Load Balancing depending on a weight

[Barry Margolin]

In article <dfsuam$1u7h$1 at sf1.isc.org>,
 Kevin Darcy <kcd at daimlerchrysler.com> wrote:

> Brad Knowles wrote:
> 
> > Moreover, most caching resolvers *don't* do round-robin, ...
> 
> Depends on your definition of "caching resolver", and your definition of 
> "round-robin". BIND is, after all, capable of functioning as a caching 
> resolver, and when doing so, gives out multiple-A-record answers in 
> different sequences, although most versions in use today just randomize 
> the sequence, or the starting point of the sequence, which most folks 
> would call "round-robin" (as opposed to fixed) even though technically 
> it isn't "round-robin"...

I'm pretty sure he was talking about client-side caches.  That's why he 
said "caching resolvers", not "caching servers".  It doesn't matter what 
the caching server does if the stub resolver performs its own caching.

And caching resolvers are not unusual -- I believe Windows has one 
built-in (I don't know much about it, but there's a command like 
"ipconfig /flushdns" that clears it), and so do many flavors of Unix 
(e.g. Solaris's nscd).

-- 
Barry Margolin, barmar at alum.mit.edu
Arlington, MA
*** PLEASE post questions in newsgroups, not directly to me ***