[linux-lvm] lvmcache with vdo - inconsistent block size

Zdenek Kabelac zkabelac at redhat.com
Thu Sep 17 21:46:26 UTC 2020


Dne 17. 09. 20 v 23:41 Gionatan Danti napsal(a):
> Il 2020-09-17 21:27 Zdenek Kabelac ha scritto:
>> You've most likely found the bug and this should be likely disable
>> (and enabled only with some force option).
> 
> Hi Zdenek, I am not sure about what bug I found - can you be more explicit?
> 
>> Problem is, when such device stack is used for XFS - where the
>> 'geometry' changes, but for some other it's not a big issue (i.e.
>> ext4).

XFS doesn't like when 4K device becomes 512b sector based.


> Again, I am not sure about what we are speaking - generic thinlvm over VDO, or 
> about the caching issue discussed before? Why this should be an issue for XFS 
> vs EXT4? From RH docs, I can see VDO + thinlvm + XFS as being fully supported.

Here you are using separate level of tooling - so you have separate dealing
with errors on each device with its own tooling.

lvm2 is not yet capable to handle this stack on its own.

> 
> I'll do, but I first need to understand the core issue better.

Create XFS on 4K sector device - copy it on 512b and use it.

Zdenek





More information about the linux-lvm mailing list