Upgrade campaign simulation and evaluation for highly available systems

High Availability (HA) is an important nonfunctional requirement of carrier grade services. These services must not experience more than five minutes of downtime per year including outage due to planned upgrades. The Service Availability Forum (SAF) developed a set of standard services for an HA enabling middleware. Among these services, the Availability Management Framework (AMF) is responsible for managing application/service availability, while the Software Management Framework (SMF) is in charge of orchestrating upgrade campaigns that perform changes on these applications. Upgrades can be performed in different ways and may have different characteristics, such as induced service outage or upgrade duration. We propose an approach which is based on the Discrete Event System Specification (DEVS) formalism for upgrade campaign simulation, evaluation and selection using the concepts of best case and worst case scenarios.