Parallel Processes in Python

Base Knowledge

Students should be able to:

• Summarize what a process is and how a running program can involve many processes.
  
• Describe the challenge of shared resources in parallel computation.
  
• Describe the purpose of interprocess communication.
  

Conceptual/Applied Knowledge

Students should be able to:

• Write embarrassingly parallel programs in Python using the multiprocessing module.
  
• Manage access to a shared resource via a lock.
  
• Write programs involving simple interprocess communication via a shared queue.
  

This module is designed for use in the latter half of a semester-long CS1 course. It is expected that the module would be useful in a course with any number of students, though it has been tested in courses of 20 - 30 students. At a minimum, students should have introductory-level experience in printing, variables, tuples, and writing and using functions with keyword arguments. Some basic understanding of object-oriented programming is also very helpful: classes, methods, object state, constructors, and method invocation. For example, it would be quite sufficient for students to be able to write a "Student" class with major and GPA fields and supporting accessors and mutators.

You can visit the module in your browser and use it in your course:

Web version: Parallel Processes in Python

or you can download the module in either PDF format or latex format.

PDF Format: ParallelProcessesinPython.pdf.
Latex Format: latex.tar.gz.
Word Format: ParallelProcessesinPython.docx.

While the main module document is written such that it can be used as an independent tutorial, instructors may wish to present this material using a combination of in-class lecture, in-class (or lab) exercises, and out-of-class homework. In this case, the document can serve as a guide for topic ordering and scope.

Regarding prerequisites, students should have at a minimum introductory-level experience in printing, variables, tuples, and writing and using functions with keyword arguments. Some basic understanding of object-oriented programming is also helpful: classes, methods, object state, constructors, and method invocation. For example, it would be quite sufficient for students to be able to write a "Student" class with major and GPA fields and supporting accessors and mutators.

The narrative of the main module document includes several specific exercises. Typically these would be done as in-class exercises interspersed with lecture, but they could also be used as homework. These exercises include:

• Making Multiple Processes
  
• Anonymous Processes
  
• Digging Holes (multiple processes cooperating on a task)
  
• Extended Communication Via a Queue
  
• Obtaining a Result from a Single Child
  

The following exercises and solutions are available from the author:

• Short-answer questions
  
• Adding Numbers programming project
  
• Restaurant Simulation programming project
  
• Caesar Cipher programming project
  
• Parallel Pipeline Computation programming project
  

Applicable assessment instruments are not yet available. The module has been class-tested several times.