A policy-driven, model-based software and services deployment architecture for heterogeneous environments (Propuesta de una arquitectura de despliegue de software y servicios para entornos heterogéneos basado en modelos y dirigido por políticas)
- Autores: José Luis Ruiz Revuelta
- Directores de la Tesis: Juan Carlos Dueñas López (dir. tes.)
- Lectura: En la Universidad Politécnica de Madrid ( España ) en 2007
- Idioma: español
- Tribunal Calificador de la Tesis: Gonzalo León Serrano (presid.), David Fernández Cambronero (secret.), Carlos Delgado Kloos (voc.), José Juan Pazos Arias (voc.), Juan Ramón Velasco Pérez (voc.)
- Materias:
- Texto completo no disponible (Saber más ...)
- Resumen
Software deployment focuses on the group of activities associated with the transition of software from software production sites to software consumers' sites. Modern systems are created by combination of multiple software components at runtime. Furthermore, they are very frequently changing. Systems are dynamically updated, extended and removed. These changes may be related to many different causes, such as including new features, improving performance or correcting failures. In consequence; software deployment issues have an undeniable impact on ITC systems' operation and maintenance costs. Despite this fact, deployment challenges have traditionally been neglected by Software Engineering. At least, if we compare it with the attention drawn by other development-time activities such as design, coding, testing or development processes.
Unfortunately, there are many factors that complicate the task of automating software deployment. Deployment units exhibit complex dependency patterns that must be understood and solved. Dependencies span elements at different but interrelated levels; software, hardware and networking. All of them must be satisfied in order to carry out a proper deployment action.
This dissertation's contributes to the area of software deployment with a software deployment architecture that supports context adaptation and distributed deployment targets. A deployment system is context aware if it is capable of adapting its operation to the current status of the deployment target, i.e. its hardware, software and networking available resources. Deployment activities can be initiated from remote locations and deployment units can be fetched from many distributed units' repositories. I have intended to handle an additional aspect of distribution; distribution at the deployment target. This implies that the deployment target can be a networked set of nodes. Eventually, this implies that units' which collaborate to provide a certain service can be allocated
