Cache file list in daemon mode?

[Peter Scott]

On 8/2/2012 7:30 PM, Jason Haar wrote:
> So what you're really saying is gluster is quite slow at doing recursive
> directory listings, so how about just using "find" on the real backend
> bricks to find the files that have changed since last run, merge those
> listings together (to get rid of dupes) and then get rsync to just
> update those against the gluster service?
>
Yes, that's pretty much the solution we're converging on.  It's probably 
20 lines max*, most of which is anticipating race conditions, but I'm 
congenitally lazy* and thought I might not be the first person with 
these requirements - not getting as specific as Gluster, for sure, but 
still with enough files and slowness on the target to raise the question 
of why not operate in some sort of incremental mode.  I dislike having 
to think of every possible race condition.

* = Perl