Automated Change Planning for Differential Update IT Systems with State Constraint

Abstract

As the life-cycle of IT systems is getting faster and shorter, the lead time requirements of the deployment for updates have become severe. Provisioning automation tools make it easy to update IT systems, but the update of a large-scale IT systems such as a large enterprise system is still time-consuming. Existing provisioning tools cannot automatically generate update procedures because they cannot take care of the constraints related to the states of the current system. For example, the web service has a constraint that requires both of the modification of the configuration files and the restart of the service for configuration update when the service is in the running state. On the other hand, it requires only the modification of the configuration files when the service is not in the running state. In this paper, we propose a novel declarative component-based model scheme that automatically generates the update procedure from the difference between the current and desired state for IT systems with the aforementioned state constraints. We present the efficiency and effectiveness of our scheme through a case study using the implementation of a prototype.