ARCC High Performance Computing Policy - Storage
Contents
ARCC provides several types of storage for researchers. Each is detailed below.
ARCC provides two options for research data storage. Alcova, and Pathfinder.
Research Data Storage on Alcova is subject to default quotas as listed on our Pricing Sheet.
ARCC high-performance storage system (HPS) for Beartooth is a high speed, tiered storage system designed to maximize performance while minimizing cost. Beartooth is intended to be used for storing data that is actively being used.
The following policies discuss the use of this space. In general, the disk space is intended for support of research using the cluster, and as a courtesy to other users of the cluster, you should try to delete any files that are no longer needed or being used. All data on the HPS, are considered to be related to your research and not to be of a personal nature. As such, all data is considered to be owned by the principal investigator for the allocation through which you have access to the cluster.
Beartooth is for the support of active research using the clusters. You should remove data files, etc. from the cluster promptly when you no longer actively working on the computations requiring them. This is to ensure that all users can avail themselves of these resources.
Note: None of the Beartooth file systems are backed up. We do data replication within the file system in order to minimize the loss of data in case of a system fault or failure. |
Each individual researcher is assigned a standard storage allocation or quota on /home, /project, and /gscratch. Researchers who use more than their allocated space will be blocked from creating new files until they reduce their use, or in the case of /project and /gscratch, request a one-time expansion or purchase additional storage. The chart below shows the storage allocations for individual accounts and the cost of additional space.
/home
Private user space for storing small, long term files such as environment settings, scripts, and source code.
/project
Project-specific space shared among all members of a project for storing short term data, input, and output files.
/gscratch
User-specific space for storing data that is actively being processed. This storage can be purged of old files as needed and is not for long term storage.
/lscratch
Node specific space for storing short-term computational data relevant to jobs running on that node. Files are deleted nightly.
Directory | Backed Up? | Default Storage Quota | Total Size | Media Type | Additional Storage Cost | Supported Protocols |
|---|---|---|---|---|---|---|
/home | No | 1.2PB | Tier 1 | Please see Cost of Resources | NFS, CIFS, GPFS | |
/project | No | 1.2PB | Tier 1 & 2 | Please see Cost of Resources | NFS, CIFS, GPFS | |
/gscratch | No | 1.2PB | Tier 1 & 2 | Please see Cost of Resources | NFS, CIFS, GPFS | |
/lscratch | No | N/A | 200GB or 1TB | N/A | N/A | N/A |
Researchers working with or generating massive data sets that exceed the default 5 TB allocation, or having significant I/O needs should consider the following options:
Rent space on shared hardware: There is a set price per TB per year. Please contact ARCC for the exact price.
Purchase additional storage disks to be incorporated into Beartooth: This option is appropriate for groups that need more space than the free offering, but don’t have the extreme space or performance demands that would require investing in dedicated hardware.
Buy your own dedicated storage hardware for Research Computing to host: If you need more than about 15 TB of storage or very high performance, dedicated hardware is more economical and appropriate. The exact choices are always evolving. Please contact ARCC for more information.
This describes ARCC's file deletion policy:
/home: Home directories will only be deleted after the owner has been removed from the university system.
/project: Project directories will be preserved for up to 6 months after project termination.
/gscratch: Files may be deleted as needed without warning if required for system productivity.
/lscratch: Files will be removed after thirty (30) days of not being used or accessed.
If users require quota increases on Beartooth file storage, they may apply for a one-time, no-cost capacity increase by providing written justification for the amount of extra storage and it’s benefit to the UW mission of research, education and economic development.
This capacity increase lasts for the life of the storage system (typically 3 years), the increase must be approved by ARCC Leadership, and is subject to a 10TB max limit.
In all other cases, the user will be charged per Cost of Resources as listed here.
Users may request capacity increases on the ARCC Service Portal.