Elm is Stanford Research Computing’s answer to the growing data archiving needs of the research community. Built with cutting-edge, open-source technology, Elm offers a fast, scalable, vendor-neutral and cost-efficient platform for storing large datasets that require long-term retention with infrequent access, making it ideal for archival and compliance purposes.
Features
- On-premises “cold” long-term storage.
- Scalable from 1 TB to hundreds of Petabytes.
- Store up to 100,000 objects (files) for every 1 TB.
- Fast data ingestion via MinIO S3. The system replicates stored data from the disk tier to tape automatically.
- Data protection via erasure coding, checksums, and encryption ensures that data stays intact and secure throughout its lifespan.
Designed for
- Archiving research data for future reference, such as field studies, simulations, and image data.
- Preserving research data to meet project or institutional requirements, such as data for regulatory/policy compliance.
- Backing up research data in the long term, such as periodic back-ups of large datasets.
Requirements
- Must be a Stanford Principal Investigator (PI) or faculty member to purchase the service.
- A Stanford PTA associated with a department, a PI/Faculty budget, or a grant for monthly payments.
Data security
May be used to store Low, Moderate, High Risk Non-PHI Data and High Risk PHI Data, as defined by the Information Security Office.
NOTE: Support for High-Risk Data on Elm will be available soon.
Rates
Get started
Submit a Help request to the Elm Storage team.
Get help
Questions? Contact Stanford Research Computing (srcc-support@stanford.edu).
See also
Oak storage for nearline High-Performance Computing (HPC) storage, mounted on Sherlock and SCG clusters.
