[Back]

C. Inzinger, W. Hummer, S. Dustdar:
"Customizable Reliability and Security for Data-Centric Applications in the Cloud, Part II";
Talk: Dagstuhl Seminar 13171: Customizing Service Platforms, Dagstuhl, Germany (invited); 04-21-2013 - 04-26-2013.

Service-oriented computing (SOC) has become a prevalent paradigm for
creating loosely coupled distributed applications and workflows. In
parallel to SOC, Event-Based Systems (EBS) in various fashions (e.g.,
data stream processing) are gaining considerable momentum as a means for encoding complex business logic on the basis of correlated, temporally decoupled event messages. More recently, advanced virtualization and resource allocation techniques advocated by Cloud computing have further shaped the implementation possibilities of SOC and EBS. Clouds have proven to be an ideal environment for flexible and elastic applications which provide scalability, resource optimization, and built-in support for multi-tenancy. Ongoing trends in the area of Data-as-a-Service (DaaS) have spurred further research efforts towards robust data processing services, leveraging the benefits of the Cloud.

Distributed computing systems in general, and applications in the Cloud in particular, are often burdened with stringent requirements concerning reliability and security, dictated by business objectives (e.g., cost-benefit tradeoffs), contractual agreements (e.g., service level agreements, SLAs), or laws. Customized support for reliability and security in service platforms is a key issue. One approach to reliability is software testing, which attempts to identify and avoid software-induced faults in the first place. A second important aspect of reliability is adaptability and fault-tolerance, which involves different runtime challenges such as fault detection, isolation, or
recovery. Additionally, due to the multi-tenancy inherently encountered in Cloud environments, security and access control play a crucial role for application provisioning. Consideration of these aspects in the software development and validation process requires precise knowledge about the type and nature of potential threats to reliability.

Within our work we tackle the aforementioned challenges and present novel techniques for reliable and secure provisioning of data-centric service platforms and applications in the Cloud. We strive for a robust, scalable, and secure execution environment for applications to integrate services and data from a plurality of sources, generating added value for service consumers. During the development phase, applications are systematically tested for incompatibilities and integration issues. At runtime, platforms should leverage
Cloud virtualization to ensure reliability and efficiency (elastic scaling, minimal resource allocation, optimized load distribution). Moreover, customized security policies need to be enforced to assure responsibilities and avoid unauthorized access.