[Box Backup] Impact of the bug #63 (hardlinks not supported)

Dmitry Astapov dastapov at gmail.com
Mon Dec 13 19:47:50 GMT 2010


On Mon, Dec 13, 2010 at 7:25 PM, Dmitry Astapov <dastapov at gmail.com> wrote:

>
>
> On Mon, Dec 13, 2010 at 5:12 PM, Thomas Capricelli <orzel at freehackers.org>wrote:
>
>>   Hi,
>>
>> I'm also having lot of such repositories (mercurial) and I would also be
>> interested in a (somewhat) official answer to this question.
>>
>> regards,
>>
>
> Ok. I found this thread from 2008, which seems highly relevant:
> http://lists.warhead.org.uk/pipermail/boxbackup/2008-April/004338.html
>
> Am I right that at least the following info in that thread (mail +
> followups) is still current: no hardlinks, no directory mtimes fixup on
> restore?
>

My experiments show that hardlinks are not supported, but nothing bad
happens with mercurial and darcs repos when you restore them (except for
code bloat).

And yes, mtimes for directories are not preserved. Bummer :(

I wish there were columns like "Supports hardlinks" and "Restores all
metadata" in http://www.boxbackup.org/trac/wiki/BoxComparison :)


>  --
>>
>> Thomas Capricelli <orzel at freehackers.org>
>>
>> http://www.freehackers.org/thomas
>>
>> On Monday 13 December 2010 15:54:49 Dmitry Astapov wrote:
>>
>> > Hello,
>>
>> >
>>
>> > I'm a new boxbackup user, and I have a massive ammount of warnings
>> desribed
>>
>> > in #63.
>>
>> >
>>
>> > Why do I have a lot of hardlinks? Because distributed version control
>>
>> > systems like darcs or hg, and, to some extent, git use hardlinks when
>>
>> > creating local branches of repositories to share history and patches
>> between
>>
>> > them. I have a lot of such development repositories - hence lots of
>>
>> > warnings.
>>
>> >
>>
>> > Now, bug #63 says that this warning should not be ignored - data
>> corruption
>>
>> > is imminent. However, specifics of the corruption are not described. My
>>
>> > guess is that only one copy of the hardlinked file is backed up, so when
>>
>> > restoring I would end up with just a single copy of the fire in one of
>> the
>>
>> > places - am I right?
>>
>> >
>>
>> > Does this mean that boxbackup should be avoided when hardlinks are
>> present
>>
>> > or not?
>>
>> >
>>
>> >
>>
>>
>> _______________________________________________
>> Boxbackup mailing list
>> Boxbackup at boxbackup.org
>> http://lists.boxbackup.org/cgi-bin/mailman/listinfo/boxbackup
>>
>>
>
>
> --
> Dmitry Astapov
>



-- 
Dmitry Astapov
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.boxbackup.org/pipermail/boxbackup/attachments/20101213/8bf4cafb/attachment.html>


More information about the Boxbackup mailing list