Dear all,

I know it is best to use small amount of mem in mapper and reduce.
However, sometimes it is hard to do so. For example, in machine
learning algorithms, it is common to load the model into mem in the
mapper step. When the model is big, I have to allocate a lot of mem
for the mapper.

Here is my question: how can I config hadoop so that it does not fork
too many mappers and run out of physical memory?

My machines have 24G, and I have 100 of them. Each time, hadoop will
fork 6 mappers on each machine, no matter what config I used. I really
want to reduce it to what ever number I want, for example, just 1
mapper per machine.

Here are the config I tried. (I use streaming, and I pass the config
in the command line)

-Dmapred.child.java.opts=-Xmx8000m  <-- did not bring down the number of mappers

-Dmapred.cluster.map.memory.mb=32000 <-- did not bring down the number
of mappers

Am I missing something here?
I use Hadoop 0.20.205

Thanks a lot in advance!
-Shaojun

Try:

-Dmapred.tasktracker.map.tasks.maximum=1

Although I usually put this parameter in mapred-site.xml.

Jeff
Dear all,

I know it is best to use small amount of mem in mapper and reduce.
However, sometimes it is hard to do so. For example, in machine
learning algorithms, it is common to load the model into mem in the
mapper step. When the model is big, I have to allocate a lot of mem
for the mapper.

Here is my question: how can I config hadoop so that it does not fork
too many mappers and run out of physical memory?

My machines have 24G, and I have 100 of them. Each time, hadoop will
fork 6 mappers on each machine, no matter what config I used. I really
want to reduce it to what ever number I want, for example, just 1
mapper per machine.

Here are the config I tried. (I use streaming, and I pass the config
in the command line)

-Dmapred.child.java.opts=-Xmx8000m  <-- did not bring down the number of mappers

-Dmapred.cluster.map.memory.mb=32000 <-- did not bring down the number
of mappers

Am I missing something here?
I use Hadoop 0.20.205

Thanks a lot in advance!
-Shaojun

Not sure about EMR, but if you install your own cluster on EC2 you can use the configs mentioned here:

>> http://hadoop.apache.org/docs/stable/capacity_scheduler.html

Arun

On Jan 18, 2013, at 2:50 PM, Shaojun Zhao wrote:

> I am using Amazon EC2/EMR.
> jps give this
> 16600 JobTracker
> 2732 RunJar
> 2504 StatePusher
> 31902 instance-controller.jar
> 23553 Jps
> 22444 RunJar
> 2077 NameNode
>
> I am not sure how I can impose capacityscheduler on ec2/emr machines.
> -Shaojun
>
> On Fri, Jan 18, 2013 at 1:18 PM, Arun C Murthy <[EMAIL PROTECTED]> wrote:
>> Take a look at the CapacityScheduler and 'High RAM' jobs where-by you can run M map slots per node and request, per-job, that you want N (where N = max(1, N, M)).
>>
>> Some more info:
>> http://hadoop.apache.org/docs/stable/capacity_scheduler.html#Resource+based+scheduling
>> http://hortonworks.com/blog/understanding-apache-hadoops-capacity-scheduler/
>>
>> hth,
>> Arun
>>
>> On Jan 18, 2013, at 12:05 PM, Shaojun Zhao wrote:
>>
>>> Dear all,
>>>
>>> I know it is best to use small amount of mem in mapper and reduce.
>>> However, sometimes it is hard to do so. For example, in machine
>>> learning algorithms, it is common to load the model into mem in the
>>> mapper step. When the model is big, I have to allocate a lot of mem
>>> for the mapper.
>>>
>>> Here is my question: how can I config hadoop so that it does not fork
>>> too many mappers and run out of physical memory?
>>>
>>> My machines have 24G, and I have 100 of them. Each time, hadoop will
>>> fork 6 mappers on each machine, no matter what config I used. I really
>>> want to reduce it to what ever number I want, for example, just 1
>>> mapper per machine.
>>>
>>> Here are the config I tried. (I use streaming, and I pass the config
>>> in the command line)
>>>
>>> -Dmapred.child.java.opts=-Xmx8000m  <-- did not bring down the number of mappers
>>>
>>> -Dmapred.cluster.map.memory.mb=32000 <-- did not bring down the number
>>> of mappers
>>>
>>> Am I missing something here?
>>> I use Hadoop 0.20.205
>>>
>>> Thanks a lot in advance!
>>> -Shaojun
>>
>> --
>> Arun C. Murthy
>> Hortonworks Inc.
>> http://hortonworks.com/
>>
>>

--
Arun C. Murthy
Hortonworks Inc.
http://hortonworks.com/