Trusted Services Network

| No Comments | No TrackBacks

I've been visiting several telecommunications companies over the last couple of years about the difficulty in getting to the convergence nirvana. Recently I've been discussing the development of the service network with colleagues of mine, including Tony Parker and Brian Whittaker. One of the exciting developments is the idea of opening up their networks for third party applications offered as services. It's the idea of providing your network as a marketplace for Software as a Service (SaaS).


One of the overriding themes Tony, Brian and I have been hearing lately is the idea that while the technology is more or less available for the rapid deployment of services on the network, consumers of those services will have difficulty trusting these services and the network. And service providers are wary of the impact that consumers or other service providers can have on their services. In order for the network provider to be successful the consumers of services must be confident in the quality of the services on the network. These consumers include the produces of new services based on composites (mashups) of other services.

The way to solve this problem, while maintaining the value of rapid service deployment, is to provide certification of the services. This initially sounds orthogonal to the idea of rapid service mashups but it can actually be achieved in an effective way. Technology exists today that can solve this problem to a great extent.
There are several elements needed:


  • Well formed service contracts
  • Certification Kits
  • Accessible Active Repository

Well formed service contracts are required because this is going to be one of the main selling points for a service. What the service provides and how it provides it, is crucial to building confidence with consumers. The design of new services or the refactoring of old contracts for deployment as new service contracts is a vital part of the process. Poorly defined services will not be consumed and will therefore lower the value of the service network.

Certification Kits provide the mechanism for rating the capabilities and therefore the usability of a service. A certified service, provided it meets the well formed service contract requirement, will be trusted and therefore consumed. Services are certified by meeting the requirements of their contract. Obviously this can have various depths.

Certification kits allow service providers not only to guarantee their own kits in terms of meeting the contract but also allow composite/mashup services to gaurantee quality against the services they consume in the mashup. Certification kits include full emulators of service consumer and service provider and include the service contract and the data required for testing. Kits may be black-box or white-box in terms of the type of logic and data required and available. Certification kits require good test data. Some of this can be generated, some can be pulled off of production systems (scrubbed for privacy reasons) and some may need to be hand input. Test data is actually the bulk of the work, not the code.

An accessible, active repository is required to provide a place for efficient storing and retrieval of service contracts, certification kits and certification status. This is not just a look-up registry. When third parties want to consume services, including for consumption within mashups, it is necessary for the process of downloading certification kits to be straightforward. Service network providers may require a fee but the confidence in the certification kits and ease of use will make this worth while. An active repository supports the service lifecycle and will schedule certification tests when new services are deployed into the test environment and can version services across test and production networks. At any point in time consumers and providers can look at the network and see the "health" of the services deployed and therefore the overall "health" of the service network.

This is a very high level look at this model. There are greater details within each of these areas. Feel free to contact me if you are interested in more details.

No TrackBacks

TrackBack URL: http://ipbabble.com/cgi-bin/mt/mt-tb.cgi/15

Leave a comment

Who is IPBabble

William Henry IP Babble is the personal blog of William Henry.

William has 20 years experience in software development and distributed computing and holds a M.Sc from Dublin City University. He is currently working in the office of CTO at Red Hat on the MRG product. This weblog is not funded by Red Hat.

Posts are intended to express independent points of view, but understand that there is probably a bias based on the influence of working with standards based middleware for over a decade. (See disclaimer below)

December 2008

Sun Mon Tue Wed Thu Fri Sat
  1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 30 31      

Disclaimer

The views expressed in this blog are solely the personal views of the author and DO NOT represent the views of his employer or any third party.

About this Entry

This page contains a single entry by William Henry published on May 10, 2007 3:44 AM.

IONA Advances SOA was the previous entry in this blog.

Sun SOA Dog Food? is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.