This winter, Research Computing has been focused on space, both physical and digital. First, we’re happy to announce that Research Computing is now located in room 350 of the Case Library and Geyer Center for Information Technology, adjacent to the north entrance. Our new space features area for collaboration and project planning with faculty and students. We plan to have an open house event sometime later in the spring, and expect to use this space when working with researchers. Please stop by and visit us!
Following the space theme, Research Computing has also been thinking about data storage and working on a long-term strategy for research storage. Last summer, ITS purchased a 160TB IBM Storwize storage system, which was used primarily to provide storage for the research virtual machines mentioned in the last newsletter. This storage is highly reliable and fast, but quite expensive — about $690 per TB. That’s about ten times more expensive than disk storage available at a retail store, because of the speed and redundancy required for such important work. That’s perfect for active storage of data that’s changed and accessed frequently, but too expensive for data that’s used once and then stored long-term.
Over the winter, we’ve been looking at what different researchers across campus are already doing for long-term storage. What’s become clear is that there’s a need for a simple long-term data storage solution that offers the reliability and redundancy that our researchers need, but without such a high cost barrier.
In the last newsletter, we mentioned a small server that we could help configure that would hold a total of 40TB of disk space at a cost closer to $50 per TB. This system is based on FreeNAS, an open-source operating system designed for storage. The FreeNAS software is free and well supported by a community of users who also desire a reliable, inexpensive storage solution. Using FreeNAS along with a relatively simple server and a set of high-quality hard drives, we are able to provide a private server to researchers here at Colgate with enterprise-quality features. These servers are small and energy efficient, so it’s easy for ITS to provide a home for them in our secure, climate-controlled, and power-protected campus data center. Researchers can simply use the storage space as a network drive from any computer on campus, much like Dropbox, Google Drive, or Netstore — but with the speeds available using the campus network.
Research Computing has also been working with the libraries to consider data management best practices so we can better assist researchers developing their own data management plans. Most of us are using Crashplan or DropBox to back up our files, but as with research data, that’s not always the best option. This spring and summer, Research Computing plans to review our backup strategy and supplement it as necessary, which means we will be reaching out to our faculty colleagues to help us better understand research data workflows and help identify any risk that can be mitigated.
If you have any questions about storage, or if you want to get a head start on preparing for your summer research, please let us know. We welcome drop-ins or can arrange to meet in your lab or office.