SfB has two computer clusters for development and testing of services, pipelines, and systems. In addition we have servers for service deployment, big memory computation, and storage.
Hardware. The two clusters named ice and ice2 are built for data intensive computing. Ice is a 5 node cluster with a total of 40 cores (80 threads), 120GB DRAM, and 22.5TB raw storage. Ice2 is a 10 node cluster with a total of 40 cores (80 threads), 320GB DRAM, and 40TB raw storage. We have a frontend (Stormbringer) used to run public services with 12 cores (24 threads), 48GB DRAM, and 2TB storage. In addition we have a machine (Burn) for running big memory tasks with 12 cores (24 threads), 256GB DRAM, and 300GB disk. Our storage resources comprise two storage systems with respectively 12TB and 34TB raw storage.
Software. The clusters run the Rocks Cluster Distribution with 64-bit CenOS Linux. We also use the Open Grid Engine, Cloudera Hadoop, and Spark for data-intensive and parallel computations. We use the Galaxy platform for workflow management. In addition we use several in-house developed infrastructure systems, pipelines, and data exploration tools.
Additional resources. SfB collaborates with the Tromsø Display Wall lab, where we have access to a wall-sized large-resolution display, 3D TV's, and various interaction systems and devices. We also use the Stallo supercomputer at the University of Tromsø and we have access to NORSTORE (through StoreBioInfo) and NOTUR resources.
Resources for SfB members:
Last updated: 07.04.2014 20:35