hadoop-mapreduce-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From rajeshbabu chintaguntla <rajeshbabu.chintagun...@huawei.com>
Subject maximum capacity of queue is not effecting beyond its capacity
Date Sat, 09 Feb 2013 13:04:35 GMT
Hi,



When I run a job its getting hanged bacause its not able to get free memory resources to map/reduce
task containers.



Total Memory avialable : 8 GB

schedular configured : CapacitySchedular

queue path cofigured : a



Application master got started on 'a' queue path by consuming 1.5 GB



Afterthat  no container is getting launched because its not getting free memory from 'a' queue
even free memory(max - cosumed) is sufficient.

When I debugged I observed queue capacity requirements are satified but user capacity requirements
are not satified.



There is some problem I think. I am not able to follow the logic behind it. Please help me
in this case.



capacity-schedular.xml - configurations

========================

<configurations>

<property>
  <name>yarn.scheduler.capacity.root.queues</name>
  <value>a,b</value>
  <description>The queues at the this level (root is the root queue).
  </description>
</property>

<property>

<name>yarn.scheduler.capacity.root.capacity</name>

<value>100<value>

</property>

<property>

<name>yarn.scheduler.capacity.root.a.capacity</name>

<value>5<value>

</property>

<property>

<name>yarn.scheduler.capacity.root.a.maximum-capacity</name>

<value>60<value>

</property>

<property>

<name>yarn.scheduler.capacity.root.b.capacity</name>

<value>95<value>

</property>

<property>

<name>yarn.scheduler.capacity.maximum-am-resource-percent</name>

<value>8<value>

</property>

<property>

<name>yarn.scheduler.capacity.root.acl_submit_jobs</name>

<value>*<value>

</property>
<property>

<name>yarn.scheduler.capacity.root.a.acl_submit_jobs</name>

<value>*<value>

</property>

<property>

<name>yarn.scheduler.capacity.root.b.acl_submit_jobs</name>

<value>*<value>

</property>

</configurations>



Thanks and Regards,

Rajeshbabu.

Mime
View raw message