A Novel Reconfigurable-by-Design Highly Distributed Applications Development Paradigm over Programmable Infrastructure

Given the inability of Highly-Distributed-Application-Developers to foresee the changes as well as the heterogeneity on the underlying infrastructure, it is considerable crucial the design and development of novel software paradigms that facilitate application developers to take advantage of the emerging programmability of the underlying infrastructure and therefore develop Reconfigurable-by-Design applications. In parallel, it is crucial to design solutions that are scalable, support high performance, are resilient-to-failure and take into account the conditions of their runtime environment being able to adapt. Towards this direction, the ARCADIA project aims to design and validate a Novel Reconfigurable-By-Design Highly Distributed Applications (HDAs) Development Paradigm over Programmable Infrastructure. The proposed framework relies on the development of an extensible Context Model which will be used by developers to produce annotated source-code and generate HDAs as service chains of application tiers and network functions containing meaningful semantics. A Smart Controller responsible for on-boarding the HDAs will undertake the tasks of translating annotations to optimal infrastructural configuration. Such a controller will enforce an optimal configuration to the registered programmable resources and will proactively adjust the configuration plan based on the Infrastructural State and the Application State to meet objectives and apply policies. Driving a HDA through its entire lifetime proves highly beneficial for all stakeholders since the synergy of the introduced applications' reconfigurability and the underlying infrastructure's programmability, facilitates the development of new fine-grained strategies able to fulfil new and complex requirements.

[1]  Carlos Becker Westphall,et al.  Cloud resource management: A survey on forecasting and profiling models , 2015, J. Netw. Comput. Appl..

[2]  Romain Rouvoy,et al.  Towards multi-cloud configurations using feature models and ontologies , 2013, MultiCloud '13.

[3]  Laurence Duchien,et al.  Cardinality-based feature models with constraints: a pragmatic approach , 2013, SPLC '13.

[4]  Brice Morin,et al.  Towards Model-Driven Provisioning, Deployment, Monitoring, and Adaptation of Multi-cloud Systems , 2013, 2013 IEEE Sixth International Conference on Cloud Computing.

[5]  Philippe Roose,et al.  Context Aware Adaptable Applications - A global approach , 2009, ArXiv.

[6]  Brice Morin,et al.  Managing multi-cloud systems with CloudMF , 2013, NordiCloud '13.

[7]  Hui Song,et al.  CloudMF: Applying MDE to Tame the Complexity of Managing Multi-cloud Applications , 2014, 2014 IEEE/ACM 7th International Conference on Utility and Cloud Computing.

[8]  Keith G. Jeffery,et al.  Research information management: the CERIF approach , 2014, Int. J. Metadata Semant. Ontologies.

[9]  Romain Rouvoy,et al.  Leveraging feature models to configure virtual appliances , 2012, CloudCP '12.

[10]  Jörg Domaschka,et al.  Towards a Generic Language for Scalability Rules , 2014, ESOCC Workshops.

[11]  Zoltán Ádám Mann,et al.  Allocation of Virtual Machines in Cloud Data Centers—A Survey of Problem Models and Optimization Algorithms , 2015, ACM Comput. Surv..

[12]  Jörg Domaschka,et al.  SRL: A Scalability Rule Language for Multi-cloud Environments , 2014, 2014 IEEE 6th International Conference on Cloud Computing Technology and Science.