The University of Arizona has partnered with Amazon Web Services (AWS) to provide a scalable, cost-effective storage solution for research data that doesn’t require immediate access. This solution, known as Tier 2 Storage, is designed for archiving inactive data or storing backups, offering a contrast to Tier 1 storage, which is tightly integrated with high-performance computing (HPC) for active analyses.

Tier 2 Storage utilizes AWS’s intelligent tiering feature to optimize costs while managing data efficiently. When data is first uploaded, it resides in the standard access class. After three months of inactivity, the data automatically transitions to Glacier storage, a more cost-effective, long-term storage option. Glacier is slower to access, meaning users need to request a restore to retrieve their data. If the data remains inactive for another three months in Glacier, it is then moved to Deep Glacier, an even more economical but slower option, requiring the same restore request for access.

This solution supports an S3 intelligent tiering bucket per fund account, with multiple buckets possible upon request. Tier 2 Storage is ideal for archiving and backing up research data, and researchers can use Globus software to transfer data between Tier 1 and Tier 2, or move data from other endpoints. Unlike Tier 1, Tier 2 is not mounted on the HPC clusters, but Globus provides an efficient way to move data when needed.

By offering a flexible, low-cost storage option, Tier 2 ensures that research data is safely archived while minimizing unnecessary expenses for researchers.

Image