Question

I made a cluster and created 4 users. I created 3 queues using capacity scheduler of capacity 70%, 10% and 20 %. I want to add user1, user2 to queueA and user3 to queueB and user4 to queueC. I deleted the default queue. But when I tried to run a mapreduce program, it is going directly to default queue and showing error. How I can configure such that if user1 runs a job, it will automatically run in queueA.?

Answer 1

In client side, you can add properity in hadoop-site.xml, like this

<property>
    <name>mapred.job.queue.name</name>
    <value>queueA</value>
</property>

Then this client`s job will submit to queueA

I don`t know whether if there is some way to config the server side.