The term service has been massively overloaded in the marketing brochures of Service Oriented Architecture (SOA) solutions.
A simple description from the point of view of computing technologies could be: a service is a self-describing software entity that is designed and developed in isolation offering near frictionless interoperability to deliver some kind of business value or input. This covers a wide spectrum including; web services, application services, hardware services, virtualization services, cloud services and others.
In sense‘s view almost all elements of computing technology and infrastructure can be considered services if
they can be interfaced and interacted with,
they can be affiliated and managed by an external actor,
they deliver some form of measurable business input or value.
By its nature, sense is a self federating system that hosts sense nodes, services, external interactions, business rules and others; with each entity being represented with sense as a service.
A sense service extends the of idea of Service Level Agreement (SLA) to declare business drivers that influence and change the behavior of a runtime scenario automatically in order to maintain an adequate SLA, without the need for human intervention. An SLA lets a service declare a series of ‘states of health’ that are advised within the sense federation during runtime. The switch between the service level states is computed by the service itself depending on used defined criteria, such as the average response time of the service or how much revenue is acquired by the service: sense evaluates the health and can overload the service and intelligently redirect the call to the best performing service provider within the federation.
As a last level fall back, if all services within the federation are failing to the lowest health state, sense is able to intercept all the calls and propagate a mock response from the mock strategy configured for each service. This protects the calling system from a possible system crash and lets the overall system catch its breath and continue operation while some of the services themselves are still in a crisis or unavailable state. sense‘s mock strategy disrupts the domino effect seen in many applications when one component of the system fails.
No Comments »
No comments yet.
RSS feed for comments on this post. TrackBack URL
Leave a comment
Comments (0)