Universities Tap Open Source Software-Defined Storage to Manage Big Data
Two universities, Canada's McMaster University and the University of Reading in the United Kingdom, have deployed Red Hat Storage to tackle the growing challenges of big data. Red Hat Storage is an open source, software-only storage solution for managing unstructured data in physical, virtual and cloud environments.
In the middle of a major ERP software deployment, McMaster University found it needed a new storage system with high availability and data replication capabilities. Having already invested heavily in storage hardware, the university turned to the software-defined Red Hat Storage Server to manage both existing and future storage requirements.
"We have a heavily virtualized environment; and we wanted software-defined storage, network and compute to work within that environment," said Wayde Nie, lead architect for University Technology Services at McMaster, in a prepared statement. "If you export the raw storage that resides in the Red Hat Storage software-defined layer, then you get the flexibility to export it and replicate it however you require. That was absolutely key for us."
The deployment can be expanded or upgraded in real time without disrupting operations, allowing McMaster to add more highly available storage to its environment without compromising on performance and without impacting downtime.
At the University of Reading, the Department of Meteorology needed a highly reliable, available and scalable storage file system to manage data for its scientific research projects in weather, climate and earth observation. Two new high-capacity servers have been deployed with Red Hat Storage, allowing the system to scale to support 300 terabytes of data. Additionally, the department is gradually adding data from older, stand-alone servers to host more than 1 petabyte of data in total.
"My priorities are to ensure not only that the department can store hundreds of terabytes of research data efficiently and securely, but that good performance is maintained as the I/O load from our growing compute cluster increases," reported Dan Bretherton, high-performance computing manager at the university, in a press release. "I also knew I'd get the extra help I needed from Red Hat to fine-tune the product for our academic, data-intensive HPC environment."
About the Author
Rhea Kelly is editor in chief for Campus Technology, THE Journal, and Spaces4Learning. She can be reached at [email protected].