Storage¶

We have 3 storage systems for you: $HOME (/home/<NetID>), $SCRATCH (/scratch/<NetID>), and $ARCHIVE (/archive/<NetID>).

In short, you should

Put all your data to $SCRATCH and run your jobs from there.
Only a small persistent fraction to $HOME (e.g., source code, executables).
For long-term storage, archive them to $ARCHIVE.
$ARCHIVE is not visible on compute nodes but mountable on your local workstation using this link, it is best suited to quick post-processing, analysis and visualization, without moving your data.

Important

Backing up is a user’s own responsibility. E.g., if a user deletes something accidentally, we can not recover, unfortunately.

Caution

Running jobs from /home is a serious violation of HPC policy. Any users who intentionally violate this policy will get their account suspended. $HOME SSDs are not designed for running jobs, it will kill the SSDs quickly.

Summary¶

	$HOME	$SCRATCH	$ARCHIVE
Use for storing	source code / executables	data	anything
Accessible From	login / compute	login / compute	login
Use to Run Jobs	No	Yes	No
Retention Time (Days)	No Limit	90	No Limit
Mountable	No	No	Yes
Default Quota	20GB, 150K Files	5TB, 500K Files	No Limit

Know Your Quota¶

Run myquota command in the terminal on the HPC to check your current usage and quota. Example output:

                   DISK SPACE                # FILES (1000's)
filesystem       size      quota            number      quota
            --------------------------   --------------------------
     /home     1406MB       20GB (  7%)        17       150 ( 12%)
  /scratch       70GB     5000GB (  1%)        93       500 ( 19%)
  /archive        8KB     5120GB (  0%)         0       125 (  0%)

Best Practices¶

Dos	Remarks
Periodically clean your `/scratch`	Files which have not been accessed for 90 days in `/scratch` are deleted.
Once a project is completed move the data over to `/archive` using this link	Moving data to `/archive` frees up space from `/scratch` and avoids deletion of files if older than 90 days.
Use tar files to archive directories with large file count	Lesser the number of files, faster is the archiving and dearchiving process