The Renaissance Computing Institute (RENCI) supports cutting-edge research in life,earth, and data sciences. RENCI needed a new storage solution to enable its high performance computing (HPC) cluster and archive large amounts of precious scientific data for years to come.

Archiving Large Volumes of Genomic and Coastal Modeling Data

Based at the University of North Carolina (UNC) at Chapel Hill, RENCI supports groundbreaking research. Scientists from UNC, North Carolina State University, Duke University, and other universities investigate the genetic causes of disease, create detailed storm surge models, facilitate the use of big data for government and industry, and more.

Providing sufficient storage for research projects is critical. “Genome sequencing can produce 100GB or more of data per person,” says Chris Bizon, senior informatics scientist at RENCI. “In addition to keeping the final products, we store temporary data as it’s generated—if something goes wrong, researchers can look at data without having to start at the beginning.”

The IT group at RENCI must also keep scientific data available for research over the long term. “Genomic data is expensive and time-consuming to collect,” says Bizon. “We want to make sure we preserve data so researchers can conduct new analyses in the future.”

Creating a Multi-Tier Environment for HPC with StorNext

Several years ago, the IT group at RENCI implemented a multi-tier storage environment powered by StorNext data management software to streamline workflows and remove impediments for scientific projects. Currently, the StorNext multi-tier environment includes StorNext metadata appliances, StorNext gateway appliances, and Quantum Scalar i6000 tape libraries as well as disk storage from other vendors.

“StorNext provides a single, multi-tier platform that simplifies administration and helps ensure researchers have the resources they need right away,” says Marcin Sliwowski, IT manager at RENCI.

The tape libraries provide a cost-effective way to archive large amounts of scientific data. “In addition to avoiding hardware costs, we can reduce power, cooling, and real estate expenses,” says Sliwowski.

Providing Easy Access to Research Data to Enable Scientific Collaboration

StorNext data management software provides a single, transparent namespace for the entire multi-tier environment, making it easy for researchers to access the data they need. “With StorNext, we can provide access to all research data, so scientists can analyze information without moving data off-site,” says Bizon.

Scaling to Accommodate Continuous Growth in Research Data

The StorNext tape archive at RENCI currently offers nearly 3PB of capacity, giving researchers plenty of room to preserve genomics data and earth science data for years to come. “The StorNext archive can scale to a very large size, so researchers don’t have to worry about how much data they’re saving,” says Erik Scott, senior research software developer at RENCI.

Improving HPC Speed with Dedicated Scratch Disk Built on StorNext

A few years ago, the RENCI IT group expanded its use of StorNext to support a new scratch disk environment for its HPC cluster. Researchers have noticed performance improvements from the dedicated scratch space. “The previous scratch storage solution slowed when multiple users tried to stream data at once—it was laptop-class performance,” says Scott. “With StorNext, that’s not a problem anymore. Multiple users can collaborate on data without performance issues.”

Download the case study