The MidSouth Bioinformatics Center has assembled a Computing Cluster to assist researchers whose workstations are insufficient to handle their data processing needs.
The Center anticipates big data in the form of next-gene fastq files as well as other formats. Our Centerâ€™s preparation for big data focused on three areas, storage, RAM and CPUs. By its nature big data requires large storage capacity. The Center has set up a new server with 22 TB of storage and 128GB of RAM. One of its roles will be as a file server for both the Centerâ€™s shared memory systems and cluster (distributed memory system).
Big data often requires large amounts of RAM for efficient processing. The new server with 128GB of RAM along with a machine that has 64GB RAM, are used for RAM intensive jobs. For CPU intensive processing the Center has several multi-CPU systems including the new server with 16 CPUs, two servers with 16 and 8 CPUs respectively as well as a computational cluster.
Computational clusters are an important part of processing big data. The Centerâ€™s Rocks cluster currently has 6 nodes attached to a front-end. Each node has 4 CPUs and 8GB of RAM. More nodes will be added as the servers in the set are refurbished. Currently the front-end is also used as a file server through NFS (network file system mount) providing Â˝ TB of data storage. This will soon be supplemented by the addition of a new file server adding 14TB of storage (through NFS).
The cluster will be used for CPU intensive medium RAM processing tasks including those requiring message passing (MPI). Along with production bioinformatics data processing the cluster will be used as a teaching tool. For example, it will be used to compare and contrast the deference between shared memory systems and distributed memory systems (clusters).
Students will have the opportunity to write and test code on the Centerâ€™s cluster before submitting to the UALR Computational Research Center cluster. Students will also become familiarly submitting jobs using schedulers on the Centerâ€™s clusters thus helping to bridge the gap for students to understand and use larger institutional clusters.
Here at the MidSouth Bioinformatics Center we work closely with The UALR Computational Research Center (CRC).
Our expertise and experience help bridge the cap in teaching our clients to take advantage of large institutional clusters like those at the CRC.
Tape backup system
Our backup server has 2TB of file storage in a raid 6 configuration. The attached tape backup system currently has 10 X 800 GB tapes installed with slots available for adding more tapes.
This server ensures backup copies of development programs and data are available to faculty, staff and students using our systems.