Re: ZFS Permanent error <0xffffffffffffffff>:<0x0>

> ----------------------------------------
> From: Andrea Venturoli <ml@netfence.it>
> Sent: Sat Dec 31 11:41:39 CET 2022
> To: <questions@freebsd.org>
> Subject: Re: ZFS Permanent error <0xffffffffffffffff>:<0x0>
> 
> The RAM alas... just removed a faulty module.
> So I hope I don't get any more troubles, even if I don't know how to 
> clear/fix the old one.
> 

Recreating the pool is what ultimately cleared it for me: Attached a backup
disk, booted into a livecd, took a recursive snapshot of zroot, sent it to the
backup disk, destroyed the old pool, recreated it, sent the backed-up datasets
to the old disk.

I initially did the waiting game like you, but then started getting CAM errors
that made the system unusable. I tried refreshing the disk, but that didn't
help:

dd if=/dev/ada0p3 of=/dev/ada0p3 bs=1m

The ZFS metadata errors vanished after recreating the pool. It's about an
hour-long project.

A recursive snapshot of zroot sends everything over with a single command:
livecd# zfs send -vR zroot@migration | zfs recv -vFud new_zroot

And then back again after recreating the pool:
livecd# zfs send -vR new_zroot@migration | zfs recv -vFud zroot

Just have to make sure you're using a livecd with a ZFS version that's equal or
newer.  After all the hassle, I did contemplate setting up a mirror raid.


-- 
Sent with https://mailfence.com  
Secure and private email