There are many ways to improve GlusterFS performances, but I will mention here two ways:

- Increasing "performance.cache-size", this is done by GlusterCLI (for example, increasing to 128MB):

> volume set VOL_NAME performance.cache-size 128738368

- Changing mount option to NFS, instead of native FUSE:

mount -t nfs -o _netdev,noatime,vers=3,direct-io-mode=disabled SERVER:/VOL_NAME /MOUNT_POINT

The key success is to measure performances, and after applying those two changes here are improvements on a Wordpress website, with RDS as a database.

BEFORE:

Before GlusterFS changes

 

AFTER:

After GlusterFS changes