I have been using Karate for our Java application & its working out pretty good. Some teams in our company have been using python in their Development & QE process. They are interested in karate, but is there a possibility to run python scripts within karate tests?
Atleast 3 different teams have come to us asking for this. Please let me know if anyone is already doing this (or) if there is a possibility to do this. Thanks.
I think you have some pretty good options here.
First, look at this answer. It will be very easy to call any command line program from a Karate test: https://stackoverflow.com/a/51150286/143475
A new development in Karate is that there is a stand-alone JAR (binary) which contains all the capabilities, including parallel running and HTML reports. This is ideal for teams that don't want to set up a Java IDE - only a JRE is sufficient. So you can even invoke Karate from the command-line. You can even edit / debug tests using the Karate UI - but this still needs work and we are looking for contributions.
Please refer the documentation here, and you can try this within a few minutes: https://github.com/intuit/karate/tree/develop/karate-netty#standalone-jar
Related
Currently, we've automated all of our web services using the Karate framework and it is super cool.Is there any way we can do the load testing with this tool?
Example: Given when I enroll a member 1000 times
Then the average response time should be 2.7 milliseconds
I'm the developer of Karate and it is great to hear this feedback, thanks !
Karate was not designed for load-testing - but we would like to support it in the future. But it is not a priority now, and there are things like mocking we want to get to sooner which we feel is a more critical need for many teams.
If you or the community would be willing to investigate and contribute, that would be great. One experiment I can think of is using the Java API try to create multiple threads and run a *.feature file.
EDIT: Karate now has Gatling integration: https://github.com/ptrthomas/karate-gatling-demo
Yesterday as part of my POC work w.r.t API demonstrated Karate in my office people were very much impressed. thanks for Karate team.
I am new to test and during my intership, I had to look for some good tools to automate functional tests.
So I made a lot of searches and decided to use Cucumber, linked with Selenium and SoapUI.
But the fact is that another search had been made before to automate load tests and Gatling (used with Jenkins) had been chosen for that.
Then I am asking to find a solution to gather the 2 solutions but it is quite hard to find any solution about that.
The only information I catch was using Taurus to have a single configuration file.
If you have any information which you think coul help me, I would be very thankful.
The purpose and techniques used by cucumber and gatling are very different so you will probably not want to push interoperation too far.
For example, you plan to use Selenium in your cucumber steps. That would not work well with the performance testing.
If you use jvm version of cucumber, you'll be able to reuse some of the utilities, for example test builders.
I'm now at the level where I am able to build my own test automation framework using TestNG, through Selenium. The next stage is for me to look at kicking off automation runs, periodically using Selenium Grid.
I've being doing some research on this, and the CI tool, Jenkins, keeps cropping up. However, there isn't anything out there (that I can find!) that will answer the questions I have on it. Any help would be much appreciated on the following:
1) Why is Jenkins used with Selenium...what does it bring to the party?
2) Can Jenkins be used in conjunction with Selenium Grid? If so, I'm just after some guidance as to how they inter-link?
Andy
See comment to your question. (new member, and I can't yet comment, so simply restating the comment as a formal answer).
As mentioned in the comment, Jenkins is simply a CI tool. As to what it brings to the party: It has some nice features and plugins. So it depends on what you want. You should just check out: http://jenkins-ci.org/
Yes. Jenkins provides several different ways to invoke. It has the ability to string build invocations together so you can even build, deploy, then start selenium grid.
is there any tool out there that i can used to set-up run automatically and i was goggling and i found selenium test runner? there are so many tools out there its hard to figured out which is best
I'm using C# and using MSTest as a test framework and I'm looking forward to see if I can get a way from testing in MSTEST
any help?
This is very subjective question. Every requirement will have its own correct answer. Anyhow I will try to address few requirements and will be updating as I learn more.
If you are automating web app browser tests (sans flash player and silverlight) I would say that selenium is the way to go. There are ways to automate flash and silverlight too, but that is answer for another question.
Selenium is anyways an automation too and your choice will rather is of which test framework to select. So here are few options:
1. Integrating with CI tools:
If you want to organize your tests as segregated atomic units and want them to be integrated to some CI server (e.g. TeamCity). I will recommend using NUnit to run your selenium tests.
2. Behavioral Tests
It is a new trend in the software development and how we test our products. Using behavioral (i.e. business specification) like language. In my experience it is also a very good format to write up acceptance tests. You can use selenium with something like Nbehave or SpecFlow
3. Centralize Test management and Execution
Now this might not fit for everyone but I have found FitNesse (and its c# binding) to be very useful in maintaining and executing selenium test cases.
Please note this answer may not be right and is certainly not complete given the scope of the question. I have nevertheless tried provide few pointers.
Does anyone know a good way to mix and merge multiple testing frameworks together so that they can all be run in batch and return an solid overall total of which tests failed for which frameworks and suites/specs.
So lets say my testing setup for a particular project looks like so:
I'm using Rails (Ruby) and using RSpec to test that.
I'm also using Cucumber with my Rails application.
I'm using MochaJS with the Testacular runner for JavaScript testing.
I'm using Jasmine to test for some NodeJS applications that I'm using as well.
Now to test each of these test groups I would have to launch each of their respective frameworks/instances, start and run the "run tests" operation for each, and then tally up the results and figure out which tests failed and which ones didn't.
Does anyone know of a tool that is designed to do this?
You probably need a build automation software to perform all these task together.
Whenever one of your test process fails you'll get a detailed feedback.
As you're developing a ruby application maybe Buildr is the best choice, but you could as well use Ant or Rant...
you can find a complete list of tools is here:
http://en.wikipedia.org/wiki/List_of_build_automation_software