Some thoughts …

Just thought I’d christen this group with some tentative ideas (dare I call it a “seeding of our cloud-making efforts”? :-) as to directions we might be pursue in the development of cloud-related standards .

These ideas are highly tentative: I definitely welcome feedback, especially from those with more particular expertise in the relevant areas than myself (being none in particular :-)

First of all, I think we’ll need some type of “engagement” ontology (meaning a description of the ways in which cloud services may be engaged) and a set of polices (possibly in the form of a RIF rule set) detailing the procedure used to select a cloud service or set of services

We’ll also need to think about how we should utilise models in our cloud processing, and how we should represent our models. I think we’ll need to two types of models: a “computational” model (referring to the flow of control of execution within a single type of cloud processor) and a “physical” model (how the configuration of processors, proxies, gateways and sub-clouds combine to collectively implement the cloud’s task).

For such modelling we could use ideas from the XMI model representation (if nothing else, because it’s in XML) within the OMG’s MDA infrastructure, as well as from XProc

However these models are represented, they’ll most likely also need some kind of API representing the execution’s environment which, in the case of the computational model, would represent interfaces to data connections with neighbouring nodes (although I don’t yet know whether these interfaces should be configurable)

I also like the idea of being able to reference related models by URI (with a fragment identifier further locating contained model elements) , which is in keeping with other W3C standards.

The RDF mapping I mentioned in the group’s summary is intended to enable RDF processors to also perform cloud processing, bringing with it the benefit of an existing DOM API for making dynamic changes to models (although this could also be accomplished by other means)

But, as mentioned, these are all tentative ideas, and further input is welcome and encouraged :-)

2 Responses to Some thoughts …

  1. Le Sun says:

    Hi…good idea..currently, I try to identify a service description model for cloud service transactions. My initial thought was on the basis of USDL(unified service description languange)

    • Russell Potter says:

      Le Sun,

      (My apologies if I’ve already responded to
      your message, but I’m still getting used
      the WordPress UI and wasn’t sure whether
      I’d already done so)

      Can’t say I’m totally familiar with USDL
      but, from what I understand, it seems to
      be concerned with the specification of
      the products of an organisation underlying
      the servicing of a request.

      Although both our effort and theirs
      use MOF as each’s approach to implem-
      enting the relevant modelling, USDL
      is specialised for factors surroun-
      ding the specification of these pr-
      oducts, while ours is specialised
      for distributed computation

      Hope this description helps :-)
      Russell

Leave a Reply

Your email address will not be published. Required fields are marked *

*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Before you comment here, note that this forum is moderated and your IP address is sent to Akismet, the plugin we use to mitigate spam comments.