How much Disk Space do I Need?

On our high-performance computing facilities, we have storage resources (i.e. disk space) available for medium-term data storage. Most of this storage is persistent, rather than permanent; in effect, your project team will lease the space off us for the duration of your project.

We have some flexibility to change your storage allotment in response to a service request. However, it is most helpful to us if you can provide an estimate of your disk space needs.

Please be aware that our data storage facility is a shared resource. We ordinarily expect data not to be copied onto it until needed for a job, and to be removed from our facilities as soon as it is no longer required there. If you have a requirement for storage of data produced by finished computational work, please contact an appropriate person or team at your institution, such as your supervisor or your information technology support team.

You can work out your disk space needs on our facilities by using the following formula:

diskspace_calculator.png

where D is your total disk space requirement, R is the amount of reference data you need to keep (e.g. databases that stay the same throughout your project), J is the number of simultaneous jobs you expect to run, and for each simultaneous job i, I, O and T are the amount of input data, output data and temporary data needed for that job. Temporary data is data that is automatically deleted when the job finishes, unlike output data, which will not ordinarily be deleted without manual intervention.

For this purpose, two or more jobs may be thought of as simultaneous if you expect them to run at the same time, or so close together in time that removing the output of one before the next starts would be impractical.

If you don't know otherwise, a good starting guess is that all jobs require the same amount of data storage space.

Was this article helpful?
0 out of 0 found this helpful