RE: New draft available

[Eric Dean <eric@crystalballinc.com>]

On Tue, 21 Nov 2000, Gary Tomlinson wrote:

> I don't see how ICP can be considered to be within the context of
> request-routing as defined within CDNP.  ICP is used between caches to
> locate content being delivered by the requestor of ICP, where "delivery" is
> the key aspect to be considered.  In CDNP request-routing involves getting a
> request to a surrogate for delivery, as opposed to the surrogate utilizing
> ICP to find content within the CDN.  

You are making an assumption that the only method to get a request is
via DNS. There are methods to completely bypass DNS when a client is
proxied into a cache. In this manner a client cache could talk directly to
a surrogate via ICP and then use DNS otherwise.
 
> Simply put, ICP happens within the
> cloud itself as a byproduct of a CLIENT REQUEST being serviced by SURROGATE.
> I see ICP as being associated with DISTRIBUTION within the CDNP context.
> 
> Gary
> 
> -----Original Message-----
> From: owner-cdn@ops.ietf.org [mailto:owner-cdn@ops.ietf.org]On Behalf Of
> Eric Dean
> Sent: Tuesday, November 21, 2000 9:06 AM
> To: Mark Day
> Cc: cdn@ops.ietf.org
> Subject: Re: New draft available
> 
> 
> 
> This memo doesn't address any implementations of using ICP RFC2186/RFC2187
> for CDN Request Mappings.  I have seen cases whereby an ISP either uses
> transparent caches, or forced browser proxy into a cache that then speaks
> ICP to nearby surrogates.
> 
> I have also seen implementations of CDN Request Mappings that use BGP
> multi-homing techniques whereby the DNS address for content is registered
> to an IP address that is shared by many surrogates in a diverse network.
> The network then routes the content requests to the nearest surrogate.
>    
> Both of these implementations are widely used and have considerable
> issues.
> 
> On Thu, 16 Nov 2000, Mark Day wrote:
> 
> > This one really is new, as opposed to a revision of an existing draft:
> > 
> > http://www.ietf.org/internet-drafts/draft-cain-cdnp-known-req-map-00.txt
> > or
> > http://www.content-peering.org/draft-cain-cdnp-known-req-map-00.html
> > 
> > Here's the abstract:
> > 
> >     This memo presents a number of known mechanisms used to direct
> >     client application requests to surrogate servers based on various
> >     policies. In this memo we group mechanisms commonly called request
> >     routing, content routing or content redirection under the term
> >     request mapping. There exist multiple request mapping mechanisms. At
> >     a high-level, these may be classified under: DNS Request Mapping,
> >     Transport-layer Mapping, and Application-layer Mapping.
> > 
> > Thanks to the multiple members of the design team who wrote this new
> draft:
> > Brad Cain, Fred Douglis, Mark Green, Markus Hoffmann, Raj Nair, Doug
> Potter,
> > and Oliver Spatscheck.
> > 
> > --Mark
> > 
> > 
> 
> Eric Dean
> President, Crystal Ball Inc.
> W 703-322-8000
> F 703-322-8010 
> M 703-597-6921 
> 
> 
> 

Eric Dean
President, Crystal Ball Inc.
W 703-322-8000
F 703-322-8010 
M 703-597-6921