Re: 0.16.0 HOD python config errors wrt missing '}' on variable names

[Jason Venner <ja...@attributor.com>]

At the present time I just manually substituted the values in the hodrc, 
and that works

Mahadev Konar wrote:
> Could you try with --resource_manager.queue=batch? 
>
> Regards
> mahadev
>
>   
>> -----Original Message-----
>> From: Jason Venner [mailto:jason@attributor.com]
>> Sent: Monday, February 25, 2008 7:41 PM
>> To: hadoop-user@lucene.apache.org
>> Subject: 0.16.0 HOD python config errors wrt missing '}' on variable
>>     
> names
>   
>> My goodness, HOD has been quite a hard climb to get going. Not being
>> particularly skilled in python makes this step a little more
>>     
> difficult.
>   
>> Our current problem is that the ${VAR_NAME} in the hodrc seem to have
>> trouble.
>> These are set in the environment
>>
>> /data/hadoop-0.16.0-dfs/contrib/hod/bin/hod -t ~/hadoop-hod.tar.gz -o
>> "allocate /tmp/ 10"
>> error: bin/hod failed to start.
>> error: invalid 'java-home' specified in section hod: ${JAVA_HOME
>> error: invalid 'batch-home' specified in section resource_manager:
>> ${RM_HOME
>> error: 2 problems found.
>>
>> Note the missing '}' on the variable names.
>> The hodrc is the stock hodrc with some changes in the mapred section,
>> not in the hod section
>>
>> If we pass in the java_home and the rm_home and the queue name we end
>>     
> up
>   
>> with this
>>
>> /data/hadoop-0.16.0-dfs/contrib/hod/bin/hod -Q batch
>> --resource_manager.batch-home=/usr  -j $JAVA_HOME -t
>>     
> ~/hadoop-hod.tar.gz
>   
>> -o "allocate /tmp/ 10"
>> INFO - Service Registry Started.
>> CRITICAL - qsub Failure : /bin/sh: -c: line 0: unexpected EOF while
>> looking for matching `}'
>> CRITICAL - qsub Command : /usr/bin/qsub -l nodes=10 -W x= -l nodes=10
>>     
> -W
>   
>> x= -N "HOD" -r n -d /tmp/ -q ${RM_QUEUE
>> CRITICAL - Scheduler failure, allocation failed.
>>
>>
>> CRITICAL - Cannot allocate cluster /tmp
>>
>> --
>> Jason Venner
>> Attributor - Publish with Confidence <http://www.attributor.com/>
>> Attributor is hiring Hadoop Wranglers, contact if interested
>>     
>
>