Does it get stuck before the creating a Hadoop job or after creating a Hadoop 
job.

In case it is stuck before creating a hadoop job you can look at Hive.log 
(wherever you are directing it) for what is taking a long time to setup the job.
In case the Hadoop job has already started you can look at the task attempt 
logs.

Sometimes if you have a lot of small files or lot of partitions Hive can take 
long to setup and start map reduce jobs.

-----Original Message-----
From: Thulasi Ram Naidu Peddineni [mailto:[email protected]] 
Sent: Tuesday, September 27, 2011 10:59 AM
To: [email protected]
Subject: Reg: Map joins in hive

-----
Regards,
Thulasi Ram P




---------- Forwarded message ----------
From: Thulasi Ram Naidu Peddineni <[email protected]>
Date: Tue, Sep 27, 2011 at 11:21 PM
Subject: Reg: Map joins in hive
To: [email protected], [email protected]


Hi,
 I have a huge table x (~150M records and ~5GB) with one partition and another 
table (~200 records and <10KB). I want to join both these tables and thought 
MapJoin is perfect optimization for this. However, my job log says..

Total MapReduce jobs = 2
...
...

Mapred Local Task Succeeded . Convert the Join into MapJoin

Launching Job 1 out of 2

Number of reduce tasks is set to 0 since there's no reduce operator

and then it is stuck at this point of time for a long time. Can you some 
explain what could be happening here ?


-----
Thanks,
Thulasi Ram P

Reply via email to