testing application server response [closed]

Closed. This question does not meet Stack Overflow guidelines. It is not currently accepting answers.

---

开发者_如何学Python

We don’t allow questions seeking recommendations for books, tools, software libraries, and more. You can edit the question so it can be answered with facts and citations.

Closed 7 years ago.

Improve this question

My application has a two completly seperated application servers. They speak with each others in various models for instance. A puts a file in a Folder B listen and reads the file, create a response FILE in a new Folder A picks up the "response" file.

Same thing with JMS: A publish something B will listen and Reply

Now, I want to test this with application A only so that I have some sort of dynamic stub simulating B that is triggered when my real application A is creating the file or the Topic.

Any framworks that can support this?

I saw Mactor, but it looks like a dead project, any other sugegstions?

---

You should be able to create a mock server for B based on the real code for B. I would use that to create your mock instance for testing purposes.

---

To solve this requirment, to perform relativly good level of integration testing with external system we have no control over. We used Jmeter. Jmeter has the capability to listen on a JMS resource, wait for a request and act upon it, we used a set of stubs to represents the logic of the real system that way.

---

I'm not sure why you need a framework for this. You need to be able to drive files and JMS from inside a JUnit (or whatever) test, but that is a simple matter of coding.

You can then write tests that do things like:

1. Create a file in A's input directory
2. Wait for a file to appear in A's output directory
3. Validate the contents of the file

For more complicated series of interactions, simply extend the sequence with more action - wait - validation cycles.

Perhaps you are thinking that you need a framework that can react to different events coming from A, so that you don't need to write sequences of interactions out explicitly. I would say this is a bad idea. A test should be an explicit sequence, so it's clear what it's doing.