Scaling problems with Windows Clients name and address lookup

[Christopher R. Hertel]

> On Wed, 7 Oct 1998, Christopher R. Hertel wrote:
> 
> > Richard,
> > 
> > I don't have an answer regarding registry entries.  I'll mention that this
> > is just the tip of the iceberg regarding scalability.  NetBIOS to DNS name
> > mapping simply doesn't work outside of a LAN.  Browsing gets very silly
> > beyond a few interconnected LANs. 
> 
> if microsoft had used the NetBIOS scope as it was intended (to mirror the
> DNS zone) then we wouldn't _be_ in this mess.  NetBIOS names _do_ scale,
> *if* you use the (variable length, limited to 64 chars) NetBIOS scope.
> 
> it's just that no-one knows what it is; ms browsers don't bother with it,
> therefore browsing doesn't scale.

Um... not exactly (at least, no the way I remember your explanation, Luke). 

The scope idea arrived before browsing and--as you say--was never
implemented properly.  If I understand the RFCs correctly, scope was meant
to segregate names.  The DNS provides a hierarchy which can be traversed. 
Scope simply gives a name to a NetBIOS name "subspace". 

Machine zoobob.mouthwash would not have a name conflict with
zoobob.stethe, because the NetBIOS name service would use the scope to
keep the names separate.  On the other hand, machines in stethe scope
would not be able to resolve names for machines in the mouthwash scope. 

This is a way of avoiding scaling problem by maintaining smaller 
"virtual" name spaces.

How close am I on that?

Chris -)-----

-- 
Christopher R. Hertel -)-----                   University of Minnesota
crh at nts.umn.edu              Networking and Telecommunications Services