[
https://issues.apache.org/jira/browse/HADOOP-2164?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12559623#action_12559623
]
Joydeep Sen Sarma commented on HADOOP-2164:
-------------------------------------------
all hell breaks loose when the root partition gets full. mysterious exceptions
appear in all shapes and sizes.
use the dfs.du.reserved option and the patch that makes it work ..
> Reducer sort failed due to wrong key class
> ------------------------------------------
>
> Key: HADOOP-2164
> URL: https://issues.apache.org/jira/browse/HADOOP-2164
> Project: Hadoop
> Issue Type: Bug
> Components: mapred
> Reporter: Runping Qi
>
> One of my job's reducers failed due to the following exception:
> java.io.IOException: wrong key class: class org.apache.hadoop.io.LongWritable
> is not class org.apache.hadoop.io.Text
> at
> org.apache.hadoop.io.SequenceFile$Sorter$SegmentDescriptor.nextRawKey(SequenceFile.java:2733)
> at
> org.apache.hadoop.io.SequenceFile$Sorter$MergeQueue.merge(SequenceFile.java:2542)
> at
> org.apache.hadoop.io.SequenceFile$Sorter.merge(SequenceFile.java:2326)
> at org.apache.hadoop.mapred.ReduceTask.run(ReduceTask.java:291)
> at
> org.apache.hadoop.mapred.TaskTracker$Child.main(TaskTracker.java:1760)
> I checked the segment files to be merged on the node running the reducer,
> and founded that one of the map output file did record
> org.apache.hadoop.io.LongWritable as its key class!
> The question is how that happened? My job conf specified Text as the map
> output key class, and other map output files had
> the correct key class!
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
