Yes, the data layout is EXACTLY the same.
See below, where Dx means Data chunk x.
Code:
RAID 1+0. A stripe of mirrors:
( D1 ) ( D2 ) ( D3 )
( | ) ( | ) ( | )
(mirror)-stripe-(mirror)-stripe-(mirror)
( | ) ( | ) ( | )
( D1 ) ( D2 ) ( D3 )
Code:
RAID 0+1. A mirror of stripes:
(D1-stripe-D2-stripe-D3)
|
mirror
|
(D1-stripe-D2-stripe-D3)
Imagine the top D1 and the bottom D2 are broken.
RAID1+0 can still read data.
RAID0+1 can't read data (officially.)
What I'm saying is, both have exactly the same data layout, with exactly the same failures. If I was writing the firmware for an array controller, I would make damn sure it could read data whenever possible, not when the technical definitions say you are "allowed" to see data.
(Cheers for the hint below Kez :) )