Collaborative Research and Learning Environment for Geographic Information Science
The CGIS includes an integrated research, learning, and high-performance computing environment for students, faculty, and visiting researchers. The Center also includes a state-of-the-art laboratory for geovisualization and advanced spatiotemporal modeling and simulation. A seminar room is available for presentations and meetings on GIScience topics, as well as a dedicated classroom for the MPS and GCPS in GEOINT programs and other training courses.
The CGIS has a Hadoop Cluster for processing big data. The cluster is built on Dell Poweredge R530 with 2 namenodes and 15 compute nodes. The cluster has 356 cores, 1.5TB of RAM, 180TB of distributed storage, interconnected with 10Gbps network. The cluster is built on a Cloudera distribution of hadoop stack, and this includes yarn resource management, and supports many ecosystem tools including spark, spark2, hive, pig, MR2, Impala, Solr to process, analyze and serve data, HDFS and HBase for data store, and Sqoop2 and Flume for data ingestion.
Researchers can also access the excellent computing facilities that are available in the Department of Geographical Sciences that includes a High Performance Computing (HPC) cluster built on Redhat Enterprise Linux (RHEL) 7 on Dell Poweredge R815/R730 based-systems consisting of 236 processor cores, 1280GB Random Access Memory (RAM) connected to backend storage of 720TB (525TB usable) of RAID6-based online disk storage system.