Thomas Narten wrote:
> "Tony Hain" <[EMAIL PROTECTED]> writes:
> 
> > The discussion that should have happened first is 'what 
> alternatives 
> > do we have to deal with the requirements that network managers are 
> > using SL to deal with?' Without a clear replacement, and 
> with comments 
> > that some real problems are 'uninteresting', the network 
> manager will 
> > insist on keeping the current tool.
> 
> You have said a number of times now that network managers 
> want and need SLs.
> 
> Question: Do you believe that if we keep SLs, this implies
> that we will indeed need to complete the work on multi-site 
> nodes and that multi-site nodes will become necessary in 
> practice to deal with what I expect the common practice of 
> nodes being in multiple sites simultaneously (as I described 
> in an earlier note)?
> 

I also believe that multi-sited nodes will be commonplace. With a viable
PI space, I would expect most all of the multi-sited nodes to use that
because the places that really want the ambiguity feature are not likely
to want external access from nodes visiting another site. With PI the
node needs to keep straight which domains should be retrieved from which
DNS servers, and then which interface on the multi-sited node is
attached to the returned prefix. The only difference I see between the
work required for this vs. SL is that the multi-sited node can use the
prefix assigned to an interface as a key, rather than having to create a
local one to track where the name resolution came from. Either way it
has to track which domains are available over which logical interface. 


> (One of the things that happened at the SF meeting was that a 
> number of people expressed skepticism at the viability of a 
> limited usage "compromise" where we could use SLs in a 
> limited fashion.)

I have never believed that any of compromise approaches made any sense.
Those are about mandates counter to the network managers goals. The
majority of network managers will only choose SL because we don't offer
a viable PI option. If we fix that and provide a way to simply configure
a massive number of local use devices, the demand for the ambiguous SL
will diminish. 

Tony



--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page:                      http://playground.sun.com/ipng
FTP archive:                      ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------

Reply via email to