sense is continually balancing the load on the system system by dynamically allocating all the available ‘healthy’ resources to satisfy the incoming requests. However, there are situations where resources are just not available and/or have been withdrawn from service (planned or unplanned), yet the system as a whole must continue to be reliable.

Every resource inside sense including services, feelings and business flows can associate a mock strategy in the case of the worst SLA situation i.e. a defined response even if the service, feeling or business flow being called is completely down

A mock strategy is used in a situation where sense does not invoke the service provider, because it is not available, and is managing the response itself by carrying out something specified in the SLA of the service itself. You can think of a mock strategy as the response in the ‘worst case SLA’ situation.

Mock strategies can vary in terms of complexity from a simple fixed response message that says that the single feature is unavailable, to a complex algorithm that takes the last ‘good’ result in the user cache and keeps track of the request call to be performed in the future when the system returns to a normal status.

Implementing a mock strategy permits sense to provide a legitimate, if perhaps empty, response so that the requester is not returned an out-of-band error, which could itself instigate a larger system, application or process crash – the classic ‘domino effect’. A well defined mock strategy allows for continuous execution by the calling applications or services under the most extreme circumstances.

go to faq