Skip to content

Storage System

We store the raw simulation data on on block storage drives (ie. hard drives) connected to the computational clusters. The present page explains some of the related conventions. The policies applied to the information in the form of unstructured data are explained further in this section of the documentation.

Storage Diagram

A representative example of an overall storage system, such as implemented on our infrastructure, is explained visually in the flowchart below.

Storage System

The above example includes multiple clusters connected to the same central Login Node, each one with its own corresponding main access Master Node and associated storage space. Here, we also show (through color labelling) how each Cluster Home directory present under the corresponding Master Node filesystem is mapped (mounted) to the Login Home.

For example, the home folder for "cluster-001" under the path /home/<username> on the Master Node is mapped to the directory path /cluster-001-home/<username> under the Login Node. A similar pattern is applied for the Shared Folders for Organizations.

Storage Quotas

These storage spaces are subject to certain quotas, limiting the size of the data that may be saved in them. These quotas are documented here.