Keep getting Job Tracker error

[Tejas]

I am using 4 node machine where cassandra is deployed. In Master
node's systems.log i am keep getting following error (info) and that
fills up my file system. Is that any way to fix this?

WARN [JOB-TRACKER-INIT] 2012-01-19 10:21:42,924 JobTracker.java (line
2384) Incorrect permissions on cfs://null/tmp/hadoop-cassandra/mapred/system.
Setting it to rwx------
INFO [JOB-TRACKER-INIT] 2012-01-19 10:21:42,929 JobTracker.java (line
2410) Cleaning up the system directory
INFO [JOB-TRACKER-INIT] 2012-01-19 10:21:52,934 JobTracker.java (line
2425) problem cleaning system directory: cfs://null/tmp/hadoop-cassandra/mapred/system
java.io.IOException: TimedOutException()
at
com.datastax.bdp.hadoop.cfs.CassandraFileSystemThriftStore.deleteINode(CassandraFileSystemThriftStore.java:
904)
at
com.datastax.bdp.hadoop.cfs.CassandraFileSystem.delete(CassandraFileSystem.java:
342)
at org.apache.hadoop.mapred.JobTracker.<init>(JobTracker.java:
2411)
at org.apache.hadoop.mapred.JobTracker.<init>(JobTracker.java:
2193)
at org.apache.hadoop.mapred.JobTracker.<init>(JobTracker.java:
2187)
at
org.apache.hadoop.mapred.JobTracker.startTracker(JobTracker.java:303)
at
org.apache.hadoop.mapred.JobTracker.startTracker(JobTracker.java:294)
at com.datastax.bdp.plugin.HadoopTrackerPlugin
$1.run(HadoopTrackerPlugin.java:210)
at java.lang.Thread.run(Thread.java:636)
Caused by: TimedOutException()
at
org.apache.cassandra.thrift.CassandraServer.doInsert(CassandraServer.java:
645)
at
org.apache.cassandra.thrift.CassandraServer.internal_remove(CassandraServer.java:
615)
at
org.apache.cassandra.thrift.CassandraServer.remove(CassandraServer.java:
623)
at
com.datastax.bdp.hadoop.cfs.CassandraFileSystemThriftStore.deleteINode(CassandraFileSystemThriftStore.java:
900)
... 8 more