RE Array size dropped from 40TB to 7TB when upgrading to 5.10

RE Array size dropped from 40TB to 7TB when upgrading to 5.10

[Ian Kumlien]

Hi,

the same thing happened to me, but it was a raid6...

Eventually I just gave up and did a:
mdadm --grow /dev/md3 --size=max (since it reflects the size it used to be)

After a few hours of syncing, everything was ok again
(according to btrfs check which also took a few hours =))

But remember to do this with a 5.0.1 kernel ;)

Re: RE Array size dropped from 40TB to 7TB when upgrading to 5.10

[Sébastien Luttringer]

[-- Attachment #1: Type: text/plain, Size: 629 bytes --]

On Tue, 2020-12-15 at 21:08 +0100, Ian Kumlien wrote:
> the same thing happened to me, but it was a raid6...
With the same kernel (5.10) or an older one?

> Eventually I just gave up and did a:
> mdadm --grow /dev/md3 --size=max (since it reflects the size it used to be)
This doesn't removed your data?
I wondering if the flag --assume-clean is mandatory to keep the data untouched.

> After a few hours of syncing, everything was ok again
> (according to btrfs check which also took a few hours =))
> 
> But remember to do this with a 5.0.1 kernel ;)
You mean 5.10?

Regards,

Sébastien "Seblu" Luttringer


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 228 bytes --]