I need to find a tool which can help us to measure performance time of our application.
When I run Jmeter inside my company's firewall Jmeter's performance time is around 40 seconds when manually it's 5 seconds. Our internet pipe is unify but any way I can not explain why this time is so big. An amount of downloaded Mb is higher then what I see from firebug but it's close.
I run the same scenario from home. Jmeter's performance time was 6 second when manually it was 3 seconds. It's close but I'd like to understand why I have this difference and can I base on Jmeter's data. I understand that I have to compare an apple with an apple. I can not compare a manual result with Jmeter's but any way I'd like to understand why those results are different.
Concurrency then - 2.5 simulates concurrent downloads. It's an old discussion, try searching around but the shortcut to the conclusion is: JMeter is not a browser, it does what it says on the tin, rather well as it goes, but will not replicate client-side, browser behaviour.
"Jmeter's performance time is around 40 seconds when manually it's 5 seconds.“
How did you measure these values? Are these values for a specific http request or for a specific page (a page has a set of http requests - unless you have tested a web service.)?
There is no proxy for my browser.
I am curious. JMeter is not a browser but what's the difference between Jmeter and browser? We don't have to accept Jmeter's result absolutely?
By the way I downloaded WART Pro 2.5 and hit our home page. It shows responsible time is 49 seconds also.