FireworksWiki:Current events
From FireworksWiki
FIRE in FIA
FIA (Future Internet Assembly) is one-year old and FI communities are after the effort finding each other and more closely working cross-domains. FIRE (Future Internet Research and Experimentation) has disseminated its offering both existing and becoming, but a lot is to be done to make the FIRE services better available, also to find out what other domains really will need.
We invite to this Wiki the FIRE community and all the FIA to work on further with FIRE output with relation to needs for experimental research hosted in FIRE.
Initial thoughts on the FIA top-down a-priori "conceptual" architecture:
- The first questions to ask are probably: do we need such "top-down" and "a-priori" architecture(*) ? what is the value of such "top-down" architecture ? what such "a-priori" architecture brings into the FI debate ? Indeed, past tentatives to dictate a top-down a-priori architecture compared to an approach consisting in experimenting i) different architectural paradigms and ii) possible design of its components has proven to be far more succesful (remember that broadband ISDN is a clear example of the failure of such an apprioach) !
- Define a monolothic architecture (even if limited to a pseudo-functional composition with loose definition of its interactions) but strong coupling between its components poses the fundamental questions of i) evolvability and so durability of this architecture, ii) disabling increasing diversity in the communities that participate to its evolution and operation (Internet "by people") and iii) removes the self-arbitration and empowerment role of people (Internet "for people"). In particular, when one of the main challenges of the Internet is to address heterogeneity of people and community needs.
- Not clear if users and testbeds are players or components of this layout but it is nevertheless clear that restricting their scope as peripherical is rather limitative (experimental facilities are part of the research space itself). Note here that the FI scenarios seem to underline development at the periphery whereas slides 13-14 are focused on the network (in a large sense).
- Putting in place a multi-level set of facilities to allow scientific and research community to develop and evaluate by experimentation the architectural paradigms and components of the (future) Internet would be much more valuable compared to an approach that dictates "a-priori" what the architecture shall be (even "conceptually") and mandates subsequently that research should be limited to the context of that architecture.
- The purpose of FIA should be to identify architectural challenges and domain of scientific & technical investigation for which cooperation could be initiated so as to progressively characterize the architecture. It is to be expected that certain common principles (based on new paradigms investigated in the FIRE context) will emerge over time but this long-lasting task would be defeated by the current approach promoted by the FIA stakeholders.
(*) What's an architecture ? an architecture can be defined as the set of design principles from which (functional, informational, etc.) models are derived to characterize/specify its components (procedures, data structures, etc.) and their interactions (messages, calls, etc.).
Other input from FIRE projects for PPP discussion Media:N4C_PPP_FIA.ppt
Suggestion: initiate a dedicated activity on metrology/measurement system and performance monitoring for large scale distributed experimental facilities. For this purpose there is a need to come with a set of well-defined performance metrics as well as a measurement and monitoring method. Question: is there a dedicate place where this topic can be discussed/debated to bootstrap this activity ?
