We propose and analyze a proportional share resource allocation algorithm for realizing real-time performance in time-shared operating systems. In a proportional share system, processes are assigned a weight which determines a share (percentage) of the resource they are to receive. The resource is then allocated in discrete-sized time quanta in such a manner that each process makes progress at a precise, uniform rate. Proportional share allocation algorithms are of interest because (1) they provide a natural means of seamlessly integrating real- and non-real-time processing requirements in a general purpose operating system, (2) they are easy to implement (and in particular, easier than more traditional forms of real-time support such as periodic tasks), (3) they provide a simple and effective means of precisely controlling the real-time performance of a process including uniform, predictable degradation in times of system overload, and (4) they provide a natural mean of policing proce...

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.

We describe a new framework for resource allocation that unifies the well-known proportional share and resource reservation policies. Each client is characterized by two parameters: a weight that represents the rate at which the client "pays" for the resource, and a share that represents the fraction of the resource that the client should receive. A fixed rate corresponds to a proportional share allocation, while a fixed share corresponds to a reservation. Furthermore, rates and shares are duals of each other. Once one parameter is fixed the other becomes fixed as well. If a client asks for a fixed share then the level of competition for the resource determines the rate at which it has to pay, while if the the rate is fixed, level of competition determines the service time the client should receive. To implement this framework we use a new proportional share algorithm, called Earliest Eligible Virtual Deadline First, that achieves optimal accuracy in the rates at which process execute...

We propose and analyze a new proportional share allocation algorithm for time shared resources. We assume that the resource is allocated in time quanta of size q. To each client, we associate a weight which determines the relative share from the resource that the client should receive. We define the notion of fairness in the context of an idealized system in which the resource is assumed to be granted in arbitrarily small intervals of time. Mainly, we show that in steady conditions our algorithm guarantees that the difference between the service time that a client should receive in the idealized system and the service time it actually receives in the real system is bounded by the size q of a time quantum. The algorithm provides support for dynamic operations, such as a client joining or leaving the competition (for the resource), and changing a client's weight. By using an efficient augmented binary search tree data structure we implement these operations in O(log n), where n represe...

Packet switching in connection-oriented networks that may have multiple parallel links between pairs of switches is considered. An e cient packet-scheduling algorithm that guarantees a deterministic quality of service toconnections with real-time constraints is proposed { this algorithm is a generalization of some recent multiprocessor scheduling algorithms, and o ers realtime performance guarantees similar to those o ered by earlier fair-scheduling strategies such as Weighted Fair Queueing and proportional-share schemes.