Member-only story
Is Upgrading From .Net 3.1 to .Net 6.0 Really Worth It?
Let’s test and find out
The .Net Core 3.1 is a great framework, and since the long-term support (LTS) ends at the end of this year (Dec 2022), it may be a great time to start considering upgrading your existing applications.
But despite the LTS end date and the new language feature, does it worth the effort from a performance perspective?
I will try to answer this answer by doing a few tests, and making the test code available, so you can reproduce or customize the tests the way you think is more appropriate.
If you’re not a Medium paid member, you can read this story for free using the following link
My Previous Test Experience
Some time ago, I ran my performance tests locally on my (humble) machine to compare the performance between .Net and Node.js. The results really impressed me. As a big fan of .Net, I decided to run a similar test, but this time between .Net 3.1 and .Net 6.0 using the same approach.
About the tests
I will follow a very similar approach to my last tests; I will define these three tests scenarios:
- Hello World: A simple endpoint that returns the text “Hello world .net ”
2 . Reading file: An endpoint that reads a file with 100 lines and returns the content.
3 . Bubble sort: An endpoint that fills a long array with numbers in descendent order, sorts the numbers using the bubble sort algorithm, and returns the list ordered
Note: On my last performance tests, I used a synchronous reading file approach for both .Net and the NodeJs apps and I received a few comments pointing that out, but even after changing to use async-await and the performance difference does not change significatively, anyway, for these tests I start using async-await since the beginning.
The test tool
I will use the K6 tool to do the tests running 10,000 requests in parallel and using 30 virtual users requesting with no time between the requests.
To set this up, just add the k6test.js file below at the root of each project and change the…
