Hi all, I am currently evaluating the performance of a cloud application(Project Management tools). There is one action I would like to evaluate first, but I got problems for how to validate which requests I should look for for this particular action.
Let's mention about the action I need to test first. There is a project page which shows the details of a project. Inside a project there will be some tasks. Let's say task 1, task2, task3 and. They have start date and end date. I will update the field start day. As the tasks are linked, if I change the start date of task1, the start date of task2 and 3 will be updated too.
(HERE COMES THE ACTION). I update the start date of task1 by choosing a new date in a calendar box then the page will automatically update the start date of other tasks for me. I would like to know the request response time for this whole action. And the below is my approach now.
I create a page break (e.g. logging)before clicking the calendar box to update. Then I click the date in the calendar box to perform the action, so the transaction/requests recorded under this page break folder will be all the requests occurred after I perform the update action. Then, I just check the request-response time from the result.
It looks working well, but I don't know if it is the correct approach. Here is some of my doubts.
1. Does my approach work?
2. To my understanding, neoload record every request and response in the user path I recorded. So, if I click on a request, will I be able to trace back what action I did? (e.g. maybe with the help of the inspection from the website's javascript or looking into the details of the request/response in the neoload built in "Check user path" function?) For example, in the attached login request.png, I can see the login parameters in the request, so I know that's the login request. However, for the other request, such as Clarizen/bundle/appCore.js , I don't know what does it mean, is there anyway to know where the request was called or what action caused this request?
3. Let's say I updated the start date of task 1 from 13 Nov to 16 Nov. Then, I change the date back to 13 Nov after the recording is done. It appears that if I run the load test using that user path. The date of my project will actually changed back to 16 Nov again. Sorry maybe it's a little dumb to ask, is Neoload really simulate a user to actually do all the things I recorded (so my record is actually updated after running the test) If so, does it mean I need to change the date back every time I execute the load test?
Thank you for your time to read through all of this. Any comments will be appreciated.
In order to get the response time of your transaction "change date" you need to create a NeoLoad transaction so your actions are recorded inside that NeoLoad transaction.
To do so it's during the NeoLoad recording that you create your transaction using the recording bar. Type the name of your transaction, change your date and then stop the recording or change the name of the transaction to create a new one.
In your application when you change your date it may fire some other requests maybe fired by client Javascript and it's not an issue. Your goal is to get the response time of the whole transaction. It does not mean that an action on the UI one request is fired it can be multiple requests.
And of course when you execute your virtual user it will make real modifications in your application. So you may need to change that date before all new tests and you can include it in your script so it's automatically done for you.
