mapred.JobClient: Error reading task output

[Shimon <sh...@dpolls.com>]

Hi All,
I'm running the wordcount example on a two node cluster and I get the
following message  -

08/05/18 18:33:00 INFO mapred.JobClient:  map 100% reduce 56%
08/05/18 18:39:02 INFO mapred.JobClient: Task Id :
task_200805181824_0001_m_000002_0, Status : FAILED
Too many fetch-failures
08/05/18 18:39:02 WARN mapred.JobClient: Error reading task
outputhttp://localhost:50060/tasklog?plaintext=true&taskid=task_200805181824_0001_m_000002_0&filter=stdout
08/05/18 18:39:02 WARN mapred.JobClient: Error reading task
outputhttp://localhost:50060/tasklog?plaintext=true&taskid=task_200805181824_0001_m_000002_0&filter=stderr
08/05/18 18:39:03 INFO mapred.JobClient:  map 89% reduce 56%
08/05/18 18:39:12 INFO mapred.JobClient:  map 93% reduce 56%
08/05/18 18:39:17 INFO mapred.JobClient:  map 95% reduce 56%
08/05/18 18:39:18 INFO mapred.JobClient:  map 96% reduce 56%
08/05/18 18:39:22 INFO mapred.JobClient:  map 97% reduce 56%
08/05/18 18:39:27 INFO mapred.JobClient:  map 99% reduce 56%
08/05/18 18:39:32 INFO mapred.JobClient:  map 100% reduce 56%
08/05/18 18:39:43 INFO mapred.JobClient:  map 100% reduce 57%
08/05/18 18:39:48 INFO mapred.JobClient:  map 100% reduce 58%
mapred.JobClient: Error reading task output

08/05/18 18:44:48 INFO mapred.JobClient: Task Id :
task_200805181824_0001_m_000009_0, Status : FAILED
Too many fetch-failures
08/05/18 18:44:48 WARN mapred.JobClient: Error reading task
outputhttp://localhost:50060/tasklog?plaintext=true&taskid=task_200805181824_0001_m_000009_0&filter=stdout
08/05/18 18:44:48 WARN mapred.JobClient: Error reading task
outputhttp://localhost:50060/tasklog?plaintext=true&taskid=task_200805181824_0001_m_000009_0&filter=stderr
08/05/18 18:44:49 INFO mapred.JobClient:  map 89% reduce 58%



The task hangs about 15 minutes before resumption. The same task completes
succesfully in a bout 4 minutes when ran on a single node.


My hadoop-site.xml file contains the following parameters


name value description fs.default.name hdfs://tolik:50000
 mapred.job.tracker tolik:50001
 hadoop.tmp.dir /home/shimon/hadoop_storage
 dfs.replication 2
 mapred.map.tasks.speculative.execution false
 mapred.reduce.tasks.speculative.execution false
 mapred.child.java.opts -Xmx500m
 mapred.map.tasks 10
 mapred.reduce.tasks 4


Any help would be appreciated ,

Thanks