Re: [Netatalk-devel] Fix for bug #477 (race condition btw. dbd and cnid_dbd)

[HAT <ha...@fa...>]

Sorry.
It's not reproduced in my environment. Since my environment is poor,
the experiment for large volume cannot be performed.

Frank Lahm <fra...@gm...>, 11 Dec 2012 11:02:17 +0100
>would you be able to test the fixes I pushed to the branch 'dbd' on github?
>It fundamentally changes the way `dbd` works, please refer to the commit
>messages for details.
>While I was at it, I made a modification to the way we generate new CNIDs in
>cnid_dbd, so the changes are not only affecting dbd. I would have refrained
>from doing so (changing things that are not related to the thing that needs to
>be fixed), but the change is necessary for correct cooperation of dbd and
>cnid_dbd. It also finally gives us a CNID database that doesn't say "sorry,
>game over" when the last used CNID reached UINT32_MAX.

-- 
HAT