On Thu, Aug 29, 2013 at 01:37:51PM -0600, Chris Murphy wrote:
> 
> On Aug 29, 2013, at 11:35 AM, Zach Brown <[email protected]> wrote:
> 
> >>   If those fail, then look in dmesg for errors relating to the log
> >> tree -- if that's corrupt and can't be read (or causes a crash), use
> >> btrfs-zero-log.
> > 
> > In a bit of a tangent:
> > 
> > btrfs-zero-log throws away data that fsync/sync could have previously
> > claimed was stable on disk.
> > 
> > Given how often this is thrown around as a solution to a broken
> > partition, should the tool jump up and down and make it clear that it's
> > about to roll the file system back?  This seems like relevant
> > information.
> > 
> > Right now, as far as I can tell, it's completely undocumented and
> > silent.
> 
> Yes, I think it helps remove some burden on the list answering questions 
> about a tool that doesn't have any documentation, to have a warning.
> 
> How much longer will btrfs-zero-log be needed? If whatever it's doing isn't 
> obviated by future improvements to btrfsck, and this sort of big hammer 
> approach is still needed in some worse case scenarios, then it probably hurts 
> no one to flag the user with essentially how you described it. I think 
> documentation is a greater burden to create, and less likely to be consulted.
> 
> "Proceeding will roll back the file system to a previous state, and may cause 
> the loss of successfully written data. Proceed? (Y/N)"

   "... the loss of up to the last 30 seconds of successfully written data."

   Give the user enough information to make a sensible decision.

   Hugo.

-- 
=== Hugo Mills: hugo@... carfax.org.uk | darksatanic.net | lug.org.uk ===
  PGP key: 65E74AC0 from wwwkeys.eu.pgp.net or http://www.carfax.org.uk
        --- emacs:  Eighty Megabytes And Constantly Swapping. ---        

Attachment: signature.asc
Description: Digital signature

Reply via email to