Dockerize your unit tests for faster feedback

Michelangelo van Dam (04.Feb.2017 at 16:00, 1 hr )
Talk at SunshinePHP 2017 (English - US)

Rating: 5 of 5

Ever got the message "Fatal error: Out of memory" when running your unit tests? Or do you have to wait for more than 5 minutes for all of your unit tests to complete? We're using Docker to mitigate these and other issues. In this session I will show you how we have split up a single testsuite of 3K+ tests into manageable unit modules which are executed by Docker micro instances in parallel. And with this set up we test our application on the current PHP version, but it also allows us to test against a new major PHP version the moment it is released or even is available in beta.

Who are you?

Claim talk

Talk claims have been moved to the new Joind.in site.

Please login to the new site to claim your talk

 
Comments closed.

Comments

Rating: 5 of 5

04.Feb.2017 at 17:28 by Eric Morris (24 comments) via Web2 LIVE

This talk illustrates a little "thinking outside the box". Mr. van Dam uses Docker in a clever manner on a micro scale to solve his (and others') macro problem of increasing Unit Test efficiency. This technique is worth a look, no just for Unit Testing, but possibly other similar scenarios that can benefit from parallelliztion.

Rating: 5 of 5

04.Feb.2017 at 18:41 by Dana Luther (50 comments) via Web2 LIVE

Short, sweet and to the point! It's nice to have such a clear example to go back with and implement.

Rating: 5 of 5

04.Feb.2017 at 22:57 by Eric Poe (96 comments) via Web2 LIVE

I and several others had the same reaction: Holy ----! Excellent work. Great job setting up the problem and how you worked towards the solution.

I can't wait to bring this idea back to my team.

Rating: 5 of 5

05.Feb.2017 at 09:19 by Emil Gallant (10 comments) via Web2 LIVE

The solution presented was specific to unit testing, but it can be applied to many types of problems. This gave us a concrete plan for updating our test suites for increased efficiency.

© Joind.in 2019