We're running v3.2.3 and we have a failing piece of hardware that runs a brick (brick A). I have another brick (Brick B) in the volume that has some data on it, but has enough spare room for the data that's on brick A. If I run replace-brick like: gluster volume replace-brick myvolume server1:/brick-a server2:brick-b start - would the data currently on brick B be wipeed? - would any new data written to the volume *not* be written to brick A? (ie: does brick A go into a sort of read-only mode?) If the data in brick B is still ok after the migration I plan on then decommissioning brick A. Thanks, Scott Barber iMemories.com
