Re: svn commit: r34088 - trunk/subversion/libsvn_fs_fs

On Fri, Nov 7, 2008 at 11:42 AM, Hyrum K. Wright
<hyrum_wright_at_mail.utexas.edu> wrote:
> David Glasser wrote:
>> I'd like to keep it off by default until I see a solid case made with
>> benchmarks etc that the space gain is more noticeable than the speed
>> loss.
>
> I'm curious as to how we can quantitatively compare space gain vs. speed loss.
> It's a tradeoff to be sure, but doing a direct comparison seems like
> apples-to-oranges. Have you run Mike's script on the new, fsfs-disabled trunk
> repo? (See http://svn.haxx.se/dev/archive-2008-10/0985.shtml)

FWIW, Size is a huge factor for us at freebsd.org. Right now the repo
fits 100% in ram on our servers. We've seen as much as an order of
magnitude (or more) slowdown as a consequence of hitting disk while
bringing things up. A 3.5GB working set is fairly easy to manage on
current servers, but I'd hate to see a massive scaling up of size.

-- 
Peter Wemm - peter_at_wemm.org; peter_at_FreeBSD.org; peter_at_yahoo-inc.com; KI6FJV
"All of this is for nothing if we don't go to the stars" - JMS/B5
"If Java had true garbage collection, most programs would delete
themselves upon execution." -- Robert Sewell
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe_at_subversion.tigris.org
For additional commands, e-mail: dev-help_at_subversion.tigris.org