Work Package 1 is concerned with project management including data management, conflict management, and coordination among partners.
- Ensure efficient communication within the Consortium, the EC and cooperation with other projects.
- Ensure strategic and everyday management of the project.
- Ensure efficient scientific and technical coordination among the project’s participants.
- Manage conflicts by application of the foreseen procedures.
- Manage the technological development and project administration of related risks; ensure the quality of the final results and deliverables (reports and prototypes).
Work Package 2 deals with dissemination and exploitation. It monitors the goals and progress with respect to these aspects.
- Develop and implement a strategy for the commercial and academic exploitation of the project’s results by identifying markets, users and areas of innovation where the project’s outputs will provide a competitive advantage.
- Design and develop the project dissemination plan and communication handbook.
- Design, develop and maintain the content, channels and communities underlying the project’s dissemination plan including the RECAP website, social networking presence and all collateral, print, digital or otherwise, associated with the project.
- Openly and proactively disseminate and promote RECAP progress and results to the relevant scientific communities and a wider group of stakeholders including relevant users, cloud service providers, policy and decision makers and lay audiences, including citizen scientists and students.
- Raise awareness, build consensus and create visible/measurable impact in terms of the ability of all stakeholders, including those at the political and decision making level, to exploit best practice related to the implementation of the new computing architectures, patterns, programming models and services being addressed in the project.
- Provide a “one stop shop” to access all the information and services relevant to RECAP activities.
- To align and liaise the activities of RECAP with standardisation bodies, European clusters and complementary initiatives at European and International levels.
Work Package 3 is concerned with the use cases including requirements analysis, use case adaptation, and validation of the platform against requirements and use case features.
- Collect requirements imposed by the various use cases.
- Ensure validation of the platform by the use cases.
- Validation of simulation models for real system experimentation.
- Validation of application and workload models.
Work Package 4 is focused on the definition and integration of the architecture of the RECAP system.
- Definition of the architecture of the RECAP.
- Definition of working rules and an integration methodology.
- Realisation of open-source strategy and definition of quality standards.
- Provisioning of integration infrastructure and build infrastructure.
- Integration of software and other artefacts from the technical work packages and integrate them.
- Operation and maintenance of an infrastructure to serve as a testbed.
- Development of techniques for runtime emulation of systems.
- Release of stable intermediate versions of the integrated prototypes.
- Realisation of demonstrators for the use cases.
Work Package 5 is designed to provide a foundation for workload modelling (WP6), simulation (WP7) and application modelling (WP8), and its primary objective is to deliver the tools required for managing and refining the data needed for this work.
- The definition of an infrastructure for acquisition and (pre-)processing of application and infrastructure monitoring data.
- The investigation and development of novel algorithms for filtering, segmentation, and cross-correlation of application and infrastructure load data.
- The provisioning of tools for annotating, archiving, managing, and visualizing data.
- Collection of data traces on real infrastructure (as made available by UULM, INTEL, BT) and publication of these traces as scientific source.
This work package provides models for decomposition, classification, and prediction of workloads; as well as models describing how load propagates in applications (i.e. between nodes in distributed systems, both in terms of paths and in terms of load vs performance trade-offs) and how applications as scaled to workloads (elasticity) and remediated when they deviate from desired application behaviour. The work package has the following main objectives:
- The statistical modelling of workload patterns and compositions, and based on these classify application and application component types.
- The investigation of the formulation of models for characterization and prediction of load propagation in data centre applications (distributed systems).
- The formulation of application elasticity and remediation models, that capture target application and component behaviour and define steps to proactively scale and remediate systems.
Work Package 7 builds on the results from WPs4-6 and addresses modelling and simulation of infrastructure resources and their administration.
- Simulate large scale data centres distributed geographically and fog computing paradigm.
- Provide mechanisms for feeding the simulation executions with footprints of the current status of data centres prior to executing the simulation.
- Use the RECAP workload models and application models created in order to validate and evaluate them, allowing the usage of simulation to predict and study the behaviour of application components, elasticity and remediation.
- Investigate the creation of distributed simulation with light processes providing simulation as a technique for optimizing and taking decisions of application placement and orchestration of resources in short time frames.
Work Package 8 is concerned with the integration of the technical work packages and the demonstration of results. This includes definition of the system architecture, provisioning of an integration strategy, supervision of test-cases for test-driven development, and release management of the project. In particular, this work package will provide releases for the use case partners that will be used for platform validation.
- The investigation of application deployment optimisation methods that build on probabilistic estimates of load variations, capacity requirements, and resource availability and trust models.
- The development of optimisation methods that exploit synergistic effects in simultaneous application-infrastructure placement and configuration optimisation.
- The development of tailored infrastructure models that suite the requirements of placement and optimisation.