https://community.cloudera.com/t5/Archives-of-Support-Questions/Per-node-label-capacity-queues-HDP-2-3/m-p/95084#M8408 <P><A href="http://community.hortonworks.com/users/76/tbenton.html">@Tom Benton</A></P><P>very nicely explained ..<A href="http://aajisaka.github.io/hadoop-project/hadoop-yarn/hadoop-yarn-site/NodeLabel.html">link</A></P><P>use case : We have 5 nodes (hostname=h1..h5) in the cluster, each of them has 24G memory, 24 vcores. 1 among the 5 nodes has GPU (assume it’s h5). So admin added GPU label to h5.</P> Sun, 01 Nov 2015 21:54:32 GMT nsabharwal 2015-11-01T21:54:32Z https://community.cloudera.com/t5/Archives-of-Support-Questions/Per-node-label-capacity-queues-HDP-2-3/m-p/95082#M8406 <P>Does HDP 2.3 support per node label capacity scheduler queues? We'd like to be able to label nodes as "high memory", "dense storage", etc and split the workload within those nodes to maintain SLAs.</P> Thu, 08 Oct 2015 20:31:40 GMT https://community.cloudera.com/t5/Archives-of-Support-Questions/Per-node-label-capacity-queues-HDP-2-3/m-p/95082#M8406 tbenton 2015-10-08T20:31:40Z https://community.cloudera.com/t5/Archives-of-Support-Questions/Per-node-label-capacity-queues-HDP-2-3/m-p/95083#M8407 <P>Yes, check out <A target="_blank" href="http://docs.hortonworks.com/HDPDocuments/HDP2/HDP-2.3.0/bk_yarn_resource_mgt/content/configuring_node_labels.html">http://docs.hortonworks.com/HDPDocuments/HDP2/HDP-2.3.0/bk_yarn_resource_mgt/content/configuring_node_labels.html</A></P> Thu, 08 Oct 2015 21:22:58 GMT https://community.cloudera.com/t5/Archives-of-Support-Questions/Per-node-label-capacity-queues-HDP-2-3/m-p/95083#M8407 deepesh1 2015-10-08T21:22:58Z https://community.cloudera.com/t5/Archives-of-Support-Questions/Per-node-label-capacity-queues-HDP-2-3/m-p/95084#M8408 <P><A href="http://community.hortonworks.com/users/76/tbenton.html">@Tom Benton</A></P><P>very nicely explained ..<A href="http://aajisaka.github.io/hadoop-project/hadoop-yarn/hadoop-yarn-site/NodeLabel.html">link</A></P><P>use case : We have 5 nodes (hostname=h1..h5) in the cluster, each of them has 24G memory, 24 vcores. 1 among the 5 nodes has GPU (assume it’s h5). So admin added GPU label to h5.</P> Sun, 01 Nov 2015 21:54:32 GMT https://community.cloudera.com/t5/Archives-of-Support-Questions/Per-node-label-capacity-queues-HDP-2-3/m-p/95084#M8408 nsabharwal 2015-11-01T21:54:32Z https://community.cloudera.com/t5/Archives-of-Support-Questions/Per-node-label-capacity-queues-HDP-2-3/m-p/95085#M8409 <P>Within the GPU labeled node, could I have to separate workloads that I can enforce on that node only?</P> Sat, 21 Nov 2015 01:15:38 GMT https://community.cloudera.com/t5/Archives-of-Support-Questions/Per-node-label-capacity-queues-HDP-2-3/m-p/95085#M8409 tbenton 2015-11-21T01:15:38Z