Community Research and Development Information Service - CORDIS


SWITCH Report Summary

Project ID: 643963
Funded under: H2020-EU.

Periodic Reporting for period 1 - SWITCH (Software Workbench for Interactive, Time Critical and Highly self-adaptive cloud applications)

Reporting period: 2015-02-01 to 2016-01-31

Summary of the context and overall objectives of the project

Many industrial applications have highly time-critical requirements for their performance in order to maintain their business value, addressing for instance issues of Quality of Service (QoS) (e.g. tsunami emergency response time) or quality of experience (QoE) (e.g. delivery of ultra-high definition television, or collaborative business interactions). These applications are often called time critical applications.

The development of such applications is usually difficult and costly, because of the high requirements for the runtime environment, and in particular the sophisticated optimizations needed during the development and integration of system components. In the meantime, a Cloud environment provides virtualised, elastic, controllable and high-quality on-demand services for supporting a range of different applications. However, the engineering methodology and software tools for developing, deploying and executing classical time critical applications have not fully accounted for the programmability and controllability provided by the Clouds—as a consequence, time critical applications cannot yet exploit the full benefits that Cloud technologies could potentially provide. There is therefore an urgent need at this time to develop new software methods and tools that account for these factors.

The overall objective of the SWITCH project is to address the entire life-cycle of time-critical, self-adaptive Cloud applications by developing new middleware and front-end tools to enable users to specify their time-critical requirements for an application interactively. Users describe their applications using a direct manipulation user interface, deploy their applications and can choose to adapt the infrastructure to changing requirements either automatically (using the specified requirements) or by human intervention if desired.

More specifically, the SWITCH project will aim to achieve the following concrete objectives:

Objective 1. Improve the development and execution model of time critical applications in the programmable, elastic and virtualised environments that Clouds provide. Using an application and infrastructure co-programming and control model, the application logic will be programmed along with considerations of the system QoS/QoE together with the programmability and controllability of the Cloud environment; both the application and the virtual runtime environment for executing the application are factored in and optimised during the design phase. A virtual runtime environment (a Cloud-based runtime environment for executing the application) can be customised for critical application requirements, and can be provisioned in the Cloud with a Service Level Agreement (SLA) negotiated for time critical application requirements. The application can then autonomously adapt its own behaviour and the virtual runtime environment when performance drops at runtime. This model will be the basis for both developing software tools and developing a formal verification mechanism for evaluating application quality constraints.

Objective 2. Increase the productivity of application development for time critical applications: to design and implement a software workbench (SIDE – SWITCH Interactive Development Environment) that provides the front end by which users interact throughout the software life-cycle. SIDE will provide intuitive interfaces for developers to define the application logic and an abstract virtual runtime environment with aid for a) specifying system QoS constraints, and b) validating and verifying QoS constraints for the development. SIDE should also provide data management solutions for applications defined within.

Objective 3. Improve the deployment efficiency for time critical applications in Clouds: to design and implement an infrastructure-planning tool (DRIP – Distributed Real-time Infrastructure Planner) that provides middleware for preparing and planning the deployment of an application. DRIP exploit

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

The project officially started Feb 2015. In the first six months, project partners first analyzed detailed application requirements and reviewed the state of the art of different technologies. The review results together with an analysis of existing technology gaps have been summarized as the first technical deliverable of the project. In the meantime, plans for managing project quality, data management and other support actions have been refined since the beginning of the project and reported across three deliverables.

Based on the requirement analysis and technology review, the technical partners (UvA/CU/UL) then focused on a detailed design of the system architecture, and on evaluating key technologies for the core functional components of the system including technologies for modelling and integrating semantic information required for developing/controlling time critical applications in Cloud, for GUI presentation during the application lifecycle, for customizing and optimizing virtual infrastructure, and for monitoring system functionality. In this context, the three pilot use case providers (BEIA/WT/MOG) actively bring the latest industrial requirements into the development cycle by fully exploring challenges in both current application scenarios and their future trends. Besides reporting those technical results in three technical deliverables on architecture design, semantic linking, and monitoring respectively, more than 10 papers have been published in a peered reviewed journal (FGCS) and a number of conferences (including CCGrid, eScience, CloudForward, WORKS etc.). Several workshops were co-organized by SWITCH partners to establish forums for discussing key issues in Cloud infrastructure (IT4RIs by UvA), time critical applications (WORKS by CU), and Cloud service models (CLASS by UL). Via the industrial partners of the project, the SWITCH concept and approach have also been presented in several exhibitions, including ServiceExpo (BEIA/MOG/UvA) and IBC (MOG).

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

To achieve the objectives of the project, the project partners proposed an architecture for SWITCH to support the development, provisioning and runtime control of time critical applications in a seamless environment. More specifically, progress has been made on the following aspects of the project:

1. A multi view based graphical interface in the SWITCH Integrated Development Environment (SIDE) based on the actives of the different roles of developers in the entire lifecycle of time critical applications. Current time-critical application programming models lack consideration of the controllability of the infrastructure; they thus do not exploit the potential benefits offered by the programmable infrastructure provided by Cloud environments. SIDE will improve the productivity of time critical cloud application development.

2. The core algorithms for customizing and optimizing virtual infrastructure have been investigated in the Dynamic Real-time Infrastructure Planner subsystem; an initial prototype has been developed based on the extension of a critical path algorithm. This technology will bring the programmability of the infrastructure into the process of application development.

3. The monitoring system developed in the Autonomous System Adaptive Platform (ASAP) subsystem aggregates information from the different levels of middleware between application and infrastructure, and employs a knowledge base to diagnose the system performance and to adapt the system behavior.

4. The semantic linking framework developed in the project models the profiles of application, virtual infrastructure, and the cloud resources, and links the semantic models of QoS/QoE at different levels. This framework serves all three subsystems, enabling the development, provisioning and runtime control of time critical cloud applications.
Besides these initial works, the project will focus on the advanced programming and control model for time critical cloud applications. The programming and control model, and the software tools developed in the SWITCH project, will have considerable impact on:
1) Improving development productivity of time critical Cloud applications. The market potential for time critical applications will be more than 11 billion USD if only looking at the three domains of real-time communication platform (4.8 billion), early warning systems (0.8 billion) and video broadcast (6 billion); given the anticipated rapid development of IoT and other ICT technologies, time critical applications or services will in fact be used in many more domains. The total market potential might be 10 times bigger than that of these three domains alone. Thus, productivity in developing such applications will be crucial in the near future.
2) Upgrading industrial technologies of time critical applications to use Cloud infrastructure. Classical domains requiring time critical applications such as early warning, live broadcast and business real-time communication face a strong push for bigger capacity, and better service quality to compete with new players in the market. There will be huge market demand for upgrading current systems and moving them into Clouds. SWITCH will have immediate impact on this part of the market.
3) Improving deployment efficiency of time critical applications. Rapid delivery and deployment of time critical services are themselves critical requirements for companies like WT, BEIA and MOG. The SWITCH environment can bring significant impact on the providers of time critical applications in relation to delivering and deploying services to their customers.
4) Reducing operational cost of time critical services. The operating cost for time critical services such as disaster early warning is, on average, 40% of the total system cost, and sometimes can rise to more than 80% in extreme cases. BEIA summarises the main problems encountered by its customers as concerning the very limited automated and intelligent adaptation that their

Related information

Record Number: 186593 / Last updated on: 2016-07-14