I want to do a performance test of a website so i am creating a script that mimics user behaviour. I am using blazemeter to record those scripts and upload it in jmeter. I have two questions: 1) Do the results of a record and play script vary when run on another machine or a different time ? 2) I am getting a 400 bad request error in one of the steps of the recorded script.
What should i do ? Is there any other way to test the web pages other than record and play ?
The chance of getting a good load test from recording is minimal as modern web applications widely use dynamic HTTP Cookies HTTP Request Parameters for different reasons (security, tracking client state, etc)
So after recording your test scenario "skeleton" most likely you will need to perform so called correlation - the process of
Detecting parameters is quite simple: just record your test 2 times and compare request defaults: if you see differences - you will need to perform the correlation.
Extracting dynamic parameters is a bigger story, choosing the right extractor depends on response type, for example:
Also be aware that there is a solution allowing to perform JMeter correlation in an automated manner so you won't to detect and handle the dynamic parameter manually, check out How to Cut Your JMeter Scripting Time by 80% article for more details.
