On Sun, 24 Oct 1999, Dave Wreski wrote:
> > On Sat, Oct 23, 1999 at 01:47:27PM -0400, Dave Wreski wrote:
> > > Yes, I'm actually suprised to see M$ using round-robin for their boxes.
> > > You'd think they'd use something like Cisco's Local Director provide real
> > > load balancing and failover...
> >
> > That wont work for locally distributed server farms, since the redirector
> > box is a single point of failure. Using round-robin to multiple local
> > redirectors is better for distributing the load over differen tproviders.
>
> Well, yes, that's true. But with M$ reporting 6 web servers, chances are
> they aren't using local directors at all. I should have completed my
> statement by saying that having local direcdtors in each of two or more
> data centers is the best approach. I guess it's possible that microsoft
> has three local directors in each of two data centers for redundancy is
> possible...
>From http://www.microsoft.com/backstage/:
"As an added bonus, we have a complete case study that describes how we've
put Cisco DistributedDirector routers to work on top of our already
world-class Single IP solution (Windows Load Balancing Service). Now
customers no longer need to pick mirror sites when they want to download
files, they just click on a link and, presto, DistributedDirector delivers
the files from the most appropriate data center. It's not magic, just pure
technological innovation."
-mike
-
[To unsubscribe, send mail to [EMAIL PROTECTED] with
"unsubscribe firewalls" in the body of the message.]
