Service Availability: Principles and Practice

Chapter 16

Overcoming Complexity: Formal Modeling Techniques at the Rescue

Maria Toeroe1 and Ferhat Khendek2

1Ericsson, Town of Mount Royal, Quebec, Canada

2Concordia University, Montreal, Quebec, Canada

16.1 Introduction

As we have seen in Chapter 6 the Availability Management Framework (AMF) [48] performs its task based on an information model, and in particular the configuration information contained in this model. This model defines the scope within which AMF manages the availability of the services. That is the guarantees AMF can provide with respect to service availability (SA) is built into the model.

The model describes each entity AMF should manage as objects of the classes we touched upon in Chapter 6. Among them the configuration object class describing an AMF component defines over 20 configuration attributes all of which need to be considered when constructing the configuration for AMF. It is easy to see how the task of designing a configuration, which is not only syntactically and semantically correct – that is compliant to AMF's expectations – but that also guarantees the availability and maybe other characteristics of the system becomes a designer's nightmare if it needs to be done manually.

The Platform Management service (PLM) [36] presented in Chapter 5 requires a similar model, but it poses somewhat different challenges. Here the main issue is to make sure that when PLM collects the platform information from the Hardware Platform Interface (HPI), it needs to ...