reduce 100% but Timed out

reduce 100% but Timed out,Index failed, how to solve this problem?

$tail -f var/druid/task/index_hadoop_wikiticker_2016-07-12T01:58:04.603Z/log

2016-07-12T01:59:22,979 INFO [task-runner-0-priority-0] org.apache.hadoop.mapreduce.Job - map 100% reduce 100%

2016-07-12T02:09:28,795 INFO [task-runner-0-priority-0] org.apache.hadoop.mapreduce.Job - Task Id : attempt_1468203275218_0008_r_000000_0, Status : FAILED

AttemptID:attempt_1468203275218_0008_r_000000_0 Timed out after 600 secs

the hadoop log:

2016-07-12 10:00:50,000 INFO BlockStateChange: BLOCK* addStoredBlock: blockMap updated: xx.xxx.xxx.xx:50010 is added to blk_1073742681_1857{blockUCState=UNDER_CONSTRUCTION, primaryNodeIndex=-1, replicas=[ReplicaUnderConstruction[[DISK]DS-1cc33608-bbbb-4ea2-8ae5-36b503a1e169:NORMAL|RBW], ReplicaUnderConstruction[[DISK]DS-d007aa7c-df6c-407b-bd14-7cb648573974:NORMAL|RBW]]} size 12586

2016-07-12 10:00:50,002 INFO org.apache.hadoop.hdfs.StateChange: DIR* completeFile: /tmp/hadoop-yarn/staging/jianran.tfh/.staging/job_1468203275218_0006/job_1468203275218_0006_1.jhist is closed by DFSClient_NONMAPREDUCE_-787349368_1

2016-07-12 10:00:50,014 INFO BlockStateChange: BLOCK* addStoredBlock: blockMap updated: xx.xxx.xxx.xx:50010 is added to blk_1073742700_1876{blockUCState=UNDER_CONSTRUCTION, primaryNodeIndex=-1, replicas=[ReplicaUnderConstruction[[DISK]DS-1cc33608-bbbb-4ea2-8ae5-36b503a1e169:NORMAL|RBW], ReplicaUnderConstruction[[DISK]DS-f7bc2307-1065-4d7a-9b77-1804256fecec:NORMAL|RBW]]} size 0

2016-07-12 10:00:50,015 INFO org.apache.hadoop.hdfs.StateChange: DIR* completeFile: /tmp/hadoop-yarn/staging/history/done_intermediate/jianran.tfh/job_1468203275218_0006.summary_tmp is closed by DFSClient_NONMAPREDUCE_-787349368_1

2016-07-12 10:00:56,073 INFO org.apache.hadoop.hdfs.StateChange: BLOCK* InvalidateBlocks: ask xx.xxx.xxx.xx:50010 to delete [blk_1073742677_1853, blk_1073742678_1854, blk_1073742679_1855, blk_1073742680_1856, blk_1073742681_1857]

2016-07-12 10:09:27,929 INFO org.apache.hadoop.hdfs.server.namenode.FSEditLog: Number of transactions: 315 Total time for transactions(ms): 15 Number of transactions batched in Syncs: 744 Number of syncs: 214 SyncTimes(ms): 165

2016-07-12 10:09:27,930 INFO org.apache.hadoop.hdfs.StateChange: DIR* completeFile: /user/jianran.tfh/var/druid/hadoop-tmp/wikiticker/2016-07-12T015804.614Z/c8ee2aecb918423db7a9de726f5ae6ea/_temporary/1/_temporary/attempt_1468203275218_0008_r_000000_0/part-r-00000 is closed by DFSClient_attempt_1468203275218_0008_r_000000_0_1405158767_1

Hi,
Depending on the data you are ingesting, sometimes the index generation can take more than 10 minutes.

try increasing mapreduce.task.timeout to 30 minutes.

you may also want to make sure that you are not generating very big segments by setting targetPartitionSize.

Hi Jianran, do you happen to have the index spec you used?

the problem solved, it’s because of the hadoop user not same as the druid, the druid user has no permission to output the results and write the files, thank you

在 2016年7月12日星期二 UTC+8上午10:16:31,jianr…@alibaba-inc.com写道: