No abstract

In real distributed systems, processes may h a v e only inexact information about the amount of real time needed for primitive operations such as process steps. This thesis studies the eeect of this timing uncertainty on the real-time behavior of distributed systems. We consider a semi-synchronous model in which the amount of real time between process steps is known to be in the interval c 1 ; c 2 and every message is known to be delivered within time d of when it is sent. We use C = c 2 =c 1 as a measure of the timing uncertainty. We rst study the problem of reaching agreement in the presence of failures. A simple argument derived from the case of synchronous processes shows that at least time f + 1 d is required to tolerate f failures, while time f + 1 Cdis suucient to tolerate f stopping or omission failures by directly simulating the rounds of any synchronous consensus algorithm. We narrow this gap for omission failures, building on the nearly optimal algorithm of Attiya, Dwork, Lynch, and Stockmeyer which tolerates only stopping failures. If fewer than half the processes are faulty n 2 f + 1, then the running time of our algorithm is 4f + 1 d + Cd, which is within a factor of 4 of optimal and has minimal dependency on the timing uncertainty factor C. If more than half the processes are faulty, then a more complicated analysis shows the running time is increased by approximately a factor of min f n,f ; p C. We also present a general simulation for n 3f + 1 tolerant of Byzantine failures that simulates any synchronous algorithm at a cost of time 2Cd+dper round. Finally, motivated by the message ineeciency of our consensus algorithm for omission failures, we deene a more realistic model of message links by limiting their capacity. If messages are sent too frequently on these message links, they may incur delay greater than d. F or message links with capacity , w e prove nearly tight upper and lower bounds of min2Cd+d; C 2 d= + Cd+d and min2Cd+d=; C 2 d= + Cd+d respectively for the time needed to detect stopping failures. 3 Acknowledgments It is a pleasure to thank many friends and colleagues for their support and assistance during the research and writing of this thesis. My advisor Nancy Lynch has been an invaluable source of encouragement and enthusiasm ; I am grateful to her for having made this thesis possible. Cynthia Dwork taught m e m uch about the consensus problem; many hours of discussion with her have been enjoyable and enlightening. Hagit Attiya suggested numbering messages in the omissions algorithm. Baruch A w erbuch helped with the proof of Lemma 3.23. John Leo provided useful comments on earlier versions of Chapters 3 and 4. Thanks also to my friends for making the past few years more enjoyable, particularly

Real-time systems usually consist of a set of periodic and sporadic tasks. Periodic tasks can be divided into two classes: synchronous and asynchronous. The first type does not define the task first release, contrary to the second. Hence, synchronous periodic tasks are assumed to be released at the worst instant: the critical instant. The schedulability test is reduced to check a single execution of the task under analysis. The integration of sporadic tasks is also straightforward: they are treated as a periodic task with maximum arrival frequency. On the other hand, asynchronous periodic tasks require a test for each release in the hyperperiod and the integration of sporadic tasks is not trivial: the worst release instant is unknown a priori. However, they do not assume that the tasks are released at the worst instant.This paper presents a new schedulability analysis method based on the Response Time Analysis (RTA) to determine the worst response time of both asynchronous periodic and sporadic tasks, scheduled by a fixed-priority preemptive algorithm with general deadlines. It also presents another method that enables the introduction of a user configurable degree of pessimism, reducing the hyperperiod dependency.

No abstract

No abstract

No abstract

The ARINC 653 specification, defined for aeronautical applications, has the goal of providing a standard interface between a given real-time operating system (RTOS) and the corresponding applications. It also provides robust partitioning, with the final goal of guaranteeing safety and timeliness in mission-critical systems. The interest in ARINC 653 has extended to the aerospace industry, which resulted in the definition of an architecture, compliant with the specification, allowing for operating system heterogeneity. In this paper, we introduce the problem of integrating generic operating systems onto this architecture, and explore the case of GNU/Linux. Adding GNU/Linux allows running existing applications or interpreted scripts without needing to port the application or interpreter to an RTOS. In embedded systems, we have to cope with scarce resources and diverse existent hardware, and a balance between both issues must be reached. For such, we show the genesis of such a solution.

No abstract