[jira] Commented: (HADOOP-3155) reducers stuck at shuffling

["Christian Kunz (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HADOOP-3155?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12594366#action_12594366 ] 

Christian Kunz commented on HADOOP-3155:
----------------------------------------

We have been  running into the same issue, very unlikely to be a network problem:

#nodes 1350
#maps: 100,000 (running about 5 minutes each)
#reduces: 2,500

The maps were finished after about 4 hours, the data shuffling took about 8 hours, the amount of data very small.

Typical pattern of reduces logs:

2008-05-05 19:32:43,215 INFO org.apache.hadoop.mapred.ReduceTask: task_200805050640_0001_r_001079_0 Need 6 map output(s)
2008-05-05 19:32:43,216 INFO org.apache.hadoop.mapred.ReduceTask: task_200805050640_0001_r_001079_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
2008-05-05 19:32:43,216 INFO org.apache.hadoop.mapred.ReduceTask: task_200805050640_0001_r_001079_0 Got 5 known map output location(s); scheduling...
2008-05-05 19:32:43,216 INFO org.apache.hadoop.mapred.ReduceTask: task_200805050640_0001_r_001079_0 Scheduled 0 of 5 known outputs (0 slow hosts and 5 dup hosts)

Also saw log messages with '(xxx slow hosts and 0 dup hosts)


> reducers stuck at shuffling 
> ----------------------------
>
>                 Key: HADOOP-3155
>                 URL: https://issues.apache.org/jira/browse/HADOOP-3155
>             Project: Hadoop Core
>          Issue Type: Bug
>            Reporter: Runping Qi
>
> This happened with hadoop-0.16.2:
> In relatively small job (a few hundreds of mappers and reducers), reducers were stuck at shuffling.
> I saw the lines like the following repeated hundreds of thousands of times over a few hours:
> 2008-04-02 17:17:44,640 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:17:44,641 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:17:44,641 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:17:44,641 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:17:46,643 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:17:46,643 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:17:46,643 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:17:46,643 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:17:48,645 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:17:48,645 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:17:48,645 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:17:48,645 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:17:50,647 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:17:50,647 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:17:50,647 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:17:50,647 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:17:52,649 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:17:52,650 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:17:52,650 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:17:52,650 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:17:54,651 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:17:54,652 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:17:54,652 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:17:54,652 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:17:56,654 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:17:56,654 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:17:56,654 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:17:56,654 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:17:58,656 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:17:58,656 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:17:58,656 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:17:58,656 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:18:00,658 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:18:00,658 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:18:00,658 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:18:00,658 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:18:02,660 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:18:02,661 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:18:02,661 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:18:02,661 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:18:04,662 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:18:04,663 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:18:04,663 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:18:04,663 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:18:06,664 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:18:06,665 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:18:06,665 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:18:06,665 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:18:08,667 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:18:08,667 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:18:08,667 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:18:08,667 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:18:10,669 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:18:10,669 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:18:10,669 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:18:10,669 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:18:12,671 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:18:12,671 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:18:12,671 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:18:12,671 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:18:14,673 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:18:14,674 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:18:14,674 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:18:14,674 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:18:16,675 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:18:16,676 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:18:16,676 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:18:16,676 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:18:18,678 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:18:18,678 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:18:18,678 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:18:18,678 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:18:20,680 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:18:20,680 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:18:20,680 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...
> 2008-04-02 17:18:20,680 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Scheduled 0 of 0 known outputs (0 slow hosts and 0 dup hosts)
> 2008-04-02 17:18:22,682 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Need 2 map output(s)
> 2008-04-02 17:18:22,682 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0: Got 0 new map-outputs & 0 obsolete map-outputs from tasktracker and 0 map-outputs from previous failures
> 2008-04-02 17:18:22,682 INFO org.apache.hadoop.mapred.ReduceTask: task_200804021200_0337_r_000008_0 Got 0 known map output location(s); scheduling...

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.