Performance

Since data is distributed across multiple nodes, we can also have multiple clients accessing the cluster at the same time. This process of accessing data from multiple sources simultaneously is called parallelism, and GlusterFS allows for increased performance by directing clients to different nodes. Additionally, performance can be increased by adding bricks or nodes—effectively, by scaling horizontally or vertically.