End-to-end scheduling to meet deadlines in distributed systems

Abstract

In a distributed system or communication network tasks may need to be executed on more than one processor. For time-critical tasks, the timing constraints are typically given as end-to-end release-times and deadlines. This paper describes algorithms to schedule a class of systems where all the tasks execute on different processors in turn in the same order. This end-to-end scheduling problem is known as the flow-shop problem. We present two cases where the problem is tractable and evaluate a heuristic for the N P-hard general case. We generalize the traditional flow-shop model in two directions. First, we present an algorithm for scheduling flow shops where tasks can be serviced more than once by some processors. Second, we describe a heuristic algorithm to schedule flow shops that consist of periodic tasks. Some considerations are made about scheduling systems with more than one flow shop.

CaDAnCE: A Criticality-Aware Deployment and Configuration Engine

“…CaDAnCE does not affect the QoS behavior of operational strings. When components are promoted from a lower-criticality operational string to a higher-criticality one, component criticality is also increased to match the criticality of the higher-criticality string, which is essential to avoid criticality inversion at deployment-time [6]. Since CaDAnCE promotes components only at deployment time, it does not change the actual real-time QoS configurations (such as thread priorities, component placement and collocation) used later during runtime, i.e., CaDAnCE does not affect the QoS behavior of operational strings.…”

Minimizing response times of automotive dataflows on multicore

“…A number of methods for modeling the real-time behavior of dataflow applications have been developed for multiprocessor and distributed systems [4,12,15,23]. Typically, such applications are modeled as directed acyclic graphs (DAGs), with nodes denoting tasks and edges denoting producer/consumer relationships.…”

Connection-oriented real-time communication for mission critical applications-an introduction to NetEx: a portable and efficient tool kit

“…(8) follows directly from the delay property of maximum rate function [1]. Using (8), subsequent servers can be analyzed. Once all the delays at individual servers are obtained, we can substitute (7) into (1) to compute the delay for connection j.…”

“…The generic solution for guaranteeing endto-end delay bounds in distributed systems consists of connection-oriented communications with some form of admission control and traffic regulation (typically based on packet scheduling at the network interface) [7][8][9][10][11][12][13][14]18]. Supplementary research in protocols for guaranteed services on networks is reported in [15][16][17].…”