One of my drives in a RAID-Z array started showing unreadable (pending) sectors. I've ordered a new drive to replace it and now I'm reviewing the steps to be taken.
The guides I've found so far assume the device is failed and offline, but in my case it's very much online, with only 2 unreadable sectors, so my question is:
Is it possible to replace the problematic drive without taking it offline first? Would a generic ZFS command 'zpool replace' work in a ssh console and not screw up anything in FreeNAS?
I'm asking because taking it offline would leave the array without any redundancy, so if during resilvering I get an URE, I'm in trouble. The 'zpool replace' command only takes the device offline after the data has been migrated off the device to be replaced.
The guides I've found so far assume the device is failed and offline, but in my case it's very much online, with only 2 unreadable sectors, so my question is:
Is it possible to replace the problematic drive without taking it offline first? Would a generic ZFS command 'zpool replace' work in a ssh console and not screw up anything in FreeNAS?
I'm asking because taking it offline would leave the array without any redundancy, so if during resilvering I get an URE, I'm in trouble. The 'zpool replace' command only takes the device offline after the data has been migrated off the device to be replaced.