Modelling changes and data transfers for architecture-based runtime evolution of distributed applications

Abstract

Architecture-based approaches for runtime evolution enable software systems to dynamically move between consistent architectural variants. Successful runtime evolution must enable the new, replacement variant to be initialized with the data of the replaced one. In distributed systems, however, the initialization is complex and may be time-consuming due to data transfers across sites. Identifying systems' components subject to change is then critical for planning evolution and reducing replacement actions, avoid unnecessary data transfers, and then, reduce downtime of system services. Addressing this issue, this paper presents an approach to runtime evolution of distributed applications. We present how a development process allows to 1) specify architectural variants of an application and 2) identify components subject to change and operations for transferring data managed by these components. Moreover, the design information is used at runtime to automatically plan evolution. © 2010 Springer-Verlag Berlin Heidelberg.