5103 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5103 Project: Hadoop Map/Reduce Issue Type: Improvement Reporter: Robert Joseph Evans Assignee: jhanver chand sharma There are a few classes that are dead or duplicate code at this point. org/apache/hadoop/mapred/JobEndNotifier.java org/apache/hadoop/mapred/QueueManager.java org/apache/hadoop/mapred/QueueConfigurationParser.java org/apache/hadoop/mapred/DeprecatedQueueConfigurationParser.java LocalRunner is currently using the JobEndNotifier, but there is a replacement for in in MRv2 org.apache.hadoop.mapreduce.v2.app.JobEndNotifier. The two should be combined together and duplicate code removed. There appears to only be one method called for the QueueManger and it appears to be setting a property that is not used any more, so it can be removed. -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (MAPREDUCE-5867) Possible NPE in KillAMPreemptionPolicy related to ProportionalCapacityPreemptionPolicy
[ https://issues.apache.org/jira/browse/MAPREDUCE-5867?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13985482#comment-13985482 ] Devaraj K commented on MAPREDUCE-5867: -- {quote} 2. As of now, there are no test classes available for testing the different AM Preemption policies. May be creating a set of test cases for that feature can be tracked with another Jira. Pls suggest. {quote} Can you add a new test class for writing test cases as part of this Jira itself, may not be needed to handle as part of another Jira. Possible NPE in KillAMPreemptionPolicy related to ProportionalCapacityPreemptionPolicy -- Key: MAPREDUCE-5867 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5867 Project: Hadoop Map/Reduce Issue Type: Bug Components: resourcemanager Affects Versions: 2.3.0 Reporter: Sunil G Assignee: Sunil G Attachments: MapReduce-5867.2.patch, Yarn-1980.1.patch I configured KillAMPreemptionPolicy for My Application Master and tried to check preemption of queues. In one scenario I have seen below NPE in my AM 014-04-24 15:11:08,860 ERROR [RMCommunicator Allocator] org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: ERROR IN CONTACTING RM. java.lang.NullPointerException at org.apache.hadoop.mapreduce.v2.app.rm.preemption.KillAMPreemptionPolicy.preempt(KillAMPreemptionPolicy.java:57) at org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator.getResources(RMContainerAllocator.java:662) at org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator.heartbeat(RMContainerAllocator.java:246) at org.apache.hadoop.mapreduce.v2.app.rm.RMCommunicator$1.run(RMCommunicator.java:267) at java.lang.Thread.run(Thread.java:662) I was using 2.2.0 and merged MAPREDUCE-5189 to see how AM preemption works. -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Moved] (MAPREDUCE-5870) Support for passing Job priority through Application Submission Context in Mapreduce Side
[ https://issues.apache.org/jira/browse/MAPREDUCE-5870?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Jason Lowe moved YARN-2002 to MAPREDUCE-5870: - Component/s: (was: api) (was: resourcemanager) client Key: MAPREDUCE-5870 (was: YARN-2002) Project: Hadoop Map/Reduce (was: Hadoop YARN) Support for passing Job priority through Application Submission Context in Mapreduce Side - Key: MAPREDUCE-5870 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5870 Project: Hadoop Map/Reduce Issue Type: Improvement Components: client Reporter: Sunil G Attachments: Yarn-2002.1.patch Job Prioirty can be set from client side as below [Configuration and api]. a. JobConf.getJobPriority() and Job.setPriority(JobPriority priority) b. We can also use configuration mapreduce.job.priority. Now this Job priority can be passed in Application Submission context from Client side. Here we can reuse the MRJobConfig.PRIORITY configuration. -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Assigned] (MAPREDUCE-5870) Support for passing Job priority through Application Submission Context in Mapreduce Side
[ https://issues.apache.org/jira/browse/MAPREDUCE-5870?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Jason Lowe reassigned MAPREDUCE-5870: - Assignee: Sunil G Support for passing Job priority through Application Submission Context in Mapreduce Side - Key: MAPREDUCE-5870 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5870 Project: Hadoop Map/Reduce Issue Type: Improvement Components: client Reporter: Sunil G Assignee: Sunil G Attachments: Yarn-2002.1.patch Job Prioirty can be set from client side as below [Configuration and api]. a. JobConf.getJobPriority() and Job.setPriority(JobPriority priority) b. We can also use configuration mapreduce.job.priority. Now this Job priority can be passed in Application Submission context from Client side. Here we can reuse the MRJobConfig.PRIORITY configuration. -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (MAPREDUCE-5638) Port Hadoop Archives document to trunk
[ https://issues.apache.org/jira/browse/MAPREDUCE-5638?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13985545#comment-13985545 ] Hudson commented on MAPREDUCE-5638: --- FAILURE: Integrated in Hadoop-Hdfs-trunk #1747 (See [https://builds.apache.org/job/Hadoop-Hdfs-trunk/1747/]) MAPREDUCE-5638. Port Hadoop Archives document to trunk (Akira AJISAKA via jeagles) (jeagles: http://svn.apache.org/viewcvs.cgi/?root=Apache-SVNview=revrev=1591107) * /hadoop/common/trunk/hadoop-mapreduce-project/CHANGES.txt * /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-core/src/site/markdown/HadoopArchives.md.vm * /hadoop/common/trunk/hadoop-project/src/site/site.xml Port Hadoop Archives document to trunk -- Key: MAPREDUCE-5638 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5638 Project: Hadoop Map/Reduce Issue Type: Sub-task Components: documentation Reporter: Akira AJISAKA Assignee: Akira AJISAKA Fix For: 3.0.0, 2.5.0 Attachments: MAPREDUCE-5638-md.patch, MAPREDUCE-5638.patch Now Hadoop Archive document exists only in branch-1. Let's port Hadoop Archives document to trunk. -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (MAPREDUCE-5638) Port Hadoop Archives document to trunk
[ https://issues.apache.org/jira/browse/MAPREDUCE-5638?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13985634#comment-13985634 ] Hudson commented on MAPREDUCE-5638: --- FAILURE: Integrated in Hadoop-Mapreduce-trunk #1773 (See [https://builds.apache.org/job/Hadoop-Mapreduce-trunk/1773/]) MAPREDUCE-5638. Port Hadoop Archives document to trunk (Akira AJISAKA via jeagles) (jeagles: http://svn.apache.org/viewcvs.cgi/?root=Apache-SVNview=revrev=1591107) * /hadoop/common/trunk/hadoop-mapreduce-project/CHANGES.txt * /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-core/src/site/markdown/HadoopArchives.md.vm * /hadoop/common/trunk/hadoop-project/src/site/site.xml Port Hadoop Archives document to trunk -- Key: MAPREDUCE-5638 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5638 Project: Hadoop Map/Reduce Issue Type: Sub-task Components: documentation Reporter: Akira AJISAKA Assignee: Akira AJISAKA Fix For: 3.0.0, 2.5.0 Attachments: MAPREDUCE-5638-md.patch, MAPREDUCE-5638.patch Now Hadoop Archive document exists only in branch-1. Let's port Hadoop Archives document to trunk. -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (MAPREDUCE-5870) Support for passing Job priority through Application Submission Context in Mapreduce Side
[ https://issues.apache.org/jira/browse/MAPREDUCE-5870?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13985729#comment-13985729 ] Devaraj K commented on MAPREDUCE-5870: -- Please refer TestTypeConverter.java for adding tests for this. Support for passing Job priority through Application Submission Context in Mapreduce Side - Key: MAPREDUCE-5870 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5870 Project: Hadoop Map/Reduce Issue Type: Improvement Components: client Reporter: Sunil G Assignee: Sunil G Attachments: Yarn-2002.1.patch Job Prioirty can be set from client side as below [Configuration and api]. a. JobConf.getJobPriority() and Job.setPriority(JobPriority priority) b. We can also use configuration mapreduce.job.priority. Now this Job priority can be passed in Application Submission context from Client side. Here we can reuse the MRJobConfig.PRIORITY configuration. -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Moved] (MAPREDUCE-5871) Estimate Job Endtime
[ https://issues.apache.org/jira/browse/MAPREDUCE-5871?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Maysam Yabandeh moved YARN-2006 to MAPREDUCE-5871: -- Key: MAPREDUCE-5871 (was: YARN-2006) Project: Hadoop Map/Reduce (was: Hadoop YARN) Estimate Job Endtime Key: MAPREDUCE-5871 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5871 Project: Hadoop Map/Reduce Issue Type: Improvement Reporter: Maysam Yabandeh Assignee: Maysam Yabandeh Attachments: YARN-1969.patch YARN-1969 adds a new earliest-endtime-first policy to the fair scheduler. As a prerequisite step, the AppMaster should estimate its end time and send it to the RM via the heartbeat. This jira focuses on how the AppMaster performs this estimation. -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (MAPREDUCE-5871) Estimate Job Endtime
[ https://issues.apache.org/jira/browse/MAPREDUCE-5871?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13985786#comment-13985786 ] Hadoop QA commented on MAPREDUCE-5871: -- {color:red}-1 overall{color}. Here are the results of testing the latest attachment http://issues.apache.org/jira/secure/attachment/12642650/YARN-1969.patch against trunk revision . {color:green}+1 @author{color}. The patch does not contain any @author tags. {color:green}+1 tests included{color}. The patch appears to include 1 new or modified test files. {color:green}+1 javac{color}. The applied patch does not increase the total number of javac compiler warnings. {color:green}+1 javadoc{color}. There were no new javadoc warning messages. {color:green}+1 eclipse:eclipse{color}. The patch built with eclipse:eclipse. {color:red}-1 findbugs{color}. The patch appears to introduce 3 new Findbugs (version 1.3.9) warnings. {color:red}-1 release audit{color}. The applied patch generated 1 release audit warnings. {color:green}+1 core tests{color}. The patch passed unit tests in hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app. {color:green}+1 contrib tests{color}. The patch passed contrib unit tests. Test results: https://builds.apache.org/job/PreCommit-MAPREDUCE-Build/4569//testReport/ Release audit warnings: https://builds.apache.org/job/PreCommit-MAPREDUCE-Build/4569//artifact/trunk/patchprocess/patchReleaseAuditProblems.txt Findbugs warnings: https://builds.apache.org/job/PreCommit-MAPREDUCE-Build/4569//artifact/trunk/patchprocess/newPatchFindbugsWarningshadoop-mapreduce-client-app.html Console output: https://builds.apache.org/job/PreCommit-MAPREDUCE-Build/4569//console This message is automatically generated. Estimate Job Endtime Key: MAPREDUCE-5871 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5871 Project: Hadoop Map/Reduce Issue Type: Improvement Reporter: Maysam Yabandeh Assignee: Maysam Yabandeh Attachments: YARN-1969.patch YARN-1969 adds a new earliest-endtime-first policy to the fair scheduler. As a prerequisite step, the AppMaster should estimate its end time and send it to the RM via the heartbeat. This jira focuses on how the AppMaster performs this estimation. -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Updated] (MAPREDUCE-5871) Estimate Job Endtime
[ https://issues.apache.org/jira/browse/MAPREDUCE-5871?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Maysam Yabandeh updated MAPREDUCE-5871: --- Attachment: MAPREDUCE-5871.patch Submitting the patch (MAPREDUCE-5871.patch) that resolves the issues raised by bugfinder. Estimate Job Endtime Key: MAPREDUCE-5871 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5871 Project: Hadoop Map/Reduce Issue Type: Improvement Reporter: Maysam Yabandeh Assignee: Maysam Yabandeh Attachments: MAPREDUCE-5871.patch, YARN-1969.patch YARN-1969 adds a new earliest-endtime-first policy to the fair scheduler. As a prerequisite step, the AppMaster should estimate its end time and send it to the RM via the heartbeat. This jira focuses on how the AppMaster performs this estimation. -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (MAPREDUCE-5652) NM Recovery. ShuffleHandler should handle NM restarts
[ https://issues.apache.org/jira/browse/MAPREDUCE-5652?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13985932#comment-13985932 ] Ming Ma commented on MAPREDUCE-5652: 1. Regarding generic interface for restore/recover, I agree there is no much benefit to generalize things for the sake of it. One scenario could be something like ShuffleHandler, some ShuffleHandlers support recovery, some don't. NM can ask if a specific ShuffleHandler if it supports recovery, NM will manage the underlying store and pass the store object to ShuffleHandler and ShuffleHandler manages the serialization and deserialization, etc. If NM decides to change the underlying store and ShuffleHandler doesn't need to change. But at this point, it seems unnecessary. 2. If ShuffleHandler gets DBException during recoverState as part of serviceStart, should ShuffleHandler ignore the exception and continue like the store doesn't exist? The argument for ignoring it is it is soft state and ShuffleHandler can still run without it. Or maybe this can be configurable. NM Recovery. ShuffleHandler should handle NM restarts - Key: MAPREDUCE-5652 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5652 Project: Hadoop Map/Reduce Issue Type: Bug Affects Versions: 2.2.0 Reporter: Karthik Kambatla Assignee: Jason Lowe Labels: shuffle Attachments: MAPREDUCE-5652-v2.patch, MAPREDUCE-5652-v3.patch, MAPREDUCE-5652-v4.patch, MAPREDUCE-5652-v5.patch, MAPREDUCE-5652-v6.patch, MAPREDUCE-5652-v7.patch, MAPREDUCE-5652-v8.patch, MAPREDUCE-5652.patch ShuffleHandler should work across NM restarts and not require re-running map-tasks. On NM restart, the map outputs are cleaned up requiring re-execution of map tasks and should be avoided. -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Updated] (MAPREDUCE-5402) DynamicInputFormat should allow overriding of MAX_CHUNKS_TOLERABLE
[ https://issues.apache.org/jira/browse/MAPREDUCE-5402?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Tsuyoshi OZAWA updated MAPREDUCE-5402: -- Attachment: MAPREDUCE-5402.4.patch DynamicInputFormat should allow overriding of MAX_CHUNKS_TOLERABLE -- Key: MAPREDUCE-5402 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5402 Project: Hadoop Map/Reduce Issue Type: Improvement Components: distcp, mrv2 Reporter: David Rosenstrauch Assignee: Tsuyoshi OZAWA Attachments: MAPREDUCE-5402.1.patch, MAPREDUCE-5402.2.patch, MAPREDUCE-5402.3.patch, MAPREDUCE-5402.4.patch In MAPREDUCE-2765, which provided the design spec for DistCpV2, the author describes the implementation of DynamicInputFormat, with one of the main motivations cited being to reduce the chance of long-tails where a few leftover mappers run much longer than the rest. However, I today ran into a situation where I experienced exactly such a long tail using DistCpV2 and DynamicInputFormat. And when I tried to alleviate the problem by overriding the number of mappers and the split ratio used by the DynamicInputFormat, I was prevented from doing so by the hard-coded limit set in the code by the MAX_CHUNKS_TOLERABLE constant. (Currently set to 400.) This constant is actually set quite low for production use. (See a description of my use case below.) And although MAPREDUCE-2765 states that this is an overridable maximum, when reading through the code there does not actually appear to be any mechanism available to override it. This should be changed. It should be possible to expand the maximum # of chunks beyond this arbitrary limit. For example, here is the situation I ran into today: I ran a distcpv2 job on a cluster with 8 machines containing 128 map slots. The job consisted of copying ~2800 files from HDFS to Amazon S3. I overrode the number of mappers for the job from the default of 20 to 128, so as to more properly parallelize the copy across the cluster. The number of chunk files created was calculated as 241, and mapred.num.entries.per.chunk was calculated as 12. As the job ran on, it reached a point where there were only 4 remaining map tasks, which had each been running for over 2 hours. The reason for this was that each of the 12 files that those mappers were copying were quite large (several hundred megabytes in size) and took ~20 minutes each. However, during this time, all the other 124 mappers sat idle. In theory I should be able to alleviate this problem with DynamicInputFormat. If I were able to, say, quadruple the number of chunk files created, that would have made each chunk contain only 3 files, and these large files would have gotten distributed better around the cluster and copied in parallel. However, when I tried to do that - by overriding mapred.listing.split.ratio to, say, 10 - DynamicInputFormat responded with an exception (Too many chunks created with splitRatio:10, numMaps:128. Reduce numMaps or decrease split-ratio to proceed.) - presumably because I exceeded the MAX_CHUNKS_TOLERABLE value of 400. Is there any particular logic behind this MAX_CHUNKS_TOLERABLE limit? I can't personally see any. If this limit has no particular logic behind it, then it should be overridable - or even better: removed altogether. After all, I'm not sure I see any need for it. Even if numMaps * splitRatio resulted in an extraordinarily large number, if the code were modified so that the number of chunks got calculated as Math.min( numMaps * splitRatio, numFiles), then there would be no need for MAX_CHUNKS_TOLERABLE. In this worst-case scenario where the product of numMaps and splitRatio is large, capping the number of chunks at the number of files (numberOfChunks = numberOfFiles) would result in 1 file per chunk - the maximum parallelization possible. That may not be the best-tuned solution for some users, but I would think that it should be left up to the user to deal with the potential consequence of not having tuned their job properly. Certainly that would be better than having an arbitrary hard-coded limit that *prevents* proper parallelization when dealing with large files and/or large numbers of mappers. -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (MAPREDUCE-5402) DynamicInputFormat should allow overriding of MAX_CHUNKS_TOLERABLE
[ https://issues.apache.org/jira/browse/MAPREDUCE-5402?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13986257#comment-13986257 ] Tsuyoshi OZAWA commented on MAPREDUCE-5402: --- Updates are as follows:: * Changed to use configuration in createSplits(..) * Changed to use configuration in getSplitRatio(..) * Added validattion in getMaxChunksTolerable, getMaxChunksIdeal and getMinRecordsPerChunk * Added tests DynamicInputFormat should allow overriding of MAX_CHUNKS_TOLERABLE -- Key: MAPREDUCE-5402 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5402 Project: Hadoop Map/Reduce Issue Type: Improvement Components: distcp, mrv2 Reporter: David Rosenstrauch Assignee: Tsuyoshi OZAWA Attachments: MAPREDUCE-5402.1.patch, MAPREDUCE-5402.2.patch, MAPREDUCE-5402.3.patch, MAPREDUCE-5402.4.patch In MAPREDUCE-2765, which provided the design spec for DistCpV2, the author describes the implementation of DynamicInputFormat, with one of the main motivations cited being to reduce the chance of long-tails where a few leftover mappers run much longer than the rest. However, I today ran into a situation where I experienced exactly such a long tail using DistCpV2 and DynamicInputFormat. And when I tried to alleviate the problem by overriding the number of mappers and the split ratio used by the DynamicInputFormat, I was prevented from doing so by the hard-coded limit set in the code by the MAX_CHUNKS_TOLERABLE constant. (Currently set to 400.) This constant is actually set quite low for production use. (See a description of my use case below.) And although MAPREDUCE-2765 states that this is an overridable maximum, when reading through the code there does not actually appear to be any mechanism available to override it. This should be changed. It should be possible to expand the maximum # of chunks beyond this arbitrary limit. For example, here is the situation I ran into today: I ran a distcpv2 job on a cluster with 8 machines containing 128 map slots. The job consisted of copying ~2800 files from HDFS to Amazon S3. I overrode the number of mappers for the job from the default of 20 to 128, so as to more properly parallelize the copy across the cluster. The number of chunk files created was calculated as 241, and mapred.num.entries.per.chunk was calculated as 12. As the job ran on, it reached a point where there were only 4 remaining map tasks, which had each been running for over 2 hours. The reason for this was that each of the 12 files that those mappers were copying were quite large (several hundred megabytes in size) and took ~20 minutes each. However, during this time, all the other 124 mappers sat idle. In theory I should be able to alleviate this problem with DynamicInputFormat. If I were able to, say, quadruple the number of chunk files created, that would have made each chunk contain only 3 files, and these large files would have gotten distributed better around the cluster and copied in parallel. However, when I tried to do that - by overriding mapred.listing.split.ratio to, say, 10 - DynamicInputFormat responded with an exception (Too many chunks created with splitRatio:10, numMaps:128. Reduce numMaps or decrease split-ratio to proceed.) - presumably because I exceeded the MAX_CHUNKS_TOLERABLE value of 400. Is there any particular logic behind this MAX_CHUNKS_TOLERABLE limit? I can't personally see any. If this limit has no particular logic behind it, then it should be overridable - or even better: removed altogether. After all, I'm not sure I see any need for it. Even if numMaps * splitRatio resulted in an extraordinarily large number, if the code were modified so that the number of chunks got calculated as Math.min( numMaps * splitRatio, numFiles), then there would be no need for MAX_CHUNKS_TOLERABLE. In this worst-case scenario where the product of numMaps and splitRatio is large, capping the number of chunks at the number of files (numberOfChunks = numberOfFiles) would result in 1 file per chunk - the maximum parallelization possible. That may not be the best-tuned solution for some users, but I would think that it should be left up to the user to deal with the potential consequence of not having tuned their job properly. Certainly that would be better than having an arbitrary hard-coded limit that *prevents* proper parallelization when dealing with large files and/or large numbers of mappers. -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Updated] (MAPREDUCE-5862) Line records longer than 2x split size aren't handled correctly
[ https://issues.apache.org/jira/browse/MAPREDUCE-5862?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] bc Wong updated MAPREDUCE-5862: --- Attachment: 0001-MAPREDUCE-5862.-Line-records-longer-than-2x-split-si.patch Thanks! Updated patch. Added an exception to the rat config. Line records longer than 2x split size aren't handled correctly --- Key: MAPREDUCE-5862 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5862 Project: Hadoop Map/Reduce Issue Type: Bug Affects Versions: 2.3.0 Reporter: bc Wong Assignee: bc Wong Priority: Critical Attachments: 0001-Handle-records-larger-than-2x-split-size.1.patch, 0001-Handle-records-larger-than-2x-split-size.patch, 0001-Handle-records-larger-than-2x-split-size.patch, 0001-MAPREDUCE-5862.-Line-records-longer-than-2x-split-si.patch, recordSpanningMultipleSplits.txt.bz2 Suppose this split (100-200) is in the middle of a record (90-240): {noformat} 0 100200 300 | split | curr | split | --- record --- 90 240 {noformat} Currently, the first split would read the entire record, up to offset 240, which is good. But the 2nd split has a bug in producing a phantom record of (200, 240). -- This message was sent by Atlassian JIRA (v6.2#6252)