A new Data Storage Allocation project is underway to provide UC Berkeley research faculty with 5TB of storage. This project is a two-year pilot that will begin rolling out late fall semester of 2025 and run through fall 2027 to develop and optimize the provisioning process for storage needs, along with helping to determine and plan for the ongoing level of capacity required. View FAQs on this project/services.
Storage Options
The Data Storage Allocation will provide Berkeley research faculty with 5 TB of storage that can be used across a set of technology services aligned with the most common use cases. These services are:
-
Collaborative Storage: File storage optimized for easy sharing among UC Berkeley researchers and across institutions. Uses Berkeley Box.
-
Computational Storage: Storage designed for use in high-performance and high-throughput computing. Requires the use of the Savio high-performance computing cluster. FAQs for Savio
-
Secure Storage: Storage designed to meet security requirements for highly sensitive research data. Requires the use of the Secure Research Data & Compute (SRDC) platform. FAQs for SRDC
-
General Purpose Storage: Storage used to house data, media, and/or file objects for use by the owner with minimal sharing. Uses on-prem Cloudian and supports S3.
-
Data Sharing & Publishing Storage: Storage used to provide public or mediated access to data, media, and/or files in accordance with funder or publisher requirements. Uses Dataverse. FAQs for Dataverse
Researchers with greater needs for one or more of these offerings may use a larger fraction of their allocation on that service or use additional funds to purchase storage capacity over and above the 5 TB allocation.
Quick Comparison of Key Features
Service |
Best Use |
Security Level |
|
Collaborative Storage (Berkeley Box) |
File sharing & collaboration |
Moderate-High: Up to P3 |
|
Computational Storage (Savio HPC) |
Compute-intensive research |
Low-Moderate: Up to P2 |
|
Secure Storage (SRDC) |
Sensitive data & compliance |
High: Up to P4 |
|
General Purpose Storage (Cloudian) |
General-purpose object storage |
Moderate-High: Up to P3; P4 with an exception |
|
Data Sharing & Publishing Storage (Dataverse) |
Publishing research datasets |
Low-Moderate: Up to P3 |
Goals of Pilot
- Boost researcher productivity with dedicated staff support services.
- Strengthen cybersecurity and meet growing compliance requirements from funding agencies.
- Enhance protection of Berkeley's intellectual property.
- Develop and optimize the storage provisioning process.
- Plan for long-term capacity and resource needs.
This pilot is in response to advocacy by the Senate Committee on Computing and Information Technology (CIT) and DIVCO with storage services provided collaboratively by Research IT, Berkeley IT, and the Library. The commitment to sustain storage provided during the pilot phase will be upheld in the transition to the resulting ongoing service.
Contact & Feedback
We appreciate the hard work of the CIT in advocating for these critical resources and welcome feedback from the research faculty community as we move forward. Send questions, suggestions, and any other feedback you may have to research-storage@berkeley.edu.
Success Metrics
The initial bundled service offering is targeted for rollout to faculty during the fall semester. Project success will be measured by several key metrics throughout the pilot, including:
- The number of requests received and fulfilled.
- The storage allocated and consumed across different tiers.
- The percentage of researchers served by the new bundled service.
Milestones
