Bay count decides which redundancy levels an enclosure can offer and how far it can grow before it has to be replaced. Those two consequences matter more than the capacity the chassis reaches on the day it is bought.
Bay count is the one specification on an enclosure that cannot be changed afterwards, which is why it deserves more thought than the processor or the memory.
What each bay count makes possible
One bay is a single drive on the network. It has no redundancy of any kind, and everything in it depends on that drive continuing to work.
Two bays is the first count that can mirror. It can also run the two drives as separate volumes, or join them into one larger one with no redundancy at all — three arrangements from the same chassis, and the maker names which it supports.
Four bays is where the parity levels become sensible. Parity spreads redundancy across the members instead of duplicating a drive, so a larger share of what is installed remains usable, and the count is high enough that losing one member still leaves a working array.
More than four buys the option of surviving two simultaneous failures, and it buys expansion room. It also buys noise, heat and a power supply that has to feed all of it.
Usable space is not the sum of the drives
This is the arithmetic that surprises people, and it belongs in the decision rather than after it. A mirrored pair presents the capacity of one drive. A parity array gives up the equivalent of one member; a double-parity array gives up two. The enclosure’s own maker publishes a calculator for its levels, and that published figure is the one to plan around.
Note the unit while you are there. Drive makers count capacity in decimal terms and enclosures often display in units of 1024 with a different suffix, so the array will report a smaller-looking figure than the drives advertise. Nothing has been lost; the two are separate conventions, and this site quotes the maker’s figure in the maker’s unit.
Growth is the argument for one more bay
An array that is full is a problem with only expensive answers. Buying a chassis with a bay you do not fill on day one costs the price of an empty bay; discovering later that you needed it costs a new chassis, a migration and the drives.
Against that: every bay is a drive that will eventually be bought, spun and powered. A four-bay enclosure with two drives in it is a sensible position. A four-bay enclosure bought because eight felt excessive, then filled within a year, is the position to avoid.
What else the bay count drags along
The chassis grows with the count, and so does everything attached to it. Cooling — more drives need more moving air, and the fan is the part you hear. Power — the supply is sized for spin-up of every drive at once, which is the peak, and it is also what a UPS has to carry. Drive format — some enclosures take 3.5-inch drives, some take 2.5-inch, and some take both with the adapter the maker supplies.
The line to read before ordering
Find the maker’s compatibility list and its maximum capacity per bay. Both are published, both are specific to the model, and both are cheaper to read than to discover. A chassis that lists neither is telling you something about how much documentation you will get later.
The questions that come up before an order
Can I start with one drive and fill the other bays later?
Most enclosures allow it, and the maker states which levels can be expanded in place and which cannot. Read that line before buying rather than after: some arrangements accept a new drive and grow, others require the array to be rebuilt from scratch, and the difference is a weekend.
Do the drives in an enclosure have to be identical?
They have to be compatible with the enclosure, and within a redundant array the usable size is governed by the smallest member. Makers publish a compatibility list of drives they have qualified, and some publish a maximum capacity per bay that is lower than the largest drive on sale.
Last reviewed 10 September 2026