Skip to content

Latest commit

 

History

History
53 lines (33 loc) · 1.87 KB

README.md

File metadata and controls

53 lines (33 loc) · 1.87 KB

OptimizeMePlease

You are probably here because you saw my post on Linkedin.

Welcome!

Steps

  • Before you run an application, you will need to create a database (I'm using MSSQL) named "OptimizeMePlease"
  • Go to Program.cs class
  • "IWillPopulateDate()" is a method which will get a script from the project directory and run in on created DB
  • Run application in Debug/Release mode
  • Comment or delete IWillPopulateData() call from Main method
  • Go to BenchmarkService.cs class
  • Start coding within GetAuthors_Optimized method

How do I submit my solution?

  • Send me a Linkedin message with your github username, I will add you and create a branch for you.

OR

  • Clone the project, create a branch and work on that branch.

OR

  • If you don't want to bother with github, after you're done with the changes, send me the results on Linkedin.

Rules

  • Only Entity Framework (Core) is allowed for using
  • The data obtained in the non-optimized version of the code must also be obtained in the optimized version
  • If you see potential optimization of something else, you can do it
  • Entities and DbContext cannot be changed (you got legacy code no matter what it might be bad :) )
  • The models returned from the method can be changed

What should the method return?

  • Given that there is a predefined database of data, the method should in any case return the list of data currently returned by the non-optimized method.

How will performance success be measured?

  • Given that each of us works on a computer with different performance and power, the execution time quotient of the non-optimized and optimized method will be checked for each separately.

Example:

Non-optimized method Execution time: 1.1s = 1100ms Optimized method Execution time: 200ms

Result: 1100ms/200ms = 5.5x faster.

In addition, I will check the performance on my computer for individual results.

GOOD LUCK!