Skip to content

mendezcode/master-workers-simulation

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Master and Workers Simulation

This project simulates a regular working shift having master processes that act as supervisors for workers. Each worker is spawned and expires after a random amount of milliseconds.

After all the workers of a given master process have completed, the master process then has completed its job. he supervisor then spawns another master process, which in turn spawns more worker processes until they are done.

All of this is done asynchronously, and it provides some good insight of how erlang processes work. Setting moderate values for master and worker processes provides interesting results.

Learning Topics

This sample project has been created for learning purposes, and has a few interesting things that can serve as a small introduction to new users to erlang:

  • Process monitoring
  • Process links
  • Message Passing
  • Using receive/after
  • Module Macros
  • Records used for state
  • Module type and spec declarations
  • Edoc tags and comments
  • Creating documentation using edoc
  • Overview page for edoc
  • Random number generation
  • Sample escript script
  • Using Emakefile

Usage

Edit src/system.erl and adjust the MASTERS and WORKERS macro values to define how many master processes to spawn, and how many workers each master process spawns.

To define the number of milliseconds each worker runs, adjust the TIMEOUT macro value.

Build:

make

Build documentation:

make docs

Run:

make run

Cleanup:

make clean

Random number generation

As a side note, there is also a pseudo random number generator built using processes, which uses the erlang random module.

Due to referential transparency, it's a bit tough to create random numbers in erlang, in the code is a simple yet effective approach to random numbers using a process that keeps the internal random state to generate random numbers.

Originally, random:uniform/0 uses the process dictionary, which means that if the function is called multiple times on the same process, it will inherently create the same random number, hence the process approach implemented herein, which creates different numbers regardless from the process on wich they are requested.

License

MIT

About

Simulating master and worker processes in erlang

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages