[sheepdog] [PATCH v4 2/2] sheep: implement atomic create and write mechanism for config

Hitoshi Mitake mitake.hitoshi at gmail.com
Wed Apr 17 07:22:40 CEST 2013

At Wed, 17 Apr 2013 13:17:10 +0800,
Liu Yuan wrote:
> On 04/17/2013 01:14 PM, Hitoshi Mitake wrote:
> > I agree with writing the function if there are two or more users. But
> > the abstraction will be a little bit difficult because of its nature:
> > we have to clean up temporal files during the initialization sequence.
> > Do you have an idea about this problem?
> > 
> We can assume the temporal file will be strictly named path.tmp. So the
> cleaning of path.tmp is up to the caller. This is no problem because we
> use it internally, it is not a library function used externally.

OK, let's delegate the cleaning to callers. I'll prepare the patch.


More information about the sheepdog mailing list