I think you’re got the basics down correctly.
RAID5 (of four disks) – All four drives are running, reading, and writing. If you lose a disk, it’ll go into “DEGRADED” mode, but all data is still available and Read/Write normally.
RAID5+Spare (of four disks) – THREE drives are running, reading, and writing in a 3-Drive RAID5 array. If you lose a drive, and automatic rebuilds are enabled, the volume goes DEGRADED, the SPARE drive becomes ACTIVE and the array rebuilds into that spare drive. It goes back to “Healthy” once the rebuild is completed.
When you replace the failed drive, it becomes the spare.
So, you’re sacrificing the space the 4th drive gives you, but you’re assured that the array is left in DEGRADED mode for as short a time as possible.
But if you have access to replacement drives fairly quickly (like within a few weeks) and you’re using RAID5 (with no spare), leaving it in DEGRADED mode for that long while you wait for a replacement isn’t a huge risk. Especially since you’re also backing it up to an outside service.
RAID6 is different than RAID5 in that TWO drives are dedicated to Parity data, whereas in RAID5, only one is.
So if you’re doing RAID6, you’d have the same capacity is RAID5+SPARE, but it can survive the concurrent failure of TWO disks before the volume is unavailable.
