Steven,
thanks for the feedback. comments see inline.
On Wed, Jul 29, 2009 at 02:51:53AM -0500, Wright, Steven wrote:
> I have been reading "draft-ietf-alto-reqs-01.txt" and had some questions
> about the requirements here, particularly from an IPv4/IPv6 transition
> perspective.
>
> 2.2. ALTO terminology
>
> o Host Group Descriptor: ...This can be, for example, a single IP
> address, an address prefix or address range that contains the
> host(s), or an autonomous system (AS) number.
>
> Q: must Host Group Descriptor: be able to include both IPv4 and IPv6
> for a dual stack host ?
(please see below)
> REQ. ARv01-8: The ALTO client protocol MUST support the host group
> descriptor types "IPv4 address prefix" and "IPv6 address prefix."
> They can be used to specify the IP address of one host, or an IP
> address range (in CIDR notation), which contains all hosts in
> question. It is also possible to specify a broader address range
> (i.e., a shorter prefix length) than the intended group of hosts
> actually uses, in order to conceal their exact identity.
>
>
> Q: Is it that "host group descriptor" is a set of "IPv4 address
> prefix" or "IPv6 address prefix." Or is it possible that a "host group
> descriptor" is a set containing both "IPv4 address prefix" and "IPv6
> address prefix." ?
The idea is that the protocol must support "atomic" host group
descriptors which are either a single (not a set of multiple!)
IPv4 prefix or a single IPv6 prefix.
Furthermore, one may define other hosts group descriptor types, e.g.,
macros, but there must be a "... facility, which can be used to translate
host group descriptors of that type to IPv4/IPv6 address prefixes ..."
[ARv01-11].
Here it says "prefixes", and the intention is to allow mixing of IPv4
and IPv6.
If we changed ARv01-11 to the following proposal, would that be more
explicit?
REQ. ARv01-11: For host group descriptor types other than "IPv4
address prefix" and "IPv6 address prefix", the host group descriptor
type identification MUST be supplemented by a reference to a
facility, which can be used to translate host group descriptors of
that type to a set of IPv4 and/or IPv6 address prefixes, e.g., by
means of a mapping table or an algorithm.
Regarding dual stack hosts one has currently two options:
- use the mandatory atomic descriptor types, and treat each dual stack
host from an ALTO perspective as two independent v4-only and v6-only
hosts, respectively, which "just happen to" have the same Host
Characteristics Attribute and the same evaluation results according to
all Rating Criteria.
- use a macro mechanism (see above) to bundle one v4 and one v6 address.
Using a general macro mechanism would have the drawback that there is
no explicit indication that the two addresses belong to the same host.
In both cases the question is, would there be any benefit for the
resource consumer to know, that a v4 and a v6 address, which are ranked
equal, are actually the same host?
> 4. Host group descriptors
>
> Q: What is the requirement text for section 4?
As stated in the first paragraph of section 4, this section is not
normative but tries to capture previous proposals and discussions.
The relevant requirements for host group descriptors are in section 3.1.2
> There are two options listed, the discussion Tuesday seemed to
> depreciate the AS option. Is it required to support both ? Is one
> optional ?
They are both optional. See ARv01-8 for required attribute types.
> Q:AS's have defined semantics. If it is not an AS are there requirements
> we should be stating about the creation/ deletion / definition of PIDs ?
> ( e.g. who can do it, when, etc.)
see ARv01-9 .. 13
> 5. Rating criteria
> Q: What is the requirement text for section 4?
As stated in the first paragraph of section 5, this section is not
normative but tries to capture previous proposals and discussions.
The relevant requirements for rating criteria are in section 3.1.3
> Q:If the intention is to support dual (IPv4/IPv6) stack hosts, would
> there be a need to be able to use these rating criteria to express a
> preference between them ?
Do we need this mechanism in the ALTO client protocol, or would that
be a function of the respective (P2P) application protocol?
thanks,
Sebastian
--
Sebastian Kiesel mailto:[email protected]
Network Research Division tel:+49-6221-4342-232 fax:+49-6221-4342-155
NEC Laboratories Europe Kurfuerstenanlage 36, 69115 Heidelberg, Germany
--
NEC Europe Limited Registered in England 2832014
Registered Office NEC House, 1 Victoria Road, London W3 6BL
_______________________________________________
alto mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/alto
