Storage areas on HPC clusters

Projects and users receive different areas to store files and other data. Some areas are used for temporary files during job execution while others are for storing project data.

Data handling and storage policy is documented here.

The following tables summarizes the different storage options for the two clusters. Read the following sections for specific details.

  • Fram
Directory Purpose Default Quota Backup
/cluster/home/$USER ($HOME) User data 20 GiB / 100 K files Yes
/cluster/work/jobs/$SLURM_JOB_ID ($SCRATCH) Per-job data N/A No
/cluster/work/users/$USER ($USERWORK) Staging and job data N/A No
/cluster/projects/<project_name> Project data 1 TiB* / 1 M files Yes
/cluster/shared/<folder_name> Shared data Individual** No
  • Saga
Directory Purpose Default Quota Backup
/cluster/home/$USER ($HOME) User data 20 GiB / 100 K files Soon
/cluster/work/jobs/$SLURM_JOB_ID ($SCRATCH) Per-job data N/A No
/cluster/work/users/$USER ($USERWORK) Staging and job data N/A No
/cluster/projects/<project_name> Project data 1 TiB* / 1 M files Soon
/cluster/shared/<folder_name> Shared data Individual** No

Note that on Saga, backup of $HOME and /cluster/projects/<project_name> has not been set up yet, but will soon be implemented.

In addition to the areas in the tables above, both clusters mount the NIRD project areas as /nird/projects/nird/NSxxxxK on the login nodes (but not on the compute nodes). (This has not been implemented on Saga yet, but will happen soon.)

The /cluster file system is a high-performance parallel file system. On Fram, it is a Lustre system with a total storage space of 2.3PB, and on Saga it is a BeeGFS system with a total storage space of 1.1PB.

For performance optimizations, consult Performance Tips page.

Home Directory ($HOME)

The home directory is /cluster/home/$USER. The location is stored in the environment variable $HOME. A quota is enabled on home directories which is by default 20GiB and 100,000 files, so it is not advisable to run jobs in $HOME. However, it is perfectly fine to store stderr and stdout logs from your batch jobs in $HOME so they are available for reviewing in case of issues with it.

$HOME is only accessible for the user. Files that should be accessible by other uses in a project must be placed in the project area.

Notes

  • The home directory should be used for storing tools, scripts, application sources or other relevant data which must have a backup.
  • Not implemented on Saga yet Backed up with daily snapshots for the last 7 days and weekly snapshots for the last 6 weeks. See Backup.

Job Scratch Area

Each job gets an area /cluster/work/jobs/$SLURM_JOB_ID that is automatically created for the job, and automatically deleted when the job finishes. The location is stored in the environment variable $SCRATCH available in the job. $SCRATCH is only accessible by the user running the job.

The area is meant as a temporary scratch area during job execution. The area is not backed up. See Backup.

There are special commands (savefile and cleanup) one can use in the job script to ensure that files are copied back to the submit directory $SLURM_SUBMIT_DIR (where sbatch was run).

Pros of Running Jobs in $SCRATCH

  • 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.

Cons of Running Jobs in $SCRATCH

  • Since the area is removed automatically, it can be hard to debug jobs that fail.
  • One must use the special commands to copy files back in case the job script crashes before it has finished.
  • If the main node of a job crashes (i.e., not the job script, but the node itself), the special commands might not be run, so files might be lost.

User Work Area

Each user has an area /cluster/work/users/$USER. The location is stored in the environment variable $USERWORK. The area is not backed up. See Backup. $USERWORK is only accessible by the user owning the area.

This directory is meant for files that are used by one or more jobs. All result files must be moved out from this area after the jobs finish, otherwise they 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. Since $USERWORK is only accessible by the user, you must move results to the project area if you want to share them with other people in the project.

Notes:

  • 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.

Pros of Running Jobs in $USERWORK

  • Since job files are not removed automatically directly when a job finishes, it is easier to debug failing jobs.
  • There is no need to use special commands to copy files back in case the job script or node crashes before the job has finished.

Cons of Running Jobs in $USERWORK

  • There is a risk of interference from other jobs unless one makes sure to run each job in a separate sub directory inside $USERWORK.
  • Because job files are not removed when the job finishes, one has to remember to clean up temporary files afterwards.
  • One has to remember to move result files to the project area if one wants to keep them. Otherwise they will eventually be deleted by the atuomatic file deletion.

Project Area

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 an application for a separate NIRD project area.

Notes:

  • Daily backup is taken to NIRD. Not implemented on Saga yet!
  • 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.

Pros of Running Jobs in Project Area

  • Since job files are not removed automatically directly when a job finishes, it is easier to debug failing jobs.
  • There is no need to use special commands to copy files back in case the job script or node crashes before the job has finished.
  • There is no need to move result files to save them permanently or give the rest of the project access to them.

Cons of Running Jobs in Project Area

  • There is a risk of interference from other jobs unless one makes sure to run each job in a separate sub directory inside the project area.
  • Because job files are not removed when the job finishes, one has to remember to clean up temporary files afterwards, otherwise they can fill up the quota.
  • There is a risk of using all of the disk quota if one runs many jobs and/or jobs needing a lot of storage at the same time.

Shared Project Area

In special cases there might be a need for sharing data between projects for collaboration and possibly preventing data duplication.

If such a need is justified, a meta-group and it's according directory in /cluster/shared is created. The area gets a disk quota based on the needs of the data. The permissions of the areas vary. In some cases all but a few users in the meta-group only have read-only access to the area. In other cases, all users on the cluster have read-only access.

results matching ""

    No results matching ""