RTS Ptu Sample paper lords mbd

pdfcrowd.comopen in browser PRO version Are you a developer? Try out the HTML to PDF API

Real Time SystemsRecommend 6k

Press Ctrl & '+' To enlarge text and pics!Chapter Notes!

HomeTOPICSChapter 1 : Introduction RealTime SystemChapter 2 : Characterizing RealTime System and TaskChapter 3 : Task Assignmentand SchedulingChapter 4 : ProgrammingLanguages and ToolsChapter 5 : Real Time DatabaseChapter 6 : Real TimeCommunicationChapter 7 : Fault - ToleranceTechniquesChapter 8 : ReliabilityEvaluation TechniquesChapter 9 : ClockSynchronization

Home

Chapter 1 : Introduction Real Time System

Q1.What is real time system?

Ans.

Any system where a timely response by the computer to external stimuli in vital is a real time system.

Example: A CAR-AND-DRIVE EXAMPLE

If we consider a familiar problem of human control driving a car. The driver is the real-time controller, the car is the controlled process, and the other cars togetherwith the road conditions make up the operating environment.

Q2 Explain all types of system? Ans.We have two types of real time system:(1)Soft real time system.(2)Hard real time system. (i) Hard real time system: The requirement that all hard timing constraints must be validated invariablyplaces many restrictions on the design and implementation of hard real-time applications as well as thearchitectures of hardware and system software used to support them.

(ii)Soft real time system: A system in which jobs have soft deadlines is a soft-real system. Thedeveloper of a real time system is surely required to prove rigorously that the system surely mee its realtime performance objective.

Example: On line transaction system, telephone switches as well as electronic games.

http://www.addtoany.com/share_save#url=http%3A%2F%2Fptucse.loremate.com%2Frts%2Fnode%2F3&title=Chapter%201%20%3A%20Introduction%20Real%20Time%20System&description=

http://ptucse.loremate.com/rts/

http://ptucse.loremate.com/rts/node/11










http://loremate.com/


http://pdfcrowd.com/html-to-pdf-api/?ref=pdf

http://pdfcrowd.com/customize/

http://pdfcrowd.com/redirect/?url=http%3a%2f%2fptucse.loremate.com%2frts%2fnode%2f3&id=ma-130126084340-71a73a5d

http://pdfcrowd.com


Example: On line transaction system, telephone switches as well as electronic games.

Q3. What are the issues in real time system? Ans.

A real time computer must be much more reliable than its individual hardware and softwarecomponent. It must be capable of working in harsh environments.

For example: Take task scheduling. Real time computer systems differ from their general-purposecounterparts in two important ways. Firstly they are much more specific in their applications, and second,the consequences of their failure all more drastic.

Q4. Define real time database?

Ans.

The term real time database system refers to a diverse spectrum of information systems, rangingfrom stock price quotation systems, to tracks records databases, to real time file systems a real timedatabase contains data objects, called Image objects, that represent real-world object the attributesof an image object are these of represented real world object for example an air traffic controldatabase contains image objects that represent aircraft in the courage area.

Q5. Explain release time, deadlines and timing Constraints?

Ans. Release Time: The release time of job is the instant of time at which the job becomes available forexecution. The job can be scheduled and executed at any time at or after its release time whenever itsdata and control dependency conditions are met.As an example: We consider a system which monitors and Controls several furnaces.

Dead lines: The dead lines of a job is the instant of time by which its execution is required to becomputed suppose that in the pervious example, each control low computation job must complete by therelease time of the subsequent job. Then, their deadlines are 120 msec, 220 msec and so on, respectively.

Timing Constraints: A timing constraints of a job can be specified in termS of its release time andrelative or absolute deadlines as illustrated by the above example: some complex timing constraintscannot be specified conveniently in terms of release times and deadlines.




http://pdfcrowd.com


Q6. Describe Trigger Generator? Ans.The “Trigger generator” is a representation at the mechanism used to trigger the execution of individualjobs. It is not really a separate hardware unit, typically it is part of the executive software Many of the jobsare periodic i.e. they execute regularly. The Schedule for these jobs can be obtained offline and loaded asa look up table to used by the scheduler.

Q7.Draw structure or block diagram of Real time system OR the components of the RTS. Ans. Schematic block diagram of a Real time system.

Q.8. Explain all types of tasks classes in real time system?

OR

Explain the concept of periodic and a periodic task with the help of suitable example? Ans. There are five types of task classes:(i) Periodic and aperiodic tasks




http://pdfcrowd.com


(i) Periodic and aperiodic tasks(ii) Sporadic task(iii) Critical task(iv) Noncritical task

(1) Periodic task: There are many tasks in real — time systems that are done repetitively. For exampleone may wish to monitor the speed altitude and attitude of an aircraft every 100 ms. This sensorinformation will be used by periodic tasks that control and control surfaces of the aircraft (e.g., theailerons, elevator, rudder and engine thrusts), in order to maintain stability and other desiredcharacteristics. The periodicity of these tasks is known to the designer, and much tasks can be pre-scheduled. (2) Aperiodic task: There are many other tasks that are aperiodic, that occur only occasionally. Forinstance, when the pilot wishes to execute a turn a large number subtasks. Associated with that action areself off aperiodic tasks cannot be predicated and sufficient completing power must be held in reserve toexecute them in a timely fashion. Aperiodic task with bounded imperatival time are called sporadic tasks.

(3) Critical tasks: Critical tasks are those whose timely executions is critical; if deadlines are missedcatastrophes occur. Example include life -support systems and the ctahility control of air craft. Criticaltasks, are often executed at a higher frequency than is absolutely necessary. (4) Non critical tasks: Non critical tasks are real times tasks, the name implies not critical to theapplication. However they do deal with time varying data and hence they useless if not completed within adeadline, the goal in scheduling these tasks is these to maximize the percentage of jobs successfullyexecuted within their deadlines.

Q.9. Explain all structure of Real time system? ORExplain the architecture of Real time system How can you classify the tasks for the real timesystem? Ans. Structure of a real system:

The state of the controlled process and of the operating environment (e.g., pressure, temperature,speed and attitude) is acquired by sensors, which provide input to the controller, the real timecomputer. There is a fixed set of application tasks or jobs, the job list”. The software for these tasks is preloaded into the computer. If the computer has a should main memory, then the entire software is loaded into that. The “trigger generator” is a representation of the mechanism used to trigger the execution ofindividual jobs. It is not really a separate hardware unit typically it is part of the executive software many of jobs areperiodic, the schedule for these jobs can be obtained offline and loaded as a lookup table to be used bythe schedule.




http://pdfcrowd.com


A schematic block diagram of a real time system. There are following phases of civilian aircraft.

(1) Take off and crisis until VHF omni range VOR/DME (distance measuring equipment) is out ofrange.(2) Cruise until VOR/DME is in range again.(3) Cruise until landing is to be initiated.(4) Land.

Q.10. Explain all types of Issues in real time system? Ans. Architecture Issues:(a) Processor architecture: For reasons of economy, off-the-shelf processors are preferred, real timedesigns seldom design their own processor, for this reason, we do not treat processor architectural.

(b) Network architecture: To make systems reliable provide sufficient processing capacity, most realtime systems are multiple processor machine.

(c) Architecture for clock synchronization: In order to facilitate the interaction between the multipleunits of a real time system, the clocks of these unit must be synchronized & tightly(d) Fault-tolerance and reliability evaluation: People can die when real — time system fail such systemsmust therefore be liguly fault-tolerant.Operating system Issues.

(a) Task assignment and scheduling: The scheduling of tasks t ensure that real time dead lines aremet is central to the mission of a real-time operating system. (b) Communication protocols: It is important to - have interprocessor communication that the has




http://pdfcrowd.com


(b) Communication protocols: It is important to - have interprocessor communication that the haspredictable delays and is controllable. (c) Failure management and recovery: when a processor or software module fails, the system mustlimit such failure and recover from it. (d) Clock synchronization algorithm: We Mentioned a hardware synchronization architecture built outof phase — locked locks. Other Issues(a) Programming languages: Real time engineers need much greater control our timing and used tointerface to special purpose device.

(b) Data bases: There are many real — time data-base applications, such as the stock market, airline,reservations and artificial, Intelligence.

(c)Performance measures: Commonly used performance measures such as conventional reliability andthrough put are useless for real time systems

Loremate.com

Like 6,065

http://goo.gl/EZv6H

http://www.facebook.com/lorematecom





http://pdfcrowd.com





Home

Chapter 2 : Characterizing Real Time System and Task

Q1. Define Reliability? Ans. Reliability is the Probability that the system will not under go failure over any part of prescribedinterval. The computational reliability R* (S, t, T) is the probability that the system can start a task. T attime t tosuccessfully execute it to completion the system can start state at time t is s. Which is assumed to be afunctional state. Related to this is the computational availability. Which is the expected value of the systemcomputation capability at any given time.

Q2.Define Availability?

Ans. Availability is the fraction of time for which the system up. Throughput is the average number ofinstructions pr unit time that the system cam process.

Q3. Define Performability?

Ans.Performability improves upon the traditional measures by explicitly and formally accounting for the factthat the performance of a real time computer should be tied to the consequent performance of the processthat it controls the controlled process is defined as having several accomplishment levels.

Q4.Define Virtual Memory?

http://www.addtoany.com/share_save#url=http%3A%2F%2Fptucse.loremate.com%2Frts%2Fnode%2F4&title=Chapter%202%20%3A%20Characterizing%20Real%20Time%20System%20and%20Task&description=
















http://pdfcrowd.com/redirect/?url=http%3a%2f%2fptucse.loremate.com%2frts%2fnode%2f4&id=ma-130126084647-71b2a8be

http://pdfcrowd.com


Ans.Virtual memory is a major source of execution time uncertainly. flat is why it is wise to avoid using virtualmemory whenever possible. The time taken to handle page faults for example, can very widely andobtaining a good bound on the page fault rate is almost impossible.

Q5.Describe Performance Measure?

Ans.Performance Measures are yardsticks by which perfonnance expressed. They are, in a very real sense,languages through which we convey system performance. A performance measures must concise. It mustencapsulate in very few numbers (preferably a single number) the performance of a system.

Q6. Explain Properties of Performance Measures?

Ans. A good perfonnance measure must:(1) Represent an efficient encoding of relevant information.(2) Provide an objective basis for the Ranking of candidate controllers for a given application.(3) Provide objective optimization criteria for design.(4) Represent verifiable facts.

Q7. Define cost functions? Ans.These functions are obtained by comparing the Performability at a real lime system with a zero Reposestime to a system with given positive response times. The cost function at a particular task execution givenby:

Where P(x) is the perfonnability associated with a response time of

Q8. Define Compiler?




http://pdfcrowd.com


Ans.The Compiler maps the source level code into a machine level program. This mapping is not unique, theactual mapping will depend on the actual implementation of the particular compiler that is being used. Theexecution time will depend on the nature of the mapping.

Q9.Define operating system?

Ans.The operating system determines such issues as task scheduling and memory management, both of whichhave major impact on the execution time. Although with machine architecture, it determines the interrupthandling overhead.

Q10. Explain Functions and Hard deadlines?

Ans.Real-time tasks frequently have hard deadlines the time by which they must finish executing if catastrophicfailure of controlled process is to be avoided. They also have cost functions of the response time. Thesefunctions are obtained by comparing the perfonnability of a real-time system with a zero response time toa system with given positive response times. The cost function of a particular task execution is given by

Where P(x) is the performability associated with a response time of Cost function and hard deadlinesare best explained through a simple and idealized.

Q11. Explain the factor which estimating program run times?

Ans.Its depend upon following factors:(1) Source code: Source code that is carefully turned and optimized takes ten time to execute.

(2) Complier: The compiler maps the source — level code into a machine level program.

(3) Machine architecture: Many aspects of the machine architecture have an effect on the executiontime that is difficult to quantity exactly. Executing a program may require much interaction between the




http://pdfcrowd.com


time that is difficult to quantity exactly. Executing a program may require much interaction between theprocessor(s) and memory and I/O devices.

(4) Operating system: The operating system determines such issues as task scheduling and memorymanagement, both of which have a major impact on the execution time. Along with the machinearchitecture. It determines the interrupt handling overhead.

Q12.What are the performance measures for Real time systems and explain its properties?

Ans. Real time systems are often used in critical applications, and must therefore be carefully designed andvalidated before being put into, operation the validation process includes checking design correctness usingformal and informal methods and characterizing performance and reliability. Performance measures areyardsticks by which performance is expressed. They are, in a very real sense, language through which weconvey system performance.

A performance measure must be concise. That is, it must encapsulate in very few numbers, theperformance of a system.

Example: Consider two system A and B. A And B have system response times with the probability densityfunctions shown system A is clearly more predictable than system B and is only slightly shown flowers, ifthe mean response time’ is our performance measure, we rank B higher than A. Properties of performance Measures:

(1) A good performance measure must. Represent an efficient encoding of relevant information.

(2) Provide an objective basis for the ranking of candidate controllers for a given application.

(3)Provide objective optimization criteria for design.

(4)Represent verifiable facts.




http://pdfcrowd.com


(i) Efficient encoding: One of the problems of dealing with complex systems is the volume ofinformation that is available about them and interactions with the environment. Determining the relevanceof individual pieces of data is impossible unless the data are viewed within a certain context or framework. Such a context suppresses the irrelevant and light lights the relevant.

(ii)Objective Basis for ranking: Performance measures must by clyination, quantify the goodness ofcomputer systems in a given application or class of applications. (iii)Objective optimization criteria: The more Complex a system, the more difficult it is to optimize orto fine- tune its structure. There are numerous side effects of even simple actions, for example changingthe number of buses in the computer.

(iv)Verifiable facts: A performance measure that is impossible to derive is of no use to anybody. To beacceptable, a performance measure should hold outprospect of being estimated reasonablysome accurately.

Loremate.com

Like 6,065

http://goo.gl/EZv6H






http://pdfcrowd.com





Home

Chapter 3 : Task Assignment and Scheduling

Q1. Explain Tasks?

Ans.Tasks may have precedence constraints, which specific if any task(s, needs to precede other tasks. If taskT’ s output is needed as input by task T, then task T is constrained to be preceded by task T- each task hasresource requirements. All tasks require some execution time on a processor.

Q2. What is relative deadline? Ans.The relative deadline of a task is the absolute deadline minus the release time that is, if task T has arelative deadline d1 and is released at lime a, it must be executed by time t ± d. The absolute deadline isthe lime by which the task must be completed. Example is the absolute deadline of T1 is t+d.

Q3. Define schedule? Ans.Schedule is said to be feasible if all the tasks start after their rules times and complete before theirdeadlines a sit of tasks is A- feasible if cm assignment scheduling algorithm A, when run on that set oftasks, result in a feasible schedule The bulk of the work in real — time scheduling deals with obtainingfeasible schedules.

Q4. Explain static and dynamic priority algorithm?

http://www.addtoany.com/share_save#url=http%3A%2F%2Fptucse.loremate.com%2Frts%2Fnode%2F5&title=Chapter%203%20%3A%20Task%20Assignment%20and%20Scheduling&description=
















http://pdfcrowd.com/redirect/?url=http%3a%2f%2fptucse.loremate.com%2frts%2fnode%2f5&id=ma-130126084832-bf4179c7

http://pdfcrowd.com


Ans.There are algorithms that assumes that the task priority does not change with in a mode; these and calledstatic — priority algorithm dynamic — priority algorithms assume that priority car change with time, thebest known examples of static — and — dynai nic — priority algorithms are the rate — monotonic (RM)algorithm and the earliest deadline first (EDF) algorithm.

Q5. Define Myopic offline scheduling (MOS) Algorithm?

Ans.The myopic. offline scheduling (MOS) heuristic is an assignment/scheduling algorithm meant for nonpreemptive tasks. This algorithm tasks account not of processing needs but also of any requirements thattasks may have for additional resources. MOS is an affline algorithm in that

iii. Explain Bin Packing Algorithm?. The bin Packing algorithm assigns tasks to processors under the constraint that the total processorutilization must not exceed a given threshold. The threshold is set in such a way that the uniprocessorschedule the tasks arrived to each processor. Tasks assumed to be premeetiable.

Q7. Define Buddy strategy?

Ans.The buddy strategy takes roughly the same approach as the focused addressing algorithm. Processors aredivided into there categories:under loaded, fully loaded overloaded processors to after to take oversome of their load.

Q.8.Explain Next — fit Algorithm?

Ans.The Next — fit Algorithm is designed to work in conjunction with the rate monotonic uniprocessorscheduling algorithm. It divides the set at tasks into various classes. A set of processors is exclusivelyassigned to each task class. Tasks are assumed to be premeetiable.

Q.9.Explain utilization Balancing Algorithm?

Ans.This algorithm assigns tasks to .Processors one by one in such a way that at the end of each step, theutilizations of the various processorsare as nearly balanced as possible. Tasks are assumed to be preemptiable.




http://pdfcrowd.com


Q10. Difference between preemptive and non preemptive scheduling? Ans. Non preemptive schedule according to the same priority assignment.This schedule is same as the preemptive schedule. However at time 4 when is in released both processorsare busy. It has to wait until by computer before it can begin execution.

Example of priority-driven scheduling. Non preemptive scheduling is not better than preemptivescheduling. Preemptive scheduling is better when the cost of preemption is ignored. Specifically, in amultiprocessor system.

Q11. Explain Rate Monotonic Algorithms?

Ans.A will — known fixed — priority algorithm is the rate monoto algorithm. This algorithm assigns priorities totasks based on the periods. The shorter the period, the higher the priority the rate of a is the inverse of itsperiods. Hence the higher its rate the higherwill refer to this algorithm as RM the algorithm for short and a e produced by algorithm as a RM schedule.example: the RM schedule of the system whose cyclic schedule is




http://pdfcrowd.com


example: the RM schedule of the system whose cyclic schedule issystem, contains three tasks T1 (4, 1), T2 = (5, 2) T3 = (20, 5) thepriority of T1 is the highest because its rate is the higher each job in the

task is placed at head of the priority. Another well—known fixed priority algorithm is deadlines — monotonic algorithm called the DM algorithmhereafter. This algorithm assigns priorities to tasks according this relative deadlines. The shorter relativedeadlines the higher priority.

Q12. Explain uniproeessor scheduling and multiprocessor scheduling. Ans.Uniprocessor scheduling is part of the process of developing multiprocessor schedule our ability to obtain afeasible multiproces schedule is therefore linked to our ability to obtain f - uniprocessor schedule.




http://pdfcrowd.com


feasible multiproces schedule is therefore linked to our ability to obtain f - uniprocessor schedule.

(1) Traditional rate — monotonic (RM): The task ;et consists of periQdic, preemptiable tasks whosedeadlines equal the task period. A task set of n tasks is schedulable under RM if its total processorutilization is not greater than n(2’-l)(2) Rate — monotonic deferred snuer (DS): This is similar to the RM algorithm except that it can handleboth periodic andaperiodic tasks.(3) Earliest deadline first (EDF): Tasks are preemptible and the task with the earliest deadline has thehighest priority. EDF isan optimal uniproussor algorithm.4) Precedence and exclusion conditions: Both the RM and EDF algorithms assume that the tasks oneIndependent andpreemptive anytime.(5) Multiple task unisons: The system has primary and alternative versions vary in their execution timeand in the quality of output they provide. Primary versions are the full — fledged tasks, providing top —quality output.Multiprocessor Scheduling:(1) Utilization balancing algorithm: This algorithm assigns tasks to processors one by one in such a waythat at the end of eachstep,(2) Next — fit algorithm: The next fit algorithm is designed to work in conjunction with the rate monotonicuniprocessor schedulingalgorithm. It divides the set of tasks into various classes.(3) Bin — packing algorithm: The bin — packing algorithm assigns tasks to processor under the constraintth,t the total processorutilization must not exceed a given threshold.(4)Myopic offline scheduling algorithm: This algorithm can deal with non-preemptive tasks, It builds up the




http://pdfcrowd.com


Myopic offline scheduling algorithm: This algorithm can deal with non-preemptive tasks, It builds up theschedule using a starch process.(5)Focused addressing and building algorithm: In this algorithm, tasks are assumed to arrive at the individualprocessors. A processor that finds itself unable to meet the deadline or other constraints of all its taskstries to offload some of its workload onto other processors.(6)Buddy strategy: The buddy strategy tasks roughly the some approach as the focused addressing algorithm.Processors are divided in to three categories under loaded, fully loaded and ourhead-our loaded processorsask the under loaded processors ask the under loaded processors to offers to take over some of their load.(7)Assignment with precedence constraints: The last task assignment algorithm takes task precedenceconstraints into account.

Q13.Explain Preemptive Earliest Deadline First? Ans.A Processor following the EDF algorithm always executes the ti whose absolute deadline is the earliest. EDFis a dynamic priorit scheduling algorithm; the task priorities are not fixed but depending on the closenessof their absolute deadline. EDF is also called the deadline monotonic scheduling algorithm.e.g. consider the following set of (aperiodic) task arrivals to a system.

When T1 arrives, it is the only task waiting to run and so starts executing immediately. T2 arrives at time4, since d2 <d1, it has higher priority than T1 and preempts it. T3 arrives at time 5, now ever since d3> d, it has lower priority than t2 finishes (at time t), T3 starts since it—has higher priority than Ti). T3 runs until 15 at which point T can resume and runs to completion.EDF is an optimal uniprocessor scheduling algorithm. EDF can not feasibly schedule a task set on auniprocessor, there is no other scheduling algorithm that can.If the total utilization of the task set is no greater than 1, the task set can be feasibly scheduled on a singleprocessor by the EDF algorithm.

Loremate.com

Like 6,065

http://goo.gl/EZv6H






http://pdfcrowd.com





Home

Chapter 4 : Programming Languages and Tools

Q1. Define Programming Language?

Ans.Programming Languages are the medium in which computer Programs are written. A Language shapes arenot only how we express an idea but also the very idea itself. A good programming languages enables theprogrammer to state, with a low probability of error, his directions to the computer. It also forces theprogrammer into the discipline of formatting his thoughts according to the computer language frame work.

Q2. Why Readability is Important?

Ans.It is important for two Reason:(1) It makes the debugging process easier.(2) It makes possible the maintenance of the_ program even if the original developers are no longeravailable.

Q3.Describe features of Real Time Programs and Language?

Ans.(1) It real time programs have to deal with a large variety at I/O devices. A real time system acquiretheir input from and deliver their output to many different kinds of sensors and actuators.

(2)A good Real time Language must give the programmer Flexibility as to the scheduling of processor andother resources.

(3) Programs must lend themselves to analysis for schedulability.

http://www.addtoany.com/share_save#url=http%3A%2F%2Fptucse.loremate.com%2Frts%2Fnode%2F6&title=Chapter%204%20%3A%20Programming%20Languages%20and%20Tools&description=
















http://pdfcrowd.com/redirect/?url=http%3a%2f%2fptucse.loremate.com%2frts%2fnode%2f6&id=ma-130126084912-4a871930

http://pdfcrowd.com


(4)A Real time Program, we must be able to analyze it to determine if it meets all the task deadline.

Q4.Define Data Typing?

Ans.Data types are used by programmers. The main purpose of Data typing is:(i)To guard against programming errors(ii)To specify the desire degree of numerical precision.

Q5. Define Record? Ans.A Record is a data type that allows the grouping of different data types and is especially useful in database.The dot notation can be used to access individual fields of a Record. e.g. student name denotes the namefield by the record student. New components may not be added.e.g. If we define the type student as a type derived from the type student, we cannot add anothercomponent, say social-security- number.

Q6. List the important features of ADA useful for Real time Programming?

Ans.In Real time Programming, with the help of ADA we can remove the limitation at Record i.e. If we can notadd another component in Record, with the help of ADA we can add the component i.e. this limitation isRemoved in ADAx by the introduction of the tagged data type that is we could define type student! asderived from student.

Q7. Define Blocks?

Ans.A block consists at two parts: a specification, which defines the variables used within the block and a bodywhich contains the statérnents to be executed. A block way be placed anywhere in a program that astatement is written. The ,main purpose of block is infuriation hiding.




http://pdfcrowd.com


Q8. Define Packages?

Ans.Packages take the concept at procedures and functions one step further. A package consist of a packagespecification, which is the interface to the outside world, in the package declaration, together with apackage body Package can be written, debugged and compiled independently of one another. Packagescan be placed ir a library and made accessible to everyone who has access to that library.

Q9.Write down the disadvantage of blocks. Ans.The main disadvantage at block is that they have to be explicitly repeated every time they are needed. Aprocedure or fu.iction can be written out once and called whenever it is needed. This allows economy andpermits a hierarchical decomposition of pro grams.

Q10.Explain Run-Time Error?

Ans.It is also know as Exception handling. There are two kinds at errors: compile time errors are detected bythe compiler, Run time errnrs escape the compiler and occur during execution of the software.

Q11.Describe Euclid? Ans.Euclid is an experimental language developed at the university at Toronto.

(1) It has the features at strong type checking and the ability to declare variables at specified memoryaddresser, which are necessary as security against errors and for handling I/O devices resp.

(2) It also has extensive used defined exception handling facilities.

(3) Euclid is that it is specifically designed to allow for reasonably accurate estimates of worst caseprogram run times. Such estimates make it possible to determine the schedulability of the Real time workload that is to find out if the Real time tasks of the work load can all meet their hard deadline.




http://pdfcrowd.com


load that is to find out if the Real time tasks of the work load can all meet their hard deadline.

These are features not unique to Euclid (i.e. similar to Ada).> In particular, dynamic data structures are not allowed.> The syntax at the process declaration in Euclid shows how much timing control the programmer is givenin this language.

Q 12.Explain the Run — Time support?

Ans.The key tools of the Run — time support are the compiler, the linker. the debugger and the kernel. (1) Compiler: The compiler translate the source code into machine language. A good compiler will havean excellent diagnostic capability, which is used to improve the efficiency of the debugging Process. Thecompiler should also list, on demand, the machine instructions of the complied unit this is useful in makingtiming estimator.

(2) Linker: It links together the parts. (Which may be separately compiled units at the program or libraryroutines) that makes up the program.

(3) Debugger:(i) It allows program execution to be suspended and then Resumed on the programmer’s command. Itmakesvariables value to the programmer.(ii) It allow the insertion at breakpoints into the program, when a program encounter breakpoint, itsuspends execution so that the programmer can inspect the variable values at that point.(iii) It allows the programmer to execute the program step by step.

(4) Kernel: The kernel is responsible for arranging the resources of the system. It includes routines forallocating and scheduling the tasks, managing the memory running the algorithm for inter processorcommunication, handling processor failures and carrying out input and output operations.

Q13. Explain the concept of Run — Time error?

Ans.There are two kinds of errors:(1) Compile time errors are detected by the compiler.(2) Run time errors escape the compiler and occur during execution of the software.




http://pdfcrowd.com


(2) Run time errors escape the compiler and occur during execution of the software.(3) It is also known as exception handling. Mechanisms for exception handling in general must have thefollowing rather obvious properties:(i) They must be simple to use without making the s/w difficult to read.(ii) They should have low run time overheads.(iii) The Programmer should have the freedom to specifythe recovery actions in response to various kinds of runtime error.(4) Ada handles run time errors by raising exceptions. An exception causes controlled to be trans ferred toan exception handler, which is a program for suitably handling chat exception.(5) Every exception condition has a unique name. There are five exception condition that are predefinedand programmer is ableto define additional conditions.(6) The Predefined conditions are: (i)Constraint — Error: Asserted whenever a variable value goes outside its prescribed bounds, if any (ii)Numeric — Error: Asserted to indicate an inability to maintain the required levels at precision (usuallydue to an attempt to divide by zero). (iii)Strong — Error: Asserted when the program rums out of memory space.. (iv)Program — Error: Asserted whenever an exception occurs that is not captured by any at the otherconditions.

(v)Tasking — Error: Asserted in connection within errors that arise due to the incurrence at use of thetasking mechanismThe Programmer has to explicitly state due response to an exception (7)The name of the exception is defined much as a variable type is declared. The programmer must thenexplicitly raise the exception where its warranted. What happens if an exception is raised in a block,procedure or function but there is no handling routine within that unit? In Ada, the normal execution a theunit ceares and exception keeps propagating until it finds an exceptional handler. If the exception occurs ina block, control is transferred to a point just after the end at the block, if it occurs in a procedure orfunction, control is transferred to a point just after where that procedure was called

Loremate.com

Like 6,065

http://goo.gl/EZv6H






http://pdfcrowd.com





Home

Chapter 5 : Real Time Database

Q1.Define Database.

Ans.Database are a structured and convenient way to manage the sharing ofi large quantities of data amongmultiple tasks. In many cases, interaction b/w tasks can be defined in terms of read and write operationsby those tasks on the database.

Q2.Give the applications of Real — Time database? Ans.Real Time database occurs in many important applications:(1) Airline Reservations(2) Banking(3) Stock Market Systems.In airline reservation and banking systems the response time needs to be kept low in order not to annoycustomer. In stock market systems, the slow execution of orders can be very expensive in volatilemarkets.

Q2.Define Transaction? Ans.Transaction is a sequence of read and write operations. If t is a set of transactions a history over t is aninterleaving of the reads and writes in the transactions in t. Transaction is said to be a query if it consistsonly of Read operations and up date otherwise.

http://www.addtoany.com/share_save#url=http%3A%2F%2Fptucse.loremate.com%2Frts%2Fnode%2F7&title=Chapter%205%20%3A%20Real%20Time%20Database&description=
















http://pdfcrowd.com/redirect/?url=http%3a%2f%2fptucse.loremate.com%2frts%2fnode%2f7&id=ma-130126084951-cd77e82d

http://pdfcrowd.com


Q3.Define Atomicity?

Ans.An action is said to be atomic if it is either done completely or not at all. Thus, if a transaction specifiesseveral steps the end Result must be as if either all or none at there steps are carried out.

Q4.Define Consistency?

Ans.The Transaction Transforms the database from one coi sistent state to another. A consistent state is onethat Results from the execution of some given sequence of transaction.

Q5.Define Isolation?

Ans.The actions of a transaction are not visible to any other transaction until and unless that transaction hascommitted.

Q6.Define Durability?

Ans.The actions of a transaction on a database (upon commitment) are permanent.

Q7.Explain Absolute Consistency.

Ans.Absolute Consistency is accuracy, that is, data about the operating environment must be consistent withthe environment. e.g. we interrogate a database to find the current temperature or pressure in a chemicalreactor vessel, we want to return data to be close to the current temperature or pressure.




http://pdfcrowd.com


reactor vessel, we want to return data to be close to the current temperature or pressure.

Q8.Define Relative consistency?

Ans.Relative Consistency means that, for multiple data, the data must have been collected reasonably close toone another.

Q9.Explain Database for hard Real Time system?

Ans.No conventional database exist that can meet the Requirements at hard Real time systems. All theReal time database of which we are aware have an average transaction time of more than 1 00 ms. This isan average time and many individual transaction suffer for longer response time.

Q10.Explain Need for Response — Time Predictability? Ans.There are many factors that make it difficult to predict the responsetime of transactions:(1) Transactions may be aborted one or more times before they are finally completed. Transaction abortsmay be needed to avoid deadlock.(2)Database are often much too large to entirely fit in the main memory and must therefore rely on discsystems.(3)Transaction access (read/writes) may be data dependent. e.g. if we have a transaction that is supposedto deduct a certain amount from a bank balance, it will have one set as accesses the balance is largeenough & another if it is not.(4) Transactions may suffer a delay because they are waiting access datum that is currently locked byanother transaction.

Q11. Explain Transaction Pñorftics?

Ans.




http://pdfcrowd.com


Ans. Transactions are granted access to the processors based on their priorities. These priorities are designedto ensure that some system measure of performance is optimized. This implicitly assume that alltransactions have equal value to the system. If this in true then some means must be found to incorporatethe value at the transaction in the priority assigned to it.

The Priority scheme is as follows:(1) Transaction is a higher value class always have priority over transaction in a lower value group.(2) If transaction A$ both belong to the HIT group of the same value class, the one with the earlierdeadline in given priority.(3) If A belongs to the ifiT and B to the MISS group of the same value class, A has priority over B.(4) If A and B belongs to the MISS group of the same value class. (a) The value of A is greater than value of B, then A has priority. (b) If they both have same value, the tie is broken on the basis at their relative positions in the list at pending transaction.

Q12. Explain Concurrency control issues in detail?

Ans.The Problem of Concurrency control is to allow transactions to execute in parallel while ensuring that thenet effect of the execution on database is as if they had been executed in some serial order. There are two basic approaches to concurrency control, pessimistic and optimistic.

(1) Pessimistic Concurrency Control: In the Pessimistic approach, we first ensure that the transactionwill not violate serialization consistency before letting it execute. The most commonly used approach is twophase locking. The Procedure Requires that the transaction Lock and unlock into two distinct phases. In thelocking phase, it acquire the Read and write locks, it needs. In the unlocking phase, it release the locks.The unlocking phase must follow the locking phase. Thus Transactions must haveobtained all the lock it willneedbefore it release any lock. This process has the potential to dedlock. Transactions are limited to Readingthe latest certified version of data. This constraint ensure that cascadingabort are avoided. (2) Optimistic Concurrency Control: In the optimistic approach, we first carry out the transactionexecution and then only check see if the execution at that transaction have violated serializationconsistency.

(3) Under the optimistic concurrency control, the execution at each transaction consist of there phases, aRead phase, validation phase and a write phase in that order. During Read phase, the transaction Readsthe data if needs and writes only into its own private address space. During validation phase, the systemcheck to See if any at the writes has potentially violated specialization consistency for every transaction A




http://pdfcrowd.com


check to See if any at the writes has potentially violated specialization consistency for every transaction Awhose timestamp antedates that at transaction T, serialization consistency is not violated due to Tis the following conditions holds.

(i) A has completed its write phase before T starts its Read phase.(ii) The read set of A is distinct from the write set of T & A has finished its write phase before T starts itswrite phase.(iii)The write set of A is distinct from both Read & Write set of T

If these conditions are not satisfied, T must be aborted. Since a transaction confines its write to its ownprivate space until it is validated such as abortion requires no cleaning up of the database itself and can bedone with little recovery overhead. Loremate.com

Like 6,065

http://goo.gl/EZv6H






http://pdfcrowd.com





Home

Chapter 6 : Real Time Communication

Q1. Define communication in Real Time system?

Ans. Effective communication b/w the various devices of a Real Time system is vital to its correctfunctioning. Hard Real systems must therefore use communication protocols that allow the communicationoverhead to the bounded. In soft Real Time systems, such as multimedia or video couferencing wherevoice and image data are being transmitted.

Q2. Explain the Reasons for Message delay? Ans.(1) Formatting and Packetizing the message.(2) Queuing the message at it wait for access to the communication medium.(3) Sending the message from the source to the destination.(4) Deformatting the messages:

Q3. Define Contant Rate?

Ans.Contant Rate: Fixed size packets are generated at periodic intervals. Many sensors produce such traffic.Contant Rate traffic is the earliest to handle since it is smooth and not bursty. The smoother the traffic, thesmaller the number of buffer that must be provided at each node.

http://www.addtoany.com/share_save#url=http%3A%2F%2Fptucse.loremate.com%2Frts%2Fnode%2F8&title=Chapter%206%20%3A%20Real%20Time%20Communication&description=
















http://pdfcrowd.com/redirect/?url=http%3a%2f%2fptucse.loremate.com%2frts%2fnode%2f8&id=ma-130126085049-050ab6cd

http://pdfcrowd.com


Q4.Define Variable Rate?

Ans. This may take the forms of fixed size packets being generated at regular intervals or of variable sizepackets being generated at regular intervals. Video packets are an example of variable sized packet beinggenerated at regular intervals.

Q5. Define Fault Tolerance?

Ans.This measures the extent to which the network can with stand the failure of the individual links and underwhile still remaining functional.

Q6.Define Token?

Ans. A token is a grant of permission to a node to transmit its packets on the network when the token holdingnode completes its transmission, it surrender the token to another node. A node is only permitted totransmit on the N/W if it currently holds the token.

Q7. Explain Packet switching?

Ans.The message is broken down into packet, which are messages at a standard or variable length. Packetshave headers, which specify their source, destination and any other information that may be required.They are then sent to their destination by the routing and flow control algorithm.

Q8. Define Circuit Switching?

Ans.A circuit is a set up bfw source and destination for such time as it required to send the message. The entirecircuit it then meant exclusively for this message, any other message that require all or part of this pathmust wait for the transmission to be completed.




http://pdfcrowd.com


must wait for the transmission to be completed.

Q9.Give advantages and Disadvantages of optical fiber?Ans.Advantages:(1) The raw bandwidth at a typical fiber can be high or several hundred GHz.(2) Optical signals are immune to the effect at electromagneticDisadvantage:(1) It is difficult to passively tap them without a significant signal loss.(2) Optical amplifier to restore signal level are expensive and so taps are impractical unless the system isvery small.

Q10.Explain Network topologies? Ans.Network topologies can be broadly classified into point to point and shared categories. In point to pointtopology nodes are connected by dedicated link. In shared topology, the holds all have access to thecommunication channel and only one node can transmit at any time over a channel.

Q.11. Explain Network Architecture Issues?

Ans.We have three types of Network Architecture Issues.(1) High Level Architecture Issues: At the highest level, a distributed system comprise a set of nodescommunicating through an interconnection network. Each made may itself be a multiprocessor comprisingapplication system and N/w processor, a shared memory segment and I/o interface. The nodes of thesystem nut be interconnected by a suitable N/w.

(2) Low Level Architectural Issues: Low Level Architectural Issues involves Packet Processing, routingand error/flow control. In the distributed real time system, there are additional issues related to thesupport for meeting deadlines, time management and housekeeping. Since the support at these low levelissues impedes the execution at application task nodes in a distributed real time system usually have acustom-designed processor for handling there chore.

(3) I/o Architecture: A real time computer can process data no faster than it acquire it from senurs &operators. The nature of Ito devices in a real time environment, consisting at it does of sensors andactuator, in quite different from the magnetic disk and taps that are the I/o devices commonlyencountered in the general purpose system.




http://pdfcrowd.com


encountered in the general purpose system.

Q12. Explain Network Topologies in detail?

Ans.The network topology for a computer or a distributed system must be carefully chosen since it affect thesystem response time and reliability. The following feature are important.

(1) Diameter: This is the maximum distance b/w any two nodes in the system, as a function of thenumber of nodes. Ideally, the diameter should increase only slowly as a function at the number of nodes.In a completely connected N/w, where each node has a dedicated link to every other node, the diameter isone, regardless of the number of nodes. (2) Node Degree: This is the number of edges adjacent to the each node and determines the number atI/o ports per node and the number of links in the system. The greater this number, the greater the cost.

(3) Fault Tolerance: This measures the extent to which the NIw can withstand the failure of individuallinks and nodes while still remaining functional.

(4) N/w topologies can be broadly classified into point to point and shared categories. In point to pointtopology, nodes are connected by dedicated links. In shared topology, the nodes all have access to thecommunication channel and only one node can transmit at any time over a channel.

Q13.Explain Protocols? Ans.Protocols: In which we describe Protocols suitable for real time system. We have different types ofProtocols.

(1) Contention Based Protocol: These are distributed Protocols that assume a broadcast medium.Nodes monitor the channel and transmit only when they detect that it is idle. If multiple nodes starttransmitting at about the same time, there is a collision of packets and the transmissions have to beaborted and thenretried.

(2) Virtual-Time Carrier-sensed Multiple Access (VTCSMA): The VTC SMA Protocol has beendesigned for single channel broadcast networks and for the bus and ring topologies. In the Carrier —sensed Multiple Access (CSNA) protocols, all the nodes can monitor the communication channel. CSMA istruly distributed algorithm with each node deciding when it will transmit.




http://pdfcrowd.com


truly distributed algorithm with each node deciding when it will transmit.

Performance of the VTCSMA Algorithms:

(i) VTCMSA Protocols have a better loss rate, in general than the CSMA Protocols.

(ii)The best performance in terms of fraction of loss rate it obtained by the VTCSMA-D Algorithm. (iii) The Performance is a function of the value of ,ñ however as long of the transmission delays are nottoo great and the network is not overloaded.

(3)Window Protocol: Like the VTCSMA Protocol, the window protocol is based on collision sensing. Againit can not be guaranteed that messages will be transmitted in time to meet their deadlines. This protocol istherefore only suitable for softreal time system. The protocol ones its name to the window maintained at each node.(4) Timed Token Protocol: The timed token protocol is a simple mechanism by which each node itguaranteed timely access to the network. It distinguishes b/w two basic classes of traffic synchronous andasynchronous.(i) Synchronous traffic is the Real time traffic, the protocol guarantees that each node can send out up to acertainamount at synchronous traffic every time unit.(ii) Asynchronous traffic is non real time traffic that takes up any bandwidth left unused by the synchronoustraffic.(5) Stop and go multihop Protocol:(i) The stop and go protocol is another technique to meet hard deadline on packet delivery times.(ii) The stop and go protocol is a distributed algorithm each node works independently without centralcontrol. (6) The Polled Bus Protocol: The poled bus protocol assumes a bus N/W within a bus-busy line. When aprocessor broadcast on the bus, it also maintain this line high. When it finishes broadcasting, this line itreset. (7) Hierarchical Round Robin Protocol: This protocol guarantees that each traffic class i can transmitup to m1 packets every t1 time units for prespecifed m1 and. T1. As with the stop and go protocol, we canbound the delay encountered by a packet at each intermediate node. (8) Deadline Based Protocols: A deadline based protocol on a point to point Network consist of eachnode transmitting the packet with the earliest deadline. There are two iariations onthis, preemptive, and non-preemptive. Loremate.com

Like 6,065

http://goo.gl/EZv6H






http://pdfcrowd.com





Home

Chapter 7 : Fault - Tolerance Techniques

Q1.Define Hardware Fault?

Ans.Hardware Fault is some physical defect that can cause a component to malfunction. A broken wire orthe output of a logic gate that is perpetually stuck at some logic value (0 or 1) are hardware faults.

Q2.Define Software Fault?

Ans.A software Fault is a “bug” that can cause the program to fail for a given set at inputs. An error in amanifestation of a fault.

Q3.What is Fault Latency?

Ans.The Fault Latency is the duration b/w the onset of a fault and its manifestation as an error. Thisduration can be considerable. Since the faults themselves are invisible to the outside world, onlyshowing themselves when they cause error, such latency can as we shall see, impact the reliability atthe overall system.

Q4.Define error latency?

http://www.addtoany.com/share_save#url=http%3A%2F%2Fptucse.loremate.com%2Frts%2Fnode%2F9&title=Chapter%207%20%3A%20Fault%20-%20Tolerance%20Techniques&description=
















http://pdfcrowd.com/redirect/?url=http%3a%2f%2fptucse.loremate.com%2frts%2fnode%2f9&id=ma-130126085119-54e9d575

http://pdfcrowd.com


Q4.Define error latency? Ans.The error Latency is the duration between when an error is produced and when it is eitherrecognized at an error or cause the failure of the system like fault latencies, error latencies can have aconsiderable impact on the overall system reliability.

Q5.Explain Error Recovery?

Ans. Error Recovery is the process by which the system attempts to recover from the effects of an error.There are two forms of error Recovery forward and backward. In forward error recovery, the error inmarked- without any computations having to be redone. In backward recovery, the system is rolledback to a moment in time before the error is believed to have occurred and computation is carried outagain.

Q6.Define ADUlT TRALIS?

Ans. Audit Trails: A second way of enabling backward error recovery is through audit trails. There areespecially popular in database. An audit trait consists of a record at all the actions that have been takenby the system, together with a time stamp indicating when each action was taken.

Q7.What is Data Diversity?

Ans. Data diversity is an approach that can be used in association with any of the redundancy techniques,considered above. Sometimes, hardware or software may fail for certain inputs but not for other inputsthat are very close to them. Instead of applying the same input data to the Redundant processor, weapply slightly different input data to there.

Q8.Explain Fault and error containment?

Ans. When a Fault or error occurs in one part of the system, it can, if unchecked, spread through the systemlike an infection disease. A fault in one part of the system. Fault and error must therefore be prevented




http://pdfcrowd.com


like an infection disease. A fault in one part of the system. Fault and error must therefore be preventedfrom spreading through the system. This is called containment.

Q9.Define Reversal Checks?

Ans. If there is a simple Relationship b/w the inputs and outputs of a system, it may then be possible tocalculate the inputs given the outputs. This can then be compared within the actual inputs as a check.Reversal is a powerful method, but its applicability is limited to tasks wherdit is relatively easy to carryout the Reverse computation.

Q10.Define Software Redundancy?

Ans. The system is provided with different slw various of tasks, preferablywritten independently by different teams at programmer. so that one version of a task fail undercertain inputs, another version can be used.

Q11. Explain Fault Detection?

Ans.There are two ways to determine that a processor in malfunctioning, online and offline.Online detection goes on in parallel with no ma! system operation. One way of doing this is to check foraz y behaviour that is inconsistent with correct operation. The folk wing action are indicative at a faultyprocessor: (1) Branching to an invalid destination.

(2) Fetching an opcode from a location containing data.

(3) Writing into a portion of Memory to which the process has no write access.

(4) Fetching an illegal opcode.

(5) Inactive for more than prescribed period.




http://pdfcrowd.com


A monitor (called a watchdog Processor) is associated with each processor, l6oking for sign that theprocessor is faulty. Offline ôetection consist o running âiagnostic test. ‘WXien a ?rocessor is runningsuch a test it obviously can not be executing the application software. Diagnostic tests can be scheduledjust like ordinary tasks. The greater the failure rate, the greater must be the frequency with whichthese tests are run.

Q12. Explain Fault Types? Ans.Faults are classified according to their temposal Behaviour and output behaviour. A fault is said to beactive when it is physically capable of generating errors and to be begin when it is not.

Temporal Behaviour classification:There are three fault types:

Permanent, intermittent and transient. A permanent fault does not die away with time, but remains until it is repaired or the affected unit isreplaced. An intermittent fault cycles b/w the fault active and fault benign states. A transient fault diesaway after sometime.

Fault type ConditionPermanent a(t) > o. b(t) a(t) = oTransient a(t)> o, b(t) o, c(t) > o, d(t) oIntermittent a(t) > o, b(t)> o, c(t) = o, d(t) > o. In fig, a(t) & b(t) are the rates at which the fault switches states, and t is the age of fault. Intermittent faults can be caused by loose wires. Transient faults can be caused by environmentaleffects. When the memory is rewritten, tl.e fault will go away transient failure are hard to catch, sincequite often by the time the system has recognized that such a failure has occurred, it has disappearedleaving behind no permanent defect that can be located output behaviours classification.

A fault is also characterized by the nature of errors that it generates. There are two categories ofoutput behaviour, non malicious and malicious, to understand the difference, consider a unit A which isproviding output to units B1 Bq If A has a non malicious fault, any errors it generates will be interpreted




http://pdfcrowd.com


providing output to units B1 Bq If A has a non malicious fault, any errors it generates will be interpretedin a consistent way by all the units B1.For e.g., if A has an output line that is stuck at logic 0, this is a non-malicious fault because all the unitsthat receive input from this line will read the line as producing a logic 0.

A malicious fault is one that is assumed to behave arbitrarily.

In this context, the terms fail-safe and fail stop can be defined. A unit is said to be fail-stop if itresponds to up a certain maximum number of failures by simply stopping, rather than putting outincorrect output.A system is said to be fail-safe if it failure node is biased so that the application processes does notsuffer catastrophe upon failure.

Q13.Explain Redundancy in detail?

Ans. Redundancy: Fault tolerance consists of using and properly managing redundancy. In other words, ifthe system is to be kept running despite the failure of some of its parts, it must have spore capacity tobegin with.There are four types of redundancy:

Hardware redundancy: The system is provided with for more hardware than it would need if all thecomponents were perfectly reliable, typically between the and three times as much.

Software redundancy: The system is provided with different software versions of tasks, preferablywritten independently by different teams of programmer.

Time Redundancy: The task schedule has some slack in it, so that some tasks can be rerun ifnecessary & still meet critical deadlines.

Information redundancy: The data are coded in such a way that a certain number of bit errors canbedetected andJor corrected.

Hardware Redundancy: Hardware Redundancy is the use of additional hardware to compensate for




http://pdfcrowd.com


Hardware Redundancy: Hardware Redundancy is the use of additional hardware to compensate forfailures, it can be used in two ways. The first is its for fault of detection, correction and masking. This isfault detection.

The second and long term use of Hardware redundancy is to replace the malfunctioning units. It ispossible for systems to be designed so that spares can be switched in to replace any faulty units.Redundancy is expensive. Duplicating or triplicating hardware is a luxury that is only justified in themost critical applications. Traditionally, such use has been largely confined to aerospace applications. Inaerospace and other applications, redundancy is more often limited by how much power consumption,heat dissipation, or volume can be accommodated.

Modular Redundancy: N-modular Redundancy is the scheme for forward error recovery it works byusing N process instead of one and voting on their output M is usually odd. To substain upto M fail units,the NMR system requires (2M+l) units in all. The most popular is the triplex which consists of total of 3units and can mask the effect of upto one failure.

Software Redundancy: Software faults are quite unlike hardware faults however software neverwears out. So the fault one never spontaneously generated during the system operation. To providereliability in the phase of software faults, we must use Redundancy however simply replicating thesame software N times will not work such diversity can be introduced by having teams ofprogrammers. A no. of software versions will depend on extend of common mode failures.

http://goo.gl/EZv6H




http://pdfcrowd.com


Loremate.com

Like 6,065






http://pdfcrowd.com





Home

Chapter 8 : Reliability Evaluation Techniques

Q1. Define Voter Reliability? Ans.There are two typical designs for voter reliability, one in which there in exactly one voter providing outputfor the cluster and the second in which there are N voters, one per processor we focus on the first designand leave the second as an exercise.

Q2. What is Fault Latency?

Ans.The Fault Latency time is zero. This is usually not the case. It is not until the fault is exercised that theFault generates an errors.Example: If a wire is broken, it is not some signal is meant to propagate down the wire that the Faultmanifests itself. Faults that have not yet manifested themselves are called latent faults.

Q3. Define Software Emr Model OR s/w Reliability?

Ans.Software - Emr Models pedict the rate at which software faults will produce errors and are meant todetermine when to stop debugging, by providing some guidance on the Reliability of the software at eachstage of debugging.The usefulness at software error Models in ultra reliable computer system is extremelylimited.

http://www.addtoany.com/share_save#url=http%3A%2F%2Fptucse.loremate.com%2Frts%2Fnode%2F10&title=Chapter%208%20%3A%20Reliability%20Evaluation%20Techniques&description=
















http://pdfcrowd.com/redirect/?url=http%3a%2f%2fptucse.loremate.com%2frts%2fnode%2f10&id=ma-130126085151-587cd4b5

http://pdfcrowd.com


Q4. Define Jelinski-Miranda Model? Aus.This the earliest and perhaps the most widely used model. It models the error generation rate asproportional to the number of faults in the software. That is, λ(i) = ji where λ(i) is the failure rate ofsoftware with i faults and J is the unknown parameters.

Q5. Explain Goel-Okumoto Model?

Ans.This model amusers that the occurrence of error is a non homogenous Poission process (NHPP) with rate λ(t) = abe-4t, where a and bore the unknown parameter. This captures the idea that some faults are easyto detect and others difficult. If no error has occurred for a long time, it is reasonable to assume that anyfaults that exist in the s/w are those that are exercised very infrequently.

Q6.Define Little wood Model? Ans. This model assumes that each fault i can generate errors according to a Poisson process with rate Φi Sincethe Φi can vary, the model allows us the flexibility at assigning different error generation rates to thevarious faults.

Q7.Explain little wood-verall model?

Ans. This model assumes that the time biw failures has the density function:

Where there are i faults in the s/w. the unknown parameter λi, is itselfnot a constant, but a random variable with the gamma density.

The function Ψ(i) determines how the rate changes with the number of faults i. It is possible to show that if Ψ(i) is an increasing function of i,λi , is stochastically greater than λf for i >j that is Prob{λf>x}>Prob{λj>x}>x>o




http://pdfcrowd.com


Prob{λf>x}>Prob{λj>x}>x>o The values of λi and ∞ are chosen by the user based on the observed data.

Q8. Explain series-Parallel systems? Ans. A set of components is connected in series, if the failure of any them will reset in system failure. A parallelconnection of components requires all the components to fail before the system fails.

Q9. Define combinatorial model?

Ans.The system will fail only if there are fewer than two functional processors left in the system. There is norepair and all the failure are assumed to be permanent, the failure probability can be found by counting allthe various ways in which fewer than two processors are left and weighting each by its probability ofoccurrence.

Q10. Define Markov chain Model?

Ans.Markov chain models, while more complex than combinational model toe such simple cases, are the




http://pdfcrowd.com


Markov chain models, while more complex than combinational model toe such simple cases, are thesolution method of choice when the systems are moral complex. The system can be modeled as a Markovchain, where the states represent the number of functional procures.

Q11. Explain Parameter Values?

Ans.The first step in developing a model is to decide what the input parameters should be: A model always bebased on parameter that can either be accurately measured estimated or attitude with confidence. (I) Obtaining — Device Failure Rates: There are two ways to obtain device failure rates, collecting fielddata and life cycle testing in the laboratory. The former is more realistic since it represents the failure ratewhen the device are being used in their normal operating conditions. The latter is the only choice when thedevices are new and field data do not exist.

(2) Measuring Error-Propagation Time: To measure now quickly an error can propagation, we usefault injection. This is best done on a prototype, special purpose hardware is used to simulate a fault on aselected line. The status of the related line is monitored using logic analyzers to determine now far andhow quickly the error propagates. If the prototype is not available, a s/w simulation can be substituted. Due to variation in execution and message propagation times, the error propagation time is a randomvariable and must be characterized by a probability distribution function (PDF).

Q12.Explain software-Error Models in detail?

Ans. Software error models predict the Rate at which siw faults will produce errors and are meant to determinewhen to stop debugging, by providing some guidance on the Reliability at the siw at each stage atdebugging.(1) Jelimki — Moranda model: (Repeated in Question 4)(2) Goel-Okunioto model (Repeated in Question 5)(3) Little wood model (Repeated in Question 6)(4) Little wood-verall model (Repeated in Question 7)




http://pdfcrowd.com


Q13.Explain Reliability models for Hardware Redundancy? Ans.The most difficult problem in reliability modeling it to keep the complexity at the models sufficiently small.In order to model the reliability of a system, we must express the reliability of each at its component andtake into account the impact of the failure of each component of the functioning of the overall system.Components1.Permanent faults only:(a) Series — Parallel System: A set of components is connected in series if the failure of any of themwill result in system failure. A parallel connection of components requires all the components to fail beforethe system fails.

2.Combinatorial model: The system will fail only if there are fewer than two functional processors left in thesystem. Since there in no repair and all the failures are a sumed to be permanent, the failure probabilitycan be found by counting all the various ways in which fewer than two processors are left and weightingeach by its probability of occurrence. 3.Markov chain model: Markov chain model while more complex than combinatorial model for such simpie caser, are the solution method of choice when the system ar complex. 4.Voter Reliability: (Repeated Question 1)5.Fault Latency: (Repeated Question 2)

Loremate.com

Like 6,065

http://goo.gl/EZv6H






http://pdfcrowd.com





Home

Chapter 9 : Clock Synchronization

Q1.Define clock synchronization? Ans.Clock synchronization in a vital to the correct operation of Real time system such activities are voting adsynchronized Rollback that the clocks are synchronized fault tightly. There ensures that the functionalprocessors remains synchronized despite a few processor or link failure.

Q2. Explain clock?

Ans.Mathematically speaking, clock c is a mapping C1: real time -+ clock timeThat is, at real time t, c1(t) in the time told by clock c1. The inverse Function c1(T) is the Real time atwhich the clock c1 tells time T. where it is important to distinguish b/w real and clock times we will use theprefixes ‘r’ and ‘c’ Respectively.

Q3. Define synchronization?

Ans.Two clocks are said to be synchronized if the times the tell are sufficiently close. More precisely, clocks C1and C, are synchronized at c time T if for some given the skew b/w clocks c, and c3at c time T.

http://www.addtoany.com/share_save#url=http%3A%2F%2Fptucse.loremate.com%2Frts%2Fnode%2F11&title=Chapter%209%20%3A%20Clock%20Synchronization&description=
















http://pdfcrowd.com/redirect/?url=http%3a%2f%2fptucse.loremate.com%2frts%2fnode%2f11&id=ma-130126085219-561edfa4

http://pdfcrowd.com


Q4. Explain Impact of Faults? Ans.When faults occurs, the simple algorithm will not work. Let us begin by assuming that all signal-propagation times are zero. e.g. clock c2 undergoes malicious failure and delivers its ticks to the clocks cand c3 at different timer. To clock e1,c2 clock c3,c2 appears to be slower than c3. As a Result both clocksc1 and c3 will work think of themselves as in the middle, the reference clock. They will not correctthemselves any way and be free to drift apart indefinitely.

Q.5.What is Loss at synchrony? Ans.Synchronization is carried out by clocks e.xchanging timing ménages and adjusting themselvesappropriately. There are two ways insynchrony can be lost as a result of some clocks becoming faulty, when multiple non overlapping cliquesare formed and the clocks are driven too fast or two slow.

Q6.. Define Hardware synchronization?

Ans.Hardware synchronization Requiresspecial hardware, which is notrequired by the software synchronization however if offers muchtighter synchronization.

Q.7.Define Software Synchronization?

AnsWhen the extremely tight synchronization that is provided by phase locking is not needed, synchronizationcan be caused out in software. In s/w based synchronization, we have an underlying hardware clock ands/w based correction. The clock time is the such at the hlw time and the correction.

Q.8.. Give the advantages at Hardware synchronization? Ans.(1) The main advantages at hardware synchronization is very small clock skews that can be attained.-These skews are on the order of nanosecond. This is due to the frequency with which theResynchronization are carried out.




http://pdfcrowd.com


Resynchronization are carried out.(2) Another Advantage is that hardware synchronization places no burden on the rest of the system.

Q.9.Give the Disadvantages at Hardware synchronization? AnsThe main disadvantage is the cost of the hardware. For application where very tight synchronization is notneeded, this approach may be to expensive and software synchronization may be the preferred approach.

Q10. What is fault-Tolerant synchronization in hardware? Ans.To synchronize in hardware, we can use phase-locked loops. The basic structure of phase-locked loop isshown in fig. The objective is to align, at closely at possible the output at the oscillator with an oscillatorysignal input.

Q.11.Give the advantages and Disadvantages of Hardware synchronization? Ans.Repeated Question 8,9.Explain Clocks in detail?Repeated in Question 1,2,3.

Q.12.Explain Fault-Tolerant Synchronization in hardware in detail? Ans.To synchronize in hardware, we can use phase locked loops. The basic structure of phase locked loop asshown in fig.

Completely connected, Zero-Propagation-Time system: Fig:




http://pdfcrowd.com


Completely connected, Zero-Propagation-Time system: Fig:shows the structures of each of the clocks. Every clock is connected by a dedicated line to every otherclock. We assume to being with that signal-propagation times are zero. Each clock has a reference circuit,which accepts as input the clock ticks from other clocks in the system an well as that of its own vco. Itgenerator a Reference signal to which its vco tries to align itself.(1)

(2)Spare-Interconnection Zero-Propagation Time system:Suppose initead at a completely connected structure, we have clocks organized into multiple clusters. Eachclock in a cluster is connected by a dedicated link to every other clock in that cluster. (3)Accounting for signal-Propagation Define: We have assumed that the signal propagation times arenegligible. This assumption is true if the geographical extent at the system is not large. (4)Multiple fault classes: We have to sustain up to us malicious faults. (5) Advantages and Disadvantages of Hardware synchronization: Repeated Question 1

Loremate.com

Like 6,065

http://goo.gl/EZv6H






http://pdfcrowd.com

RTS Ptu Sample paper lords mbd

Documents

browser pro

press ctrl

markov chain

combinational

units b1 bq

markov chain

stock market

interrupt