1U NAS Server -- baseline storage tier. The device most other systems (hosted games, crypto rigs, client contracts) depend on indirectly, since they all need somewhere to actually put data.
8 disks, 64TB usable, RAIDZ2, 75000 IOPS rating, 16GB cache out of the box -- solid general-purpose numbers, no standout specialty in either direction.
No built-in redundancy against a real failure or intrusion out of the box (0% built-in redundancy chance, the lowest of the three NAS-category devices) -- pair it with a genuine backup mirror share (see `man nas`) if the data on it actually matters, or buy a RAID Controller upgrade (one-time, real redundancy chance) from this device's own panel.
A share on this device with real used blocks and no configured backup mirror is exactly the scenario the disaster-recovery system punishes: a hardware fault or a landed intrusion on this NAS wipes that share's data outright, with a wallet activity-log line marking the loss.
Tip: this is the tier to use for low-stakes storage (scratch space, a game server you don't mind losing progress on) and to explicitly NOT use unmirrored for anything a contract's reputation payout depends on.
Total capacity is locked, not free-editable -- it only grows via a real Extra Disk purchase (repeatable, +20TB each, up to 10) from this device's own panel or the top bar's Upgrade Console.
See also: the Game Systems manual's Data Management section for backup mirrors and disaster recovery, and Hardware Upgrades & the Upgrade Console for the full upgrade system.
Takes the Redundant PSU upgrade ($1800, Inspector → Upgrades): it adds a second power input, PSU2. Wire PSU2 to a DIFFERENT PDU or UPS than PSU1 and this device keeps running when either one dies. Two feeds from the SAME PDU protects nothing -- the redundancy panel simulates that PDU failing and watches the device fail with it regardless of cable count. The spare idles at zero draw, so it costs nothing extra on the power bill. See Redundancy & Failover.