Network Scratch¶
Network scratch, also known as global scratch or user scratch, provides temporary, high-performance storage for data used in active computations. It is available on login nodes and also shared across compute nodes, making it suitable for data that needs to be accessed during jobs running on multiple nodes. Each user has access to this directory at /scratch/$USER or $USER_SCRATCH. Use it directory to store very large datasets or temporary pipeline intermediates for a short period of time while running your jobs. The maximum amount of data a single user can store in network scratch is 100 TB at any given time.
Important
Starting January 2023, scratch data will have limited retention. See Scratch Retention Policy for more information.
Warning
Research Computing expects each user to keep their scratch areas clean. The cluster scratch areas are not to be used for archiving data. In order to keep scratch clear and usable for everyone, files older than 30 days will be eligible for deletion.
Policies and Expectations¶
Data stored in /scratch is subject to two limited retention policies.
- Each user will have a quota of 100 TB of scratch storage.
- Files will be retained for a maximum of 30 days, after which they become eligible for deletion.