Forwarding: The decision, whether to employ it, is *complex*, kids.
dbrash at NOSPAM.gmail.com
Sat Dec 11 04:43:10 UTC 2004
I prefer to configure forwarders because it makes sense to me from a
performance perspective. Also, because I live in China, with forwarders I
can forward my DNS to a western DNS Server getting around many issues.
That said: point taken. This may not be the best solution for all cases,
and I should probably better qualifiy my opinions.
However, I would appreciate in the future that if you take issue with my
posts, please either email me directly or post back with a clear description
of your differences. I am neither stupid nor lazy, and I am simply here
trying to be helpful, as I know that you are. I also don't believe that I
am infallible, so I am very open to being corrected and learning more from
the community. In fact I welcome contrary discussion.
I look forward to reading more of your insightful and intelligent posts in
MCSE, MCDBA, MCSA
dbrash at NOSPAM.gmail.com
"Jonathan de Boyne Pollard" <J.deBoynePollard at Tesco.NET> wrote in message
news:c1.01.2thvzy$5AJ at J.de.Boyne.Pollard.localhost...
> If you're not using a forwarder (well, then you should be.
> Dana Brash
> MCSE, MCDBA, MCSA
> In the discussion forum for ISC's BIND, DevilsPGD has just written:
> Why are you using forwarding? This is silly, dangerous and pointless.
> I agree 100% with the above -- I'm curious, can anybody tell me why the
whole forwarding fetish seems to be so popular, especially in printed books
and documentation and MCSEs?
> Most crackerjack box MCSEs are not only stupid, but lazy -- Setting up
forwarding takes a few more seconds then not setting up forwarding, so
what's the benefit?
> I was going to point out that, thanks to a lot of effort in several
discussions during 2003 and 2004, the received wisdom in the Microsoft
discussion fora that one should, unequivocally, use forwarding has been
wiped out, and has been largely replaced by the far better wisdom that
choosing whether to employ forwarding is a complex decision that involves
quite a number of criteria, and two different situations aren't necessarily
going to yield the same answer.
> But it appears that I would have been wrong: It hasn't been wiped out.
You even provided DevilsPGD with more ammunition for his comments on
"crackerjack box MCSEs".
> I can still point out that the current received wisdom in the ISC
discussion fora that he is parroting - that one should, unequivocally, never
use forwarding - is just as absolute and thus just as wrong as the erstwhile
Microsoft discussion fora received wisdom was, and that the ISC discussion
forum is suffering from the disease of entrenched bad received wisdom on
this subject more than the Microsoft ones are at the moment.
> I can also point out that what he said about documentation wasn't quite
true. Microsoft's own KnowledgeBase article on the subject, whilst being
very misleadingly phrased, does not, if read carefully, say that one is
required to configure forwarding. It says "If this server needs to resolve
names from its Internet service provider (ISP), you should configure a
forwarder.". The first word, "if", is very important, but much overlooked.
> The reason that it is popular to unequivocally recommend forwarding in
other documentation is that in the large majority of cases that other
documentation is written and published by ISPs, and as I have pointed out
before, it is in ISPs' own self-interests to encourage their customers to
use ISP-provided proxy services, be they (as here) proxy DNS services, proxy
HTTP services, or others. It cuts down on the bandwidth costs that the ISPs
have to pay. It isn't necessarily better for the customers themselves.
More information about the bind-users