AnsweredAssumed Answered

Hive Joins in Mapr2 with Hive 0.9 Fail across table types

Question asked by mandoskippy on Oct 15, 2012
Latest reply on Oct 23, 2012 by gera
I have two seperate MapR clusters. Both running Mapr 2.0 and running Hive 0.9  On both clusters if I join two tables that STORED AS RCFILE all works fine. If I join two tables that are TEXTFILE the work fine.  IF I try to join across the types, then I get the error below. Now, what's interesting, if I take the exact same data and move into the data to the different storage types, I can recreate every time.  I have access to a Cloudera Cluster running Hive 0.9 tomorrow and will run the same test. But I think that I JOIN between table types all the time on that cluster, I just need to confirm it before I can guarantee that.  I am stumped here. 
<pre>
Examining task ID: task_201210142240_0053_m_000001 (and more) from job job_201210142240_0053
Exception in thread "Thread-38" java.lang.IllegalArgumentException: port out of range:-1
at java.net.InetSocketAddress.<init>(InetSocketAddress.java:118)
at org.apache.hadoop.net.NetUtils.createSocketAddr(NetUtils.java:166)
at org.apache.hadoop.net.NetUtils.createSocketAddr(NetUtils.java:124)
at org.apache.hadoop.hive.ql.exec.JobTrackerURLResolver.getURL(JobTrackerURLResolver.java:42)
at org.apache.hadoop.hive.ql.exec.JobDebugger.showJobFailDebugInfo(JobDebugger.java:209)
at org.apache.hadoop.hive.ql.exec.JobDebugger.run(JobDebugger.java:92)
at java.lang.Thread.run(Thread.java:662)
</pre>

Outcomes