JobTracker balks at empty String for locations
----------------------------------------------
Key: MAPREDUCE-1724
URL: https://issues.apache.org/jira/browse/MAPREDUCE-1724
Project: Hadoop Map/Reduce
Issue Type: Bug
Components: jobtracker
Affects Versions: 0.20.1
Environment: CDH 2 hadoop-0.20.1+169.68
Reporter: Craig Macdonald
Priority: Minor
If a split has locations which are "" (empty String), then the JobTracker will
get upset during initialization:
2010-04-22 19:09:20,395 ERROR org.apache.hadoop.mapred.JobTracker: Job
initialization failed:
java.lang.StringIndexOutOfBoundsException: String index out of range: 0
at java.lang.String.charAt(String.java:687)
at org.apache.hadoop.net.NetUtils.normalizeHostName(NetUtils.java:420)
at org.apache.hadoop.net.NetUtils.normalizeHostNames(NetUtils.java:443)
at
org.apache.hadoop.net.CachedDNSToSwitchMapping.resolve(CachedDNSToSwitchMapping.java:42)
at
org.apache.hadoop.mapred.JobTracker.resolveAndAddToTopology(JobTracker.java:2411)
at
org.apache.hadoop.mapred.JobInProgress.createCache(JobInProgress.java:360)
at
org.apache.hadoop.mapred.JobInProgress.initTasks(JobInProgress.java:462)
at org.apache.hadoop.mapred.JobTracker.initJob(JobTracker.java:3183)
at
org.apache.hadoop.mapred.EagerTaskInitializationListener$InitJob.run(EagerTaskInitializationListener.java:79)
at
java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:885)
at
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:907)
at java.lang.Thread.run(Thread.java:637)
Two key points:
* This is different from Hadoop 0.18
* CombineFileSplit has a constructor where String[] location is not specified,
and hence the location array is populated with empty Strings.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
