OSN

RP account needed

OSN is a distributed object-storage resource consisting of S3-compatible storage pods hosted at research institutions across the US. It is particularly well suited for active research data that needs to be reachable from anywhere over HTTPS rather than from one cluster, and is often used for moving large datasets off instruments, sharing long-tail datasets, and staging data for compute that runs elsewhere. It is reached through standard S3 tooling - AWS CLI, Cyberduck, Rclone and Restic - rather than a login shell or a mounted filesystem.

Get Support

GET AN ACCOUNT ON OSN

OSN Coldfront portal

Storage Documentation

OSN is S3-compatible object storage - no directories, mount paths, or shell access. Storage is organised as buckets within an OSN project; each bucket has an endpoint URL, a name, and up to two key pairs, and bucket names follow AWS S3 rules - lowercase, no spaces or underscores, and unique across all of OSN.

Project allocations run from 10 TB to 50 TB through the standard ACCESS process. Buckets are created in the portal, not from the command line.

1 ACCESS credit = 1 GB; Campus Champions may request 1 TB for testing. Above 50 TB, contact [OSN support].

File System

Directory Path Quota Purge Backup Notes
Bucket storage 50 TB Never purged Not backed up Per-project quota, shared across all buckets in the project. Requested and managed in the OSN portal.

File Transfer Documentation

OSN has no data transfer node and no command-line transfer host. You connect S3 tooling directly to your bucket's endpoint URL, using the access keys issued in the portal. Every client needs three things: the endpoint, the bucket name, and a key pair - read-write or read-only - unless the bucket is open for anonymous read.

Supported Methods Data Transfer Node / Globus Collection Notes
AWS CLI | RECOMMENDED OSN AWS CLI
CYBERDUCK OSN Cyberduck
RCLONE OSN Rclone
RESTIC OSN Restic