Re: Known CDN Request Mapping Mechanisms

["Abbie Barbir" <abbieb@nortelnetworks.com>]

I think adding the section will be a good idea. It is important to have an up
to date running draft on various mapping methods.
Furthermore, I do second your concerns regarding the reliance on DNS for
mapping


----

abbie


John Martin wrote:

> Firstly, apologies for my recent absence.
>
> I've just quickly scanned "Known CDN Request Mapping Mechanisms" -
> excellent document. I have two comment:
>
> 1. I cannot find reference to the "GlobalIP" -type request mapping
> mechanism. I know the draft is already submitted - and I am trying to
> speed-read email before today's call to catch up - but I'd be happy to
> draft a section to add to this, if appropriate.
>
> 2. Section 3, "Transport Layer Mapping" appears to imply (to me at least)
> that this redirection is usually performed at DNS resolution when it can
> also be achieved in an ongoing way to reach a "steady state", e.g. client
> gets redirected to various random surrogates in the first instance and
> these collate measurements between SYN and SYN-ACK in order to determine
> the closest.
>
> I realise the specifics of the latter are noted as being out-of-scope but I
> am (still!) nervous about our reliance / assumption of DNS for mapping.
>
> Again, apologies if this has already been thrashed to death.
>
> John
> ---------------------------------------------------------------
> Network Appliance           Direct / Voicemail: +31 23 567 9615
> Kruisweg 799                               Fax: +31 23 567 9699
> NL-2132 NG Hoofddorp               Main Office: +31 23 567 9600
> ---------------------------------------------------------------