Coordinated Allocation and Scheduling of Multiple Resources in Real-time Operating Systems
Kartik Gopalan and Kyoung-Don Kang
Workshop on Operating Systems Platforms for Embedded Real-Time applications (OSPERT 2007)
Pisa, Italy, July 3, 2007
Summary
Distributed real-time embedded (DRE) systems are key components of critical infrastructure including surveillance, target tracking, electric grid management, traffic control, avionics, and communications systems. They require (1)the coordinated management of multiple resources, such as the CPU, network, and disk, (2)end-to-end (E2E) real-time guarantees across the use of multiple resources, and (3)feedback control across multiple resources. None of these properties is supported as a first-class feature within the state-of-the-art real-time operating systems, but is left out as an inconvenient detail to be managed by DRE application programmers. In this paper, we shed light on this fundamental problem and make the case for greater research into the development of theory and a runtime systems for coordinated allocation and scheduling of multiple resources in real-time operating systems. We present the outlines of our proposed solution approach, called the Multiple Resource Allocation and Scheduling (MURALS) framework, that aims to bridge this gap between the need for E2E timing requirements and the techniques to coordinate the use of multiple resources. We discuss four key open research problems and possible solution strategies: (1)deadline partitioning techniques during admission control to efficiently partition E2E deadlines among multiple underlying resources, (2)explicit coordination among schedulers for multiple resources at runtime to guarantee the E2E deadlines, (3)statistical performance guarantees using measurement-based techniques that exploit the statistical multiplexing nature of resource usage and different tolerance levels to QoS violations, and (4)feedback control across multiple resources using control theoretic techniques to regulate QoS overshoot and settling time. We also outline our ongoing design and implementation of a proof-of-concept MURALS testbed that includes kernel-level coordinated multi-resource allocation and scheduling, measurement-based mapping of high-level QoS to low-level resources, and feedback control in addition to declarative APIs for E2E timing specification and QoS negotiation.