Homelab Hero · No. 007: your result

Pool's gone. 0 Hero Points.

RAIDZ1 survives exactly one dead disk, and the first one is already dead. The erroring disk is erroring, but it is still serving reads the resilver needs — pull it and the pool faults on the spot, before any replacement matters. Replacing both at once is a fine plan for a pool with a spare parity disk; this pool does not have one.

The lesson: In RAIDZ1, an erroring disk that still reads is load-bearing. Never pull the second disk.

Share your result:
Homelab Hero No. 007 — I blew the call on the mid-resilver second-disk call (0/200). Think you would call it better? Play it: https://homelabherald.com/hero/007?s=0

Share it: Post to X WhatsApp Text Reddit Facebook LinkedIn Email

Tomorrow’s incident drops with the next issue at 7:00 AM PT.

Play it again · Read Issue No. 007