posted Apr 10, 2014, 8:47 AM by Anthony Simons
updated Apr 10, 2014, 8:48 AM
Cloud Service Brokerage - Towards the Multi-Cloud Ecosystem (CSB 2014)
- Dr Iraklis Paraskakis, South-East European Research Centre, Thessaloniki, Greece
- Dr Anthony J H Simons, University of Sheffield, United Kingdom
- Dr Alessandro Rossini, SINTEF, Oslo, Norway
- Dr Jens Jensen, STFC, Rutherford Appleton Laboratory, Didcot, Oxfordshire, UK
This workshop looks to a future in which a multi-cloud ecosystem exists, within which many cloud providers and consumers interact to create, discover, negotiate and use software services. Supporting this ecosystem are cloud brokers, whose role is to bring together providers and consumers, by offering service portals with added value for all parties. A central feature of the broker’s role will be to assist with software service generation (from abstract models to platform-specific deployments), multi-cloud translation (model-driven adaptation and deployment of services) and assure quality control (governance; functional testing and monitoring), service continuity (failure prevention and recovery; service substitution) and market competition (arbitrage; service optimization; service customization).
The following (non exclusive) topic areas are of direct relevance and interest:
- Abstract service models – current service description languages are only up to the task of describing the service interface; what is needed are abstract models to describe the complete behaviour and performance of services, such that model-translation algorithms can generate equivalent services on different platforms.
- Model-driven generation – current software services are designed in vendor-specific ways that prevent them being ported onto different platforms; what is needed are sets of translation algorithms for converting abstract service models into platform-specific applications, or sets of equivalent applications deployed across multi-clouds.
- Service behaviour certification – current quality control is mostly achieved by in-house developer-based testing; what is needed is a means of determining whether alternative services are equivalent, certified by generating standard test sets from functional specifications of services, and grounding these for each of the service protocol technologies described above;
- Service performance monitoring – current service monitoring technology is limited to SLAs for response-times and availability of end-points; what is needed is a more sophisticated data fusion approach, such as complex event processing, with trend prediction, supporting service optimisation and substitution.
- Service optimization – current service platforms offer single-vendor services with failover substitution, or manual selection from several providers; what is needed is a means of offering multi-vendor services on a competitive basis, with automatic arbitrage between different providers, to support constrained optimization of cloud performance.
- Service governance – current services and platforms are developed following in-house software processes; what is needed are explicit standards and methods for governing the whole service lifecycle, ensuring common quality standards and interfaces, supporting convergent service development and service customisation.
Details of the submission process via EasyChair will be made available on the CSB 2014 Workshop Website. The main important dates to remember are the following:
- Paper submission: 30 May 2014
- Reviews completed: 30 June 2014
- Camera-ready copy: 15 July 2014
Publication of the workshop proceedings will be in the Springer CCIS series (Communications in Computer and Information Science). For further details about ESOCC 2014, please consult the ESOCC 2014 Conference Website