Extracting 100% of Data From a Stubborn, Dying ZFS Pool
This title could be clearer and more informative.Try out Clickbait Shieldfor free (5 uses left this month).
A personal account of recovering 100% of data from a decade-old ZFS pool on failing hard drives. The drives caused random I/O lockups that left rsync stuck in uninterruptible sleep (D+ state), immune to kill signals and even shutdown commands. The author automated the recovery process using a bash script to detect hung transfers by monitoring rsync output, combined with a TP-Link Kasa smart plug controlled via IFTTT webhooks to perform automated hard reboots remotely. After 151 automated power cycles over two days, all data was successfully transferred. The post also reflects on a better approach: using dd to image the raw drives first, then importing them as a ZFS pool and using zfs send | zfs recv for a local transfer.
