Make informed choices with OpenZFS

Start with the decision in front of you

ZFS Storage Hub is a reading guide for people planning or maintaining an OpenZFS system. Start with the question you need to answer, then follow the relevant article to its documentation and examples. The guides below connect the choices made before buying disks with the work of looking after an existing installation. There is no requirement to read them in publication order.

Pool design: turn a workload into a plan

Write down what the system will store, how much room it needs now, and how you expect it to grow. Use the parity guide for vdev boundaries and fixed layout choices, then put the proposed disk count into the calculator. For the broader mirrors comparison, the linked TrueNAS Guide article is the network's reference. Keeping these questions separate makes it easier to compare two plans without confusing capacity with recovery or performance.

• ZFS and RAIDZ capacity calculator
• RAIDZ1 vs RAIDZ2 vs RAIDZ3
• RAIDZ vs mirrors on TrueNAS Guide

Parity: compare the recovery tradeoffs

The RAIDZ parity comparison puts the three levels in one table and follows the choice through a device failure, a replacement and future expansion. Read it before spending the last available bay on capacity. It also explains where support devices fit in the design, which settings stay attached to a vdev, and what belongs in a scrub and backup plan.

Hardware: budget memory and check the disk path

Use the ZFS hardware and RAM guide when choosing a host or reviewing an existing parts list. It separates workload memory planning from storage capacity, explains the often repeated RAM-per-TB shortcut, and covers controllers, sector alignment and drive recovery behavior. If the filesystem itself is still undecided, the ZFS vs Btrfs comparison provides a separate starting point for that choice.

ARC: understand an installed system's cache

Already have a host and want to understand its memory graph? The ZFS ARC size guide starts with the current configuration and memory pressure before discussing limits. Follow that route for inspection commands, distribution differences and rollback steps. Return to the hardware guide when the question is how much RAM to purchase, so a runtime cache setting does not become an accidental buying rule.

Troubleshooting: read the evidence before acting

When a pool reports a problem, start with reading a DEGRADED zpool status. It shows an example status block and explains the device states, error counters, replacement identifiers and checks after a resilver. Keep the complete output alongside the recovery plan, then use the hardware and parity guides for the longer-term decisions behind the incident. Browse topics or the tag index to find the rest of the published coverage.