I also had a lot of issues with Gluster -- when it was broken, it was hard to know what was wrong. With MooseFS, the dashboard would usually give you a good id

I also had a lot of issues with Gluster -- when it was broken, it was hard to know what was wrong. With MooseFS, the dashboard would usually give you a good idea, and the logs were pretty good at showing how it was replicating data after a node or disk failed.