Learning Module II: Real-Time Systems Design

Transcription

1 Learning Module II: Real-Time Systems Design

2 Outline Introduction to real-time systems Timing requirements and timing analysis - Concept of Worst-Case Execution Time (WCET) - Why it is hard for analyzing WCET? - Overall approach: modularization - Program path analysis - Static analysis Uniprocessor scheduling algorithms - Task models - Performance metrics of scheduling algorithms - Static scheduling algorithms - Dynamic scheduling algorithms Multicore and distributed real-time systems - Multiprocessor scheduling - Distributed real-time systems What happens when things go wrong?

3 Suggested Readings Books: Giorgio C. Buttazzo, Hard Real-Time Computing Systems: Predictable Scheduling Algorithms and Applications, Springer, Second Edition, ISBN: Jane W.S. Liu, Real-Time Systems, Prentice Hall, First Edition, ISBN: John A. Stankovic, Marco Spuri, Krithi Ramamritham, Giorgio C Buttazzo, Deadline Scheduling for Real- Time Systems: EDF and Related Algorithms, Springer Science & Business Media, ISBN Papers: Robert I. Davis and Alan Burns A survey of hard real-time scheduling for multiprocessor systems. ACM Comput. Surv. 43, 4, Article 35 (October 2011), 44 pages. Sanjoy K. Baruah, Kirk Pruhs: Open problems in real-time scheduling. J. Scheduling 13(6): (2010)

4 Outline Introduction to real-time systems Timing requirements and timing analysis - Concept of Worst-Case Execution Time (WCET) - Why it is hard for analyzing WCET? - Overall approach: modularization - Program path analysis - Static analysis Uniprocessor scheduling algorithms - Task models - Performance metrics of scheduling algorithms - Static scheduling algorithms - Dynamic scheduling algorithms Multicore and distributed real-time systems - Multiprocessor scheduling - Distributed real-time systems What happens when things go wrong?

5 A real-time system is a system whose specification includes both logical and temporal correctness requirements. Logical correctness ( the results are correct ) - Requires functional analysis Temporal correctness ( the results are delivered in/on time ) - Requires non-functional analysis High reactivity and high dependability are more important than performance In these four lectures, we focus on techniques and technologies for achieving and checking temporal correctness.

6 When we design real-time systems, we need to consider how much system resources we have to realize the timing requirements of the applications. The Window of Scarcity Resources may be categorized as: Abundant: Virtually any system design methodology can be used to realize the timing requirements of the application. Insufficient: The application is ahead of the technology curve; no design methodology can be used to realize the timing requirements of the application. Sufficient but scarce: It is possible to realize the timing requirements of the application, but careful resource allocation is required.

7 Scarcity of Resources Requirements (performances, scale, etc.) Interesting Region for Real Time Systems Design Interactive Video High-quality Audio Insufficient resources Sufficient but scarce resources Network File Access Abundant resources Remote Login Hardware resources in year X

8 Examples for Real-Time Systems - Chemical & Nuclear Power Plants - Railway Switching Systems - Flight Control Systems - Space Mission Control - Automotive Systems - Healthcare Systems - Robotics - Telecommunications Systems - Stock Market and Trading System - Multimedia Systems - Virtual Reality... Hard Real-Time Systems Catastrophic results if some deadlines are missed Firm Real-Time Systems The results are useless if the deadlines are missed Soft Real-Time Systems The results are not very useful if the deadlines are missed

9 Typical Characteristics of Real-Time Systems - Timeliness - High cost of failure - Concurrency/multiprogramming - Stand-alone/continuous operation - Design for worst cases - Reliability/fault-tolerance requirements - Predictable behavior

10 Frequent Misconceptions about Real-Time Systems There is no science in real-time system design. - We shall see in the following lectures. Real-time computing is equivalent to fast computing. - Real-time computing means predictable and reliable computing. Real-time is performance engineering/tuning. - Timeliness is more important than raw performance. Advances in hardware will take care of real-time requirements. - Buying a faster processor may result in timeliness violation.

11 Frequent Misconceptions about Real-Time Systems (Cont.) Real-time programming is assembly coding. - We would like to automate (as much as possible) real-time systems design Real-time problems have all been solved in other areas of computer sciences and operations research. - OR people typically use stochastic queuing models or one-shot scheduling models to reason about systems. - In other CS areas, people are usually interested in optimizing average-case performance. Real-time systems function only in a static environment. - We also consider systems in which the environment may change dynamically.

12 Real-Time Systems are Stimulus/Response Systems Given a stimulus, the real-time system must produce a response within a specified time (timing constraint). Periodic stimuli. Stimuli which occur at predictable time intervals - For example, a temperature sensor may be polled 10 times per second. Aperiodic stimuli. Stimuli which occur at unpredictable times - For example, a system power failure may trigger an interrupt which must be processed within the timing constraint by the system.

13 Architectural Considerations for Real-Time Systems Because of the need to respond to timing demands made by different stimuli/responses, the system architecture must allow for fast switching between stimulus handlers. Timing demands of different stimuli are different so a simple sequential loop (cyclic executive) is not usually adequate. Real-time systems are therefore usually designed as cooperating concurrent processes with a real-time executive controlling these processes.

14 Reasons for Needs of Concurrency Functional - allow multiple users - perform many operations concurrently Performance - take advantage of blocking time - parallelism in multi-processor machines Expressive Power - many control applications are inherently concurrent - concurrency support helps in expressing concurrency, making application development simpler

15 Multi-tasking The execution entities (tasks, processes, threads, etc.) are competing from each other for shared resources. Scheduling decision policy is needed - When to schedule an entity? - Which entity to schedule? - How to schedule entities? The focus of the following lectures is to design and analyze scheduling algorithms for different settings in real-time systems.

16 Real-Time System Design Process Identify the stimuli to be processed and the required responses to these stimuli. For each stimulus and response, identify the timing constraints. Aggregate the stimulus and response processing into concurrent processes. A process may be associated with each class of stimulus and response. Design algorithms to process each class of stimulus and response. These must meet the given timing requirements. Design a scheduling system which will ensure that processes are started in time to meet their deadlines. Integrate using a real-time operating system (RTOS).

17 Real-Time Operating Systems Real-time operating systems are specialised operating systems which manage the processes in real-time systems. Responsible for process management and resource (processor and memory) allocation. May be based on a standard kernel which is used unchanged or modified for a particular application. Do not normally include facilities such as file management.

18 Examples of RTOS QNX: VxWorks: LynxOS: Windows CE: embos: RTAI: FreeRTOS: μc/os-ii:

19 The Focus of these four lectures: Scheduling Algorithms for Real-Time Systems - Earliest-Deadline-First Scheduling, Rate Monotonic Scheduling, Deadline Monotonic Scheduling, etc. in uniprocessor system - Global scheduling and partitioned scheduling in multiprocessor systems Analysis of Timeliness in Real-Time Systems Worst-case Execution Time (WECT) analysis Schedulability analysis of scheduling algorithms After these four lectures, you are expected to know Fundamental scheduling theories in real-time systems Schedulability analysis of scheduling algorithms in real-time systems