TrueNAS ZFS Storage Architecture: Vdev Layout, Capacity and Failure Domains
Build predictable TrueNAS storage by aligning ZFS vdev design, capacity planning, protection level and workload behaviour.
Core design principles
Choose vdev layout first. Mirrors favour random I/O and predictable latency; RAIDZ layouts favour capacity efficiency but have different performance and expansion characteristics.
Avoid running pools near capacity. Snapshot growth, copy-on-write behaviour and resilver operations need headroom; establish alerts and a capacity threshold before production use.
Separate workload classes when practical. Virtual machines, backup targets and sequential archive data have different I/O patterns and should not compete blindly.
Operational checklist
Define ownership, document the architecture and dependencies, protect privileged access, monitor the service, test recovery or rollback, and review the design after every material change.