My hosting provider preconfigures the servers I rent with RAID-1. However, for redundancy, I am planning to have another copy of the data on another server using GlusterFS.
Can I disable the RAID-1 to gain additional disk space since I am already going for redundancy?
The thing that'll tell you whether or not this is a good idea are three things:
If your application accesses the Gluster data through a Gluster mount, it should stay up even if one node is down. However, if you're just using Gluster as a replication mechanism and configured your app-servers to use local storage (but replicated through Gluster), if one drive goes bad that one app-server may be still serving bad data. Obviously, I recommend using a Gluster mount whenever possible.
How long it takes your provider to replace dead hardware is a key concern. It tells you how long you have to live with a bad config, and how long you'll be experiencing any pain with point 1. 4-hour response is good, next-business-day is bad if it dies Friday night.
How your systems behave during resync is something you should really test. It depends a lot on how much data you've got to move from point-good to point-repaired. Your Gluster-mount I/O is likely to be bad during this time, so be sure you can withstand this state.
If you have all three states taken care of, R0 may be survivable for you. If your failed-state performance is unacceptable, you'll probably have to use a higher RAID level for your base Gluster bricks.
It's all about availability. If you're OK with failing over to your other server(s) in the event of a single disk failure and you have the infrastructure for that, then go for it.
Generally, I like to layer my availability. I don't want my applications to fail over because of a single disk failure. It's really up to you, though. I'd make sure I tested this thoroughly, though - disk failures are relatively common.
Since your edit, the below is no longer applicable.
Your question is essentially invalid, because RAID 0 has no redundancy and has the maximum amount of disk space possible. It is dangerous and bad for availability. If you lose a single disk in the set, all data is gone. If you plan on keeping RAID 0, I'd test my failover procedure pretty thoroughly.
