際際滷

際際滷Share a Scribd company logo

Gatling Performance Workshop

Sai Krishna
Sai Krishna

This document provides an overview of Gatling, an open-source load testing tool developed using Scala. It discusses Gatling's focus on web applications, simplicity, high performance, good reports, and easy integration with CI tools. It then covers setting up Gatling, including prerequisites, writing a sample test, and key concepts like scenarios, simulations, and reports.

1 of 20
Downloaded 13 times
PERFORMANCE TESTING WITH GATLING Thoughtworks Bangalore
Overview What is covered? Brief about Gatling Pre-requisites Writing first test - hands on What is not covered? Performance load modelling Scala language itself Gradle - Build tool that we have used
What is Gatling? Performance testing framework developed using Scala with a focus on web applications Server side Written in Scala
Why Gatling?.. 1. Open-source tool 2. Simple 3. High performance 4. Good reports 5. Easy integration with CI
Pre-requisites 1. Java 2. Scala 3. Any build tool (optional) - We will be using Gradle 4. Any IDE like Eclipse or IntelliJ
Machine set up https://github.com/swetashegde/gatling-for-beginners has detailed steps for Mac and Windows set up with a sample project.
Lets begin..
Run a sample server to test To run a local server on machines: mb start --configfile=mb-config.ejs Use postman to check the endpoints.
Create a Gradle Project
Create a Gradle Project Edit your build.gradle file: Add dependencies Add task gatling
Create the Project Skeleton Create the following structure for your project:
Some Keywords val - constant value object - class with a single instance exec - execution step http - declares an http request scenario- declares a scenario Find more : http://gatling.io/#/cheat-sheet/2.2.2
Setting up the protocol This is where we set up: Base domain for our tests Example: http://www.example.com/search Request headers that the target APIs accept Example: Content-Type: application/json
Writing a simple query A query can be constructed for each endpoint under test. Writing this way would make the tests simpler and readable Query consists of: Query name and request name Endpoint with request method Any specific params Expected response Syntax: val <name> = http(<query_name>) .<request-method>(/endpoint) .check(<expected_status>)
Scenarios Scenarios are essentially user flows that we would want to simulate. Usual way to construct a simple scenario is: Add queries in a sequential manner to create a user journey Add waits in between if required Syntax: val name = scenario(<scenario_name>) .exec(query1) .pause() .exec(query2) ...
Simulations Simulations are where we define the amount of load we want to inject into our servers Things required to construct a simulation: Scenario Protocol Assertions Syntax: setUp(<scenario>.inject(<injection_step>)) .protocols(<protocol_name>) .assertions(<condition>)
Get..set..run!! gradle clean gatling
REPORTS
References & credits http://gatling.io/ https://github.com/lkishalmi/gradle-gatling-plugin http://www.mbtest.org/ https://www.thoughtworks.com/insights/blog/gatling-take-your-performance-tests-next-level http://gatling.io/#/cheat-sheet/2.1.7
THANK YOU

More Related Content

Gatling Performance Workshop

  • 2. Overview What is covered? Brief about Gatling Pre-requisites Writing first test - hands on What is not covered? Performance load modelling Scala language itself Gradle - Build tool that we have used
  • 3. What is Gatling? Performance testing framework developed using Scala with a focus on web applications Server side Written in Scala
  • 4. Why Gatling?.. 1. Open-source tool 2. Simple 3. High performance 4. Good reports 5. Easy integration with CI
  • 5. Pre-requisites 1. Java 2. Scala 3. Any build tool (optional) - We will be using Gradle 4. Any IDE like Eclipse or IntelliJ
  • 6. Machine set up https://github.com/swetashegde/gatling-for-beginners has detailed steps for Mac and Windows set up with a sample project.
  • 8. Run a sample server to test To run a local server on machines: mb start --configfile=mb-config.ejs Use postman to check the endpoints.
  • 9. Create a Gradle Project
  • 10. Create a Gradle Project Edit your build.gradle file: Add dependencies Add task gatling
  • 11. Create the Project Skeleton Create the following structure for your project:
  • 12. Some Keywords val - constant value object - class with a single instance exec - execution step http - declares an http request scenario- declares a scenario Find more : http://gatling.io/#/cheat-sheet/2.2.2
  • 13. Setting up the protocol This is where we set up: Base domain for our tests Example: http://www.example.com/search Request headers that the target APIs accept Example: Content-Type: application/json
  • 14. Writing a simple query A query can be constructed for each endpoint under test. Writing this way would make the tests simpler and readable Query consists of: Query name and request name Endpoint with request method Any specific params Expected response Syntax: val <name> = http(<query_name>) .<request-method>(/endpoint) .check(<expected_status>)
  • 15. Scenarios Scenarios are essentially user flows that we would want to simulate. Usual way to construct a simple scenario is: Add queries in a sequential manner to create a user journey Add waits in between if required Syntax: val name = scenario(<scenario_name>) .exec(query1) .pause() .exec(query2) ...
  • 16. Simulations Simulations are where we define the amount of load we want to inject into our servers Things required to construct a simulation: Scenario Protocol Assertions Syntax: setUp(<scenario>.inject(<injection_step>)) .protocols(<protocol_name>) .assertions(<condition>)
  • 19. References & credits http://gatling.io/ https://github.com/lkishalmi/gradle-gatling-plugin http://www.mbtest.org/ https://www.thoughtworks.com/insights/blog/gatling-take-your-performance-tests-next-level http://gatling.io/#/cheat-sheet/2.1.7