PWG Mail Archive: PWG> FW: [Srvloc-discuss] FYI: eXtensible

PWG> FW: [Srvloc-discuss] FYI: eXtensible Service Discovery Framework (XSDF)

From: McDonald, Ira (imcdonald@sharplabs.com)
Date: Thu Mar 25 2004 - 19:15:14 EST

  • Next message: McDonald, Ira: "PWG> Thu 1 April - Process meeting today??"

    Hi,

    VERY interesting new work on a next generation Service Discovery
    protocol and framework, heavily based on SLPv2 (RFC 2608).

    Much more sophisticated (load balancing, GUID service IDs rather
    than mere URI for unique services, etc.) than any service
    discovery protocol I've ever seen. If this advances in IETF,
    it will be a force to be reckoned with in the future.

    Recommended reading.

    Cheers,
    - Ira

    Ira McDonald (Musician / Software Architect)
    Blue Roof Music / High North Inc
    PO Box 221 Grand Marais, MI 49839
    phone: +1-906-494-2434
    email: imcdonald@sharplabs.com

    -----Original Message-----
    From: Manuel Urueña [mailto:muruenya@it.uc3m.es]
    Sent: Thursday, March 25, 2004 11:35 AM
    To: srvloc-discuss@lists.sourceforge.net
    Subject: [Srvloc-discuss] FYI: eXtensible Service Discovery Framework
    (XSDF)

    Hi all,

    I've been working on a evolution of the SLP protocol called XSDF and I
    am very interested in the feedback from the SLP community, as most of
    the ideas, including the architecture itself, are owed from it.

    I've started to think about an enhanced SLP after reading Guttman
    serviceid draft, which identifies the problems derived from identifying
    a Service by its URL. Also related to this, I started believing that SLP
    service model was too simple to represent certain services, for example
    when multiple network/transport/application protocols are available.

    That led to define a new Service model and a compact binary encoding to
    serialize XSDF messages. That new service model defines a set of
    standard attributes, for example to enable load balancing when multiple
    servers are found (from Rserpool WG requirements).

    As I was designing a new protocol to overcome these issues, I also
    wanted to integrate some optional capabilities defined for SLP, which
    are very interesting. For example Zhao et al. remote DA discovery, mSLP;
    and Kempf et al. Notification & Subscription.

    However, the protocol became so big that I did prefer to split it in
    several ones. For example XSLP is employed to query SAs/DAs for
    services, SAs employ XSRP to register their services in DAs (if any),
    and XSSP and XSTP are employed to keep DAs from the same Realm in sync.

    At last, I have been able to bring all together and I've just submitted
    several drafts that define XSDF:

    http://www.ietf.org/internet-drafts/draft-uruena-xsdf-overview-00.txt
    http://www.ietf.org/internet-drafts/draft-uruena-xsdf-common-00.txt
    http://www.ietf.org/internet-drafts/draft-uruena-xslp-00.txt
    http://www.ietf.org/internet-drafts/draft-uruena-xsrp-00.txt
    http://www.ietf.org/internet-drafts/draft-uruena-xssp-00.txt
    http://www.ietf.org/internet-drafts/draft-uruena-xstp-00.txt
    http://www.ietf.org/internet-drafts/draft-uruena-xbe32-00.txt

    The xsdf-overview draft resumes the whole thing in just 20 pages.
    If you are interested in a particular protocol, they are specified in
    separate drafts, while xsdf-common contains the elements and procedures
    shared by all of them. At last, the xbe32 one defines the binary
    encoding for XSDF messages.

    Thank you very much for your time !

    Best regards,
    --Manuel

    -- 
    Manuel Uruen~a - Universidad Carlos III de Madrid
    GPG FP: 68A1 164B EE28 52C9 87CB  EBF9 616E 52B5 451A B6B2
    



    This archive was generated by hypermail 2b29 : Thu Mar 25 2004 - 19:15:36 EST