Hi, Enrico:
Comments inline below.

Regards!
-Qin

-----Original Message-----
From: alto [mailto:[email protected]] On Behalf Of Enrico Marocco
Sent: Thursday, February 20, 2014 12:04 AM
To: [email protected]
Subject: [alto] Text proposal for charter update

Hi all,

the text below tries to reflect the many discussions about rechartering the 
ALTO WG. We believe that it includes the most relevant work items people have 
shown interest for, yet it is something that can be completed in a resonable 
timeframe (some may find it ambitious, esp.
according to the WG history, but the level of energy the WG has shown recently 
in the finishing of our main deliverables provides good reason for being 
optimistic).

Any kind of comment (including nits, +/-1...) are at this point extremely 
important.

Enrico & Vijay


Application-Layer Traffic Optimization Working Group Charter

The ALTO working group was established in 2008 to devise a request/ response 
protocol for allowing a host to benefit from a server that is more cognizant of 
the network infrastructure than the host would be.
The working group has developed an HTTP-based protocol (RFC-to-be) to allow 
hosts to benefit from the network infrastructure by having access to a pair of 
maps: a topology map and a cost map.

The origins of the ALTO protocol lay in peer-to-peer (P2P) applications, where 
the host was a peer in a P2P network and desired to be rendezvoused with other 
peers for file sharing, real-time communications, etc.  It is a testament to 
the flexibility of the ALTO protocol that it is now being considered as a 
solution for problems outside the P2P domain, such as in datacenter networks 
and in content distribution networks (CDN) where exposing abstract topologies 
helps applications.

To support the emerging new uses of ALTO, certain extensions are being sought.  
These extensions can be classified as follows:

 o (Standards Track) Protocol extensions for reducing the volume of
   on-the-wire data exchange required to align the ALTO server and
   clients.  Extensions under consideration are mechanisms for
   delivering server-initiated notifications and partial updates of
   maps.  Efforts developed in other working groups such as Websockets
   and JSON-path will be considered, as well as bespoke mechanisms
   specific to the ALTO protocol.

 o (Standards Track) Extensions to the base ALTO server discovery
   mechanism (RFC-to-be) for deployment in heterogeneous network
   environments.  Mechanisms under consideration are extensions for
   third-party and anycast-based server discovery.

 o (Standards Track) Protocol extensions to convey a richer set of
   attributes to allow applications to determine not only "where" to
   connect but also "when" to connect.  Such additional information
   will be related both to endpoints (e.g. conveying server load and
   cache geo-location information for CDN use cases) and to
   endpoint-to-endpoint costs (e.g. bandwidth calendaring to represent
   time-averaged cost values in datacenter networks).

[Qin]:I think conveying server load and geo-location information is used to
Allow application to choose better connection, so it would be good to 
have the following change:
OLD TEXT:
"
o (Standards Track) Protocol extensions to convey a richer set of
   attributes to allow applications to determine not only "where" to
   connect but also "when" to connect.  
"
NEW TEXT:
"
  o (Standards Track) Protocol extensions to convey a richer set of
    attributes to allow applications to choose better connection and 
    determine not only "where" to connect but also "when" to connect.  
"


   The working group will specify such extension in coordination with
   other working groups that are also working on the related use cases
   (e.g. cdni, i2rs, lmap).

[Qin]: I think this paragraph is also applied to the 1st bullet.

 o (Informational) A survey of techniques to formalize the structure
   of a network graph (that can derived from a set of related ALTO
   network and cost maps) in a format that would facilitate advanced
   graph computation.  Such survey will cover both models used in
   popular open-source software (e.g. NetworkX, Blueprints) and models
   being considered in other working groups (e.g. netmod, i2rs).

[Qin]: Can the structure of a network graph be used to discover
SF node in the service function chain?

 o (Informational) An document on deployment-related issues and
   lessons learned from early implementation experiences.

[Qin]: is this existing work item we have not completed? Or you intend to
Expand deployment document to support more CDN network and data center network 
use cases?

When the WG considers standardizing information that the ALTO server could 
provide, the following criteria are important to ensure real
feasibility:

 - Can the ALTO service realistically discover that information?

 - Is the distribution of that information allowed by the operators of
   that service?

 - Can a client get that information without excessive privacy
   concerns?  Extensions defining new endpoint properties should focus
   on exposing attributes of endpoints that are related to the goals
   of ALTO -- optimization of application-layer traffic -- as opposed
   to more general properties of endpoints.  Privacy aspects of new
   endpoint properties will in any case be evaluated to the guidelines
   provided in the IANA considerations and Security Considerations of
   the ALTO protocol specification (RFC-to-be, sections 14.3 and 15.4
   at IESG review time).

 - Is it information that a client cannot find easily some other way?

After these criteria are met, the importance of the data will be considered for 
prioritizing standardization work, for example the number of operators and 
clients that are likely to be able to provide or use that particular data.  In 
any case, this WG will not propose standards on how congestion is signaled, 
remediated, or avoided, and will not deal with information representing 
instantaneous network state.

Issues related to the specific content exchanged in systems that make use of 
ALTO are also excluded from the WG's scope, as is the issue dealing with 
enforcing the legality of the content.

Milestones

Jul 2014 Working Group Last Call for third-party server discovery
         document

Jul 2014 Working Group Last Call for anycast-based server discovery
         document

Jul 2014 Working Group Last Call for partial updates document

Sep 2014 Submit third-party server discovery document

Sep 2014 Submit anycast-based server discovery document

Sep 2014 Submit partial updates document

Sep 2014 Working Group Last Call for deployment considerations
         document

Sep 2014 Working Group Last Call for network graph format survey
         document

Nov 2014 Submit deployment considerations document

Nov 2014 Submit network graph format survey document

Nov 2014 Working Group Last Call for server-initiated notifications
         document

Jan 2015 Submit server-initiated notifications document

Jan 2015 Working Group Last Call for endpoint property extension
         document

Jan 2015 Working Group Last Call for cost property extension document

Mar 2015 Submit endpoint property extension document

Mar 2015 Submit cost property extension document

[Qin]: I think the last two milestones timeframe should be corrected as "Sep 
2014" or 
Endpoint property extension document may be changed into "July 2014". I am not 
sure
We should wait for quite a long time to get these work done.

_______________________________________________
alto mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/alto

Reply via email to