Institution  

Leibniz University Hannover

L3S Research Center
Mediator

Jurek Leonhardt

Oleh Astappiev

For further information, interested users can contact the institution mediator by mail from the Mediators page.

Clusters description and Technical Characteristics

  • LUH/L3S HADOOP Cluster

     

    		 Master Nodes Compute Nodes Login Nodes
    Node count 2 40 2
    CPU cores 56 925 32 + 80
    Hard disk space 6TB 2.4PB (HDFS) 24TB + 64TB
    Main memory 256GiB 6528GiB 768GiB + 4TB

Technical Information to Access the Computing Resources

Login nodes

  • master03 (hadoop3.kbs.uni-hannover.de)
    • Has a limited CPU and Memory usage per user. This guarantee that your jobs will not be affected by the jobs of other users and improve general stability.
    • Gateway roles: HDFS, HBase, Hive, Spark, Yarn
  • master04 (hadoop4.kbs.uni-hannover.de)
    • Doesn't have any limits per user. Can be used for many jobs, you can use as many resources as needed, but it also means that others can too.
    • Gateway roles: HDFS, HBase, Hive, Spark, Yarn

Master nodes

  • master01 (hadoop.kbs.uni-hannover.de)
    • Active server roles: HDFS, HBase, Hive, Oozie, ZooKeeper
  • master02 (hadoop2.kbs.uni-hannover.de)
    • Active server roles: Yarn, Spark, Cloudera, Prometheus
    • Standby server roles: HDFS, HBase, Hive, Oozie, ZooKeeper

Availability of the Resources

 Updated information will be provided soon

Administrative Access Mechanisms and Requisites
  • In order to access the resources, one should request it via mail. There will be a privacy agreement that needs to be signed before access is granted.