Hi all,

I tried to run a MapRe= duce-Job on my tow node FreeBSD-Cluster with Hadoop 2.4.1 and HBase 0.98.4 = and ran into below Exception. I then tried the example provided here, running a MapReduc= e-Job without involving HBase or the HDFS on a single FreeBSD machine and g= ot the same Exception:

Exception in thread "main" java.util.NoSuchElementException=C2=A0=C2=A0=C2=A0 at java.util.StringTokenizer.nextToken(StringTokenizer= .java:349)
=C2=A0=C2=A0=C2=A0 at org.apache.hadoop.fs.RawLocalFileSystem= $DeprecatedRawLocalFileStatus.loadPermissionInfo(RawLocalFileSystem.java:56= 5)
=C2=A0=C2=A0=C2=A0 at org.apache.hadoop.fs.RawLocalFileSystem$DeprecatedRaw= LocalFileStatus.getPermission(RawLocalFileSystem.java:534)
=C2=A0=C2=A0= =C2=A0 at org.apache.hadoop.fs.LocatedFileStatus.<init>(LocatedFileSt= atus.java:42)
=C2=A0=C2=A0=C2=A0 at org.apache.hadoop.fs.FileSystem$4.ne= xt(FileSystem.java:1697)
=C2=A0=C2=A0=C2=A0 at org.apache.hadoop.fs.FileSystem$4.next(FileSystem.jav= a:1679)
=C2=A0=C2=A0=C2=A0 at org.apache.hadoop.mapreduce.lib.input.File= InputFormat.singleThreadedListStatus(FileInputFormat.java:302)
=C2=A0=C2= =A0=C2=A0 at org.apache.hadoop.mapreduce.lib.input.FileInputFormat.listStat= us(FileInputFormat.java:263)
=C2=A0=C2=A0=C2=A0 at org.apache.hadoop.mapreduce.lib.input.FileInputFormat= .getSplits(FileInputFormat.java:375)
=C2=A0=C2=A0=C2=A0 at org.apache.ha= doop.mapreduce.JobSubmitter.writeNewSplits(JobSubmitter.java:493)
=C2=A0= =C2=A0=C2=A0 at org.apache.hadoop.mapreduce.JobSubmitter.writeSplits(JobSub= mitter.java:510)
=C2=A0=C2=A0=C2=A0 at org.apache.hadoop.mapreduce.JobSubmitter.submitJobInt= ernal(JobSubmitter.java:394)
=C2=A0=C2=A0=C2=A0 at org.apache.hadoop.map= reduce.Job$10.run(Job.java:1285)
=C2=A0=C2=A0=C2=A0 at org.apache.hadoop= .mapreduce.Job$10.run(Job.java:1282)
=C2=A0=C2=A0=C2=A0 at java.security= .AccessController.doPrivileged(Native Method)
=C2=A0=C2=A0=C2=A0 at javax.security.auth.Subject.doAs(Subject.java:415)=C2=A0=C2=A0=C2=A0 at org.apache.hadoop.security.UserGroupInformation.doAs= (UserGroupInformation.java:1556)
=C2=A0=C2=A0=C2=A0 at org.apache.hadoop= .mapreduce.Job.submit(Job.java:1282)
=C2=A0=C2=A0=C2=A0 at org.apache.ha= doop.mapreduce.Job.waitForCompletion(Job.java:1303)
=C2=A0=C2=A0=C2=A0 at test.NewMaxTemperature$NewMaxTemperatureMapper.main(N= ewMaxTemperature.java:39)

I ran the exact same project on a Li= nux machine and it worked just fine. So it seems this is a FreeBSD related = issue.

I looked into the code and found the following part, whe= re the exception happens:

/// loads permissions, owner, and group from `ls -ld`
private void loadPermissionInfo() {
  IOException e =3D null;
  try {
    String output =3D FileUtil=
.execCommand(new File(getPath().toUri()),=20
        Shell.getGetPermissionCommand());
    StringTokenizer t =3D
        new StringTokenizer(ou=
tput, Shell.TOKEN_SEPARATOR_RE=
GEX);
    //expected format
    //-rw-------    1 username groupname ...=

    String permission =3D t.nextToken();


So here the StrinTokenizer tries to call nextToken() on an empty String. Ho=
w can this happen?
I checked that "ls -ld" provides the same o=
utput on FreeBSD and Linux.

Does anybody have any suggestion what ha=
ppens here?


Thanks and regards,
Malte