This paper presents stride scheduling, a deterministic scheduling technique that efficiently supports the same flexible resource management abstractions introduced by lottery scheduling. Compared to lottery scheduling, stride scheduling achieves significantly improved accuracy over relative throughput rates, with significantly lower response time variability. Stride scheduling implements proportional-share control over processor time and other resources by cross-applying elements of rate-based flow control algorithms designed for networks. We introduce new techniques to support dynamic changes and higher-level resource management abstractions. We also introduce a novel hierarchical stride scheduling algorithm that achieves better throughput accuracy and lower response time variability than prior schemes. Stride scheduling is evaluated using both simulations and prototypes implemented for the Linux kernel.

This thesis presents flexible abstractions for specifying resource management policies, together with efficient mechanisms for implementing those abstractions. Several novel scheduling techniques are introduced, including both randomized and deterministic algorithms that provide proportional-share control over resource consumption rates. Such control is beyond the capabilities of conventional schedulers, and is desirable across a broad spectrum of systems that service clients of varying importance. Proportional-share scheduling is examined for several diverse resources, including processor time, memory, access to locks, and disk bandwidth. Resource rights are encapsulated by abstract, first-class objects called tickets. An active client consumes resources at a rate proportional to the number of tickets that it holds. Tickets can be issued in different amounts and may be transferred between clients. A modular currency abstraction is also introduced to flexibly name, share, and protect ...

Enabling technologies in high speed communication and global process scheduling have pushed clusters of computers into the mainstream as general-purpose high-performance computing systems. More generality, however, implies more sharing and this raises new questions in the area of cluster resource management. In particular, in systems where the aggregate demand for computing resources can exceed the aggregate supply, how to allocate resources amongst competing applications is an important problem. Traditional solutions to this problem have focused mainly on global optimization with respect to system-centric performance metrics, metrics which ignore higher level user intent. In this paper, we propose an alternative market-based approach based on the notion of a computational economy which optimizes for user value. Starting with fundamental requirements, we describe an abstract architecture for market-based cluster resource management based on the idea of proportional resource sharing of...

This paper presents a general method for computing quantitative information about finite-state real-time sys-tems. We have developed algorithms that compute exact bounds on the delay between two speci$ed events and on the number of occurrences of an event in a given inter-val. This technique allows us to determine performance measures such as schedulability response time, and system load. Our algorithms produce more detailed information than traditional methods. This information leads to a better understanding of system behavior in addition to determin-ing its correctness. The algorithms presented in this paper are efficiently implemented using binary decision diagrams and have been incorporated into the SA4V symbolic model checker. Using this method, we have verified a model of an aircraft control system with lOI states. The results obtained demonstrate that our method can be successfully applied in the verification of real-time system designs.

Abstract: Symbolic model checking is a technique for verifying finite-state concurrent systems. Models with up to 10 a ° states can often be verified in minutes. In this paper, we present a new tool to analyze real-time systems, based on this technique. We have designed a language, called Verus, for the description of real-time systems. Such a description is compiled into a state-transition graph and represented symbolically using binary decision diagrams. We have developed new algorithms for exploring the state space and computing quantitative information about the system. In addition to determining the exact bounds on the length of the time interval between two specified events, we compute the number of occurrences of an event in such an interval. This technique allows us to determine performance measures such as schedulability, response time, and system load. Our algorithms produce more detailed information than traditional methods. This information leads to a better understanding of the behavior of the system, in addition to verifying if its timing requirements are satisfied. We integrate these ideas into the Verus tool, currently under development. To demonstrate how our technique works, we have verified a robotics control system. The results obtained demonstrate that our method can be successfully applied in the analysis of realtime system designs. 1

This paper proposes efficient scheduling algorithms for the joint scheduling of hard aperiodic, sporadic and periodic real time tasks, in systems based on preemptive, fixed-priority dispatching. Our scheme guarantees or rejects hard aperiodic real time tasks without any prior knowledge of their attributes, by managing the idle processor capacity dynamically. The method assigns fixed priorities to periodic tasks based on the Deadline Monotonic (DM) scheme and analyzes their schedule off-line. We derive closed form solutions for the idle processor capacity process Z(t) within a schedule. In the absence of pending dynamic tasks, periodic tasks execute in their earliest possible schedule S , called the Fixed-Priority First (FPF). Upon the arrival of a non-periodic task Ja , the scheduler directly determines its admissibility, based on closed form equation of the available processor time in the current schedule, until the deadline of Ja . If Ja cannot be guaranteed under FPF, the scheduler evaluates the idle processor capacity of an alternative schedule S , where periodic tasks are delayed to execute at their latest possible times, called the Latest Deadline Last (LDL). If LDL offers sufficient idle capacity, the scheduler switches all periodic tasks from FPF to LDL, assigns Ja the lowest priority and admits it into the system. Otherwise, it immediately rejects Ja . We develop the theoretical framework and derive efficient algorithms to compute the idle processors capacity Z(a; b) within a time interval [a; b], and maintain it when the schedule is adjusted. The algorithms can also reclaim unused capacity from guaranteed tasks. Our admission control procedure has computational complexity \Theta(n) when the non-periodic task queue is serviced in FIFO orde...

In this work we propose a verification methodology consisting of selective quantitative analysis and interval model checking. Our methods can aid not only in determining if a system works correctly, but also in understanding how well the system works.

When developing multitasking real-time systems, schedulability tests are used to formally prove that a given task set will meet its deadlines. A wide range of such tests have appeared in the literature. This tutorial acts as a guide to the major tests available for preemptive multitasking applications.

Error-free multimedia data processing and communication includes providing guaranteed services such as the colloquial telephone. A set of problems have to be solved and handled in the control-management level of the host and underlying network architectures. We discuss in this paper `resource management' at the host and network level, and their cooperation to achieve global guaranteed transmission and presentation services, which means end-to-end guarantees. The emphasize is on `network resources' (e.g., bandwidth, buffer space) and `host resources' (e.g., CPU processing time) which need to be controlled in order to satisfy the Quality of Service (QoS) requirements set by the users of the multimedia networked system. The control of the specified resources involves three actions: (1) properly allocate resources (end-to-end) during the multimedia call establishment, so that traffic can flow according to the QoS specification; (2) control resource allocation during the multimed...

The task of checking if a computer system satisfies its timing specifications is extremely important. These systems are often used in critical applications where failure to meet a deadline can have serious or even fatal consequences. This paper presents an efficient method for performing this verification task. In the proposed method a real-time system is modeled by a state-transition graph represented by binary decision diagrams. Efficient symbolic algorithms exhaustively explore the state space to determine whether the system satisfies a given specification. In addition, our approach computes quantitative timing information such as minimum and maximum time delays between given events. These results provide insight into the behavior of the system and assist in the determination of its temporal correctness. The technique evaluates how well the system works or how seriously it fails, as opposed to only whether it works or not. Based on these techniques a verification tool called Verus...