Storage areas on Fram
Projects and users receive different file systems to store files and other data. Some areas are used for temporary files during job execution while others are for storing project data.
The following table summarizes the different temporary storage and project storage options. Read the following sections for specific details.
Data handling and storage policy is documented here.
||Job data. Set to
||Staging and cross-job data. Set to
||User data. Set to
$SCRATCHcontents are deleted after the job finishes execution.
$USERWORKis subject to automatic file deletion.
- See file deletion policies for both in the sections below.
- Users may request larger quota on the Project area (see Project are below)
The /cluster/work is used as a temporary work area for storing files during job execution. The work area is a high-performance parallel - Lustre - file system with a total storage space of 2.3PB. The work area is not backed up. See Backup
There are several reasons why the work areas are suitable for jobs:
- Work areas are on a faster file system than user home directories.
- There is less risk of interference from other jobs because every job ID has its own scratch directory
- Because the scratch directory is removed when the job finishes, the scripts do not need to clean up temporary files.
- It avoids taking unneeded backups of temporary and partial files, because
$SCRATCHis not backed up. More information about the automatic cleanup is found at the Prolog and Epilog.
The work area contains two subdirectories: jobs (
$SCRATCH) and users (
This area is automatically created when a job starts and deleted when the job
finishes. There are special commands (e.g.,
savefile) one can use in the job
script to ensure that files are copied back to the submit directory
sbatch was run).
- set to the
- it is automatically deleted as soon as the job finishes. Use
cleanupin the batch script to ensure files are copied back before the directory is deleted.
This directory is meant for staging files that are used by one or more jobs. All data after processing must be moved out from this area or deleted after use, otherwise it will be automatically deleted after a while (see notes below). We highly encourage users to keep this area tidy, since both high disk usage and automatic deletion process takes away disk performance. The best solution is to clean up any unnecessary data after each job.
- set to the
- File deletion depends on the newest of the creation-, modification- and access time and the total usage of the file system. The oldest files will be deleted first and a weekly scan removes files older than 42 days.
- When file system usage reaches 70%, files older than 21 days are subject to automatic deletion.
- It is not allowed to try to circumvent the automatic deletion by for instance running scripts that touch all files.
- It is not backed up. See Backup.
For performance optimizations, consult Performance Tips page.
All HPC projects have a dedicated local space to share data between project
members, located at /cluster/projects/
The project area is quota controlled and the default project quota for HPC projects is 1TB, but projects can apply for more during the application process or request at a later point in time if needed. The maximum quota for the project area is 10TB. Greater needs will require a separate application for NIRD project area.
- Daily backup is taken to NIRD.
- For backup, snapshots are taken with the following frequency:
- daily snapshots of the last 7 days
- weekly snapshots of the last 6 weeks.
- To see disk usage and quota information for your project, run
dusage -p <project_name.
- See Backup.
User Area (
The user area is mounted to /cluster/home/
stdout logs from your batch jobs in
$HOME so they are
available for reviewing in case of issues with it. The user area is
geo-replicated between Tromsø and Trondheim. Additionally daily snapshots are
taken and kept for the last 7 days and weekly snapshots for the last 6 weeks.
- set to the
- The home directory should be used for storing tools, scripts, application sources or other relevant data which must have a backup.
- backed up with daily snapshots for the last 7 days and weekly snapshots for the last 6 weeks. See Backup.