Size a RustFS erasure set before you deploy

Model raw capacity, usable capacity, stripe layout, and failure tolerance from the hardware shape you plan to run.

Servers

Total drives

128

Raw capacity

1.0 PiB

Cluster input

Hardware first, then choose the erasure layout.

ServersDrives / serverDrive TiB

Stripe size (K + M)

Drives per erasure set. Larger stripes improve efficiency when the hardware shape allows it.

Parity (M)

Higher parity increases fault tolerance and reduces usable capacity.

Data shards

Parity shards

Erasure set

16 drives

Recommended layout

Stripe 16, parity 4. Balanced for availability and efficiency.

Production check

Failure domainrack / node

Rebuild headroomcapacity + I/O

Ops baselinealerts + spares

Need a sizing review?

Share your node count, drive model, expected growth, and failure-domain target with the RustFS team before ordering hardware.

Usable data after parity reservation.

Usable capacity

768 TiB

Capacity split75% usable

Raw

1.0 PiB

Efficiency

75%

Drive failure tolerance

32 drives across cluster

4 out of 16 drives per stripe

Server failure tolerance

2 servers across cluster

2 out of 8 servers per shard

Results are rounded down to quorum-safe limits and should be treated as planning estimates before production sizing.