brad at danga.com
Thu Apr 3 03:24:55 UTC 2008
I'm against it because it's deceptive. It makes people think it does
something and it doesn't, really.
If you see a measurable effect, I'd love to see numbers and your
On Wed, Apr 2, 2008 at 12:46 PM, Paul Baker™ <bakerp at google.com> wrote:
> Who doesn't have a weighted random patch? Mine's at
> On Wed, Apr 2, 2008 at 6:27 AM, Jonty <jonty at last.fm> wrote:
> > On Tue, 2008-04-01 at 15:51 -0700, Mark Smith wrote:
> > > > AFAIK, perlbal actually only supports the "random" load-balancing
> > algorithm.
> > > > I've never actually tried using round-robin, but the last time I was
> > in the
> > > > source-code I thought I remember there being code to throw an error
> > if
> > > > anything other than "random" was specified.
> > >
> > > Yeah, it's random. At one point I wrote a patch to make weighted
> > > random, but never got it committed...
> > I've got a similar patch to send upstream, I'll see if I can make some
> > time for that. I also implemented normal round-robin, just for the hell
> > of it.
> > --jonty
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the perlbal