High-sensitivity fs checker (not repairer) for btrfs

High-sensitivity fs checker (not repairer) for btrfs

[Bob Marley]

Hello all
I would like to know if there exists a tool to check the btrfs 
filesystem very thoroughly.
It's ok if it needs the FS unmounted to operate. Also mounted is OK.
It does not need repair capability
It needs very good checking capability: it has to return Good / Bad 
status with the "Bad" meaning that there is at least ONE inconsistency. 
Good means that it is really really 100% consistent.

Does something like this exists?

We need to detect as much ahead of time as possible if the btrfs 
filesystem has become even just a little bit inconsistent

Thank you

Re: High-sensitivity fs checker (not repairer) for btrfs

[Hugo Mills]

[-- Attachment #1: Type: text/plain, Size: 1011 bytes --]

On Sat, Nov 10, 2012 at 10:18:42PM +0100, Bob Marley wrote:
> Hello all
> I would like to know if there exists a tool to check the btrfs
> filesystem very thoroughly.
> It's ok if it needs the FS unmounted to operate. Also mounted is OK.
> It does not need repair capability
> It needs very good checking capability: it has to return Good / Bad
> status with the "Bad" meaning that there is at least ONE
> inconsistency. Good means that it is really really 100% consistent.
> 
> Does something like this exists?
> 
> We need to detect as much ahead of time as possible if the btrfs
> filesystem has become even just a little bit inconsistent

   The closest thing is btrfsck. That's about as picky as we've got to
date.

   What exactly is your use-case for this requirement?

   Hugo.

-- 
=== Hugo Mills: hugo@... carfax.org.uk | darksatanic.net | lug.org.uk ===
  PGP key: 515C238D from wwwkeys.eu.pgp.net or http://www.carfax.org.uk
         --- 2 + 2 = 5,  for sufficiently large values of 2. ---         

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 828 bytes --]

Re: High-sensitivity fs checker (not repairer) for btrfs

[Bob Marley]

On 11/10/12 22:23, Hugo Mills wrote:
>     The closest thing is btrfsck. That's about as picky as we've got to
> date.
>
>     What exactly is your use-case for this requirement?

We need a decently-available system. We can rollback filesystem to 
last-known-good if the "test" detects an inconsistency on current btrfs 
filesystem, but we need a very good test for that (i.e. if 
last-known-good is actually bad we get into serious troubles).

So do you think btrfsck can return a false "OK" result? can it "not-see" 
an inconsistency?

Thank you

Re: High-sensitivity fs checker (not repairer) for btrfs

[cwillu]

On Sat, Nov 10, 2012 at 4:32 PM, Bob Marley <bobmarley@shiftmail.org> wrote:
> On 11/10/12 22:23, Hugo Mills wrote:
>>
>>     The closest thing is btrfsck. That's about as picky as we've got to
>> date.
>>
>>     What exactly is your use-case for this requirement?
>
>
> We need a decently-available system. We can rollback filesystem to
> last-known-good if the "test" detects an inconsistency on current btrfs
> filesystem, but we need a very good test for that (i.e. if last-known-good
> is actually bad we get into serious troubles).

Scrub is probably more useful as a check, combined with "does the
filesystem actually mount".

> So do you think btrfsck can return a false "OK" result? can it "not-see" an
> inconsistency?

No set of checks will ever be perfect, so yes.