|
Computer System Engineering 6.033 Spring 2006
Author:
Prof.Samuel MaddenProfessor
MIT Massachusetts Institute of Technology
USA
Access: |
2.1 Software development process Read Online
2.2 Requirements analysis Read Online
2.3 Software design Read Online
2.4 Software construction Read Online
2.5 Software testing Read Online
2.6 Software maintenance Read Online
2.7 Software configuration management Read Online
Virtually all countries now depend on complex computer-based systems. More and more products incorporate computers and controlling software in some form. The software in these systems represents a large and increasing proportion of the total system costs. Therefore, producing software in a cost-effective way is essential for the functioning of national and international economies.
Software engineering is an engineering discipline whose goal is the cost-effective development of software systems. Software is abstract and intangible. It is not constrained by materials, governed by physical laws or by manufacturing processes. In some ways, this simplifies software engineering as there are no physical limitations on the potential of software. In other ways, however, this lack of natural constraints means that software can easily become extremely complex and hence very difficult to understand.
Software engineering is still a relatively young discipline. The notion of ‘software engineering’ was first proposed in 1968 at a conference held to discuss what was then called the ‘software crisis’. This software crisis resulted directly from the introduction of powerful, third generation computer hardware. Their power made hitherto unrealisable computer applications a feasible proposition. The resulting software was orders of magnitude larger and more complex than previous software systems.
Early experience in building these systems showed that an informal approach to software development was not good enough. Major projects were sometimes years late. They cost much more than originally predicted, were unreliable, difficult to maintain and performed poorly. Software development was in crisis. Hardware costs were tumbling whilst software costs were rising rapidly. New techniques and methods were needed to control the complexity inherent in large software systems.
These techniques have become part of software engineering and are now widely although not universally used. However, there are still problems in producing complex software which meets user expectations, is delivered on time and to budget. Many software projects still have problems and this has led to some commentators (Pressman, 1997) suggesting that software engineering is in a state of chronic affliction.
As our ability to produce software has increased so too has the complexity of the software systems required. New technologies resulting from the convergence of computers and communication systems place new demands on software engineers. For this reason and because many companies do not apply software engineering techniques effectively, we still have problems. Things are not as bad as the doomsayers suggest but there is clearly room for improvement.
This course covers topics on the engineering of computer software and hardware systems: techniques for controlling complexity; strong modularity using client-server design, virtual memory, and threads; networks; atomicity and coordination of parallel activities; recovery and reliability; privacy, security, and encryption; and impact of computer systems on society. Case studies of working systems and readings from the current literature provide comparisons and contrasts. Two design projects are required, and students engage in extensive written communication exercises.
There are 12 questions and 12 pages in this quiz booklet. Answer each question according to the instructions given.
You have 50 minutes to answer the questions.
Most questions are multiple-choice questions. Next to each choice, circle the word True or False,
as appropriate. A correct choice will earn positive points, a wrong choice may earn negative
points, and not picking a choice will score 0.
Question: 1. [6 points]: Which of the following statements are true of the X Windows System (as described in the X Windows paper, reading # 5)?
Choices:
True
False
Question: 3. [8 points]: Ben Bitdiddle is using a MapReduce cluster (as described in reading #8) to process some data. He finds that some of his programs are running slowly and needs your help to understand which performance enhancements to try. For each of the following MapReduce configurations and suggested enhancements, circle True if the statement about the enhancement is true, and False otherwise.
Choices:
True
False
Question: 1. [6 points]: Which of the following statements are true of the X Windows System (as described in the X Windows paper, reading # 5)?
Choices:
True
False
Question: 3. [8 points]: Ben Bitdiddle is using a MapReduce cluster (as described in reading #8) to process some data. He finds that some of his programs are running slowly and needs your help to understand which performance enhancements to try. For each of the following MapReduce configurations and suggested enhancements, circle True if the statement about the enhancement is true, and False otherwise.
Choices:
True
False
Question: 2. [8 points]: The UNIX designers chose to have the process created by a fork() share (inherit) the open file descriptors of its parent. Which of the following would become more complicated had they decided on an alternate fork() semantics, in which the child process is created with no open file descriptors?
Choices:
True
False
Question: 3. [8 points]: Ben Bitdiddle is using a MapReduce cluster (as described in reading #8) to process some data. He finds that some of his programs are running slowly and needs your help to understand which performance enhancements to try. For each of the following MapReduce configurations and suggested enhancements, circle True if the statement about the enhancement is true, and False otherwise. the entire 1000-node cluster can complete these MapReduce operations in 100 seconds then doubling the CPU performance (task processing rate) of each machine in the cluster may improve performance somewhat but not significantly.
Choices:
True
False
Question: 2. [8 points]: The UNIX designers chose to have the process created by a fork() share (inherit) the open file descriptors of its parent. Which of the following would become more complicated had they decided on an alternate fork() semantics, in which the child process is created with no open file descriptors?
Choices:
True
False
Question: 1. [6 points]: Which of the following statements are true of the X Windows System (as described in the X Windows paper, reading # 5)?
Choices:
True
False
Question: 3. [8 points]: Ben Bitdiddle is using a MapReduce cluster (as described in reading #8) to process some data. He finds that some of his programs are running slowly and needs your help to understand which performance enhancements to try. For each of the following MapReduce configurations and suggested enhancements, circle True if the statement about the enhancement is true, and False otherwise.
Choices:
True
False
Question: 2. [8 points]: The UNIX designers chose to have the process created by a fork() share (inherit) the open file descriptors of its parent. Which of the following would become more complicated had they decided on an alternate fork() semantics, in which the child process is created with no open file descriptors?
Choices:
True
False
Question: 2. [8 points]: The UNIX designers chose to have the process created by a fork() share (inherit) the open file descriptors of its parent. Which of the following would become more complicated had they decided on an alternate fork() semantics, in which the child process is created with no open file descriptors?
Choices:
True
False