Volume type

The volume chosen affects the performance in different ways, since GlusterFS allocates data differently for each type.

For example, a replicated volume mirrors data across nodes, while a dispersed volume tries to maximize node usage and uses them in parallel.

If performance is the primary aim for a dispersed or distributed volume, consider that distributed volumes offer no redundancy, while a dispersed volume does it at the expense of performance degradation.