Re: Hadoop 2.2 Built-in Counters

[Harsh J]

You can see them in the MR Job HistoryServer UI if you have one
configured. You can find config properties and launch command in
http://hadoop.apache.org/docs/stable2/hadoop-project-dist/hadoop-common/ClusterSetup.html

On Thu, Aug 14, 2014 at 11:44 AM, ou senshaw  wrote:
> Hi all,
> I'm trying to analyze my mapreduce job performance via built-in counters
> such as physical memory usage, heap memory usage...
> When the job is running, I can watch these counters via Resource manager
> website(namenode:8088). However, when the job is done, counter information
> is not available in resource manager website anymore. I know I can get them
> from client output. I was wondering if there is other place in name node or
> data node to get the final counter measures regarding job id?
> Thanks,
> Shaw



-- 
Harsh J

Re: Hadoop 2.2 Built-in Counters

[Chris MacKenzie]

Hi,

This is the content of my shell script for running the job history server:

cd $HADOOP_PREFIX
hadoop fs -mkdir -p /mr-history/tmp
hadoop fs -chmod -R 1777 /mr-history/tmp
hadoop fs -mkdir -p /mr-history/done
hadoop fs -chmod -R 1777 /mr-history/done

sbin/mr-jobhistory-daemon.sh start historyserver

These configurable variables are in mapred-site.xml


mapreduce.jobhistory.address
137.195.143.129:10020
Default port is 10020.



mapreduce.jobhistory.webapp.address
137.195.143.129:19888
Default port is 19888.


I start the history server on the same node as my resource manager


The counters are available from when the job is running from:

http://your-server:8088/proxy/application_1408007466921_0002/mapreduce/job/
job_1408007466921_0002

Drill down through the application master to the job.

If you don¹t have the history server running the job data is not
persistent.

Hope this helps.


Regards,

Chris MacKenzie
telephone: 0131 332 6967
email: stu...@chrismackenziephotography.co.uk
corporate: www.chrismackenziephotography.co.uk









From:  ou senshaw 
Reply-To:  
Date:  Thursday, 14 August 2014 07:14
To:  
Subject:  Hadoop 2.2 Built-in Counters


Hi all,
I'm trying to analyze my mapreduce job performance via built-in counters
such as physical memory usage, heap memory usage...
When the job is running, I can watch these counters via Resource manager
website(namenode:8088). However, when the job is done, counter information
is not available in resource manager website anymore. I know I can get
them from client output. I was wondering if there is other place in name
node or data node to get the final counter measures regarding job id?
Thanks,
Shaw

Re: Hadoop 2.2 Built-in Counters

[ou senshaw]

Thanks guys! It's working.


On Thu, Aug 14, 2014 at 2:39 AM, Chris MacKenzie <
stu...@chrismackenziephotography.co.uk> wrote:

> Hi,
>
> This is the content of my shell script for running the job history server:
>
> cd $HADOOP_PREFIX
> hadoop fs -mkdir -p /mr-history/tmp
> hadoop fs -chmod -R 1777 /mr-history/tmp
> hadoop fs -mkdir -p /mr-history/done
> hadoop fs -chmod -R 1777 /mr-history/done
>
> sbin/mr-jobhistory-daemon.sh start historyserver
>
> These configurable variables are in mapred-site.xml
>
> 
> mapreduce.jobhistory.address
> 137.195.143.129:10020
> Default port is 10020.
> 
>
> 
> mapreduce.jobhistory.webapp.address
> 137.195.143.129:19888
> Default port is 19888.
> 
>
> I start the history server on the same node as my resource manager
>
>
> The counters are available from when the job is running from:
>
> http://your-server:8088/proxy/application_1408007466921_0002/mapreduce/job/
> job_1408007466921_0002
>
> Drill down through the application master to the job.
>
> If you don¹t have the history server running the job data is not
> persistent.
>
> Hope this helps.
>
>
> Regards,
>
> Chris MacKenzie
> telephone: 0131 332 6967
> email: stu...@chrismackenziephotography.co.uk
> corporate: www.chrismackenziephotography.co.uk
> 
> 
> 
>
>
>
>
>
>
> From:  ou senshaw 
> Reply-To:  
> Date:  Thursday, 14 August 2014 07:14
> To:  
> Subject:  Hadoop 2.2 Built-in Counters
>
>
> Hi all,
> I'm trying to analyze my mapreduce job performance via built-in counters
> such as physical memory usage, heap memory usage...
> When the job is running, I can watch these counters via Resource manager
> website(namenode:8088). However, when the job is done, counter information
> is not available in resource manager website anymore. I know I can get
> them from client output. I was wondering if there is other place in name
> node or data node to get the final counter measures regarding job id?
> Thanks,
> Shaw
>
>
>