On 16 January 2012 11:06, Oliver Lloyd <[email protected]> wrote: > As per Seb, it's a breeze to tell JMeter to use a proxy, if your browser is > pointing to a pac file just find out what that resolves to and use that > address instead. > > On a point of principle though, you should consider where the application > you are testing is and from where you want to simulate load. > > If your application is located externally to your corporate network then you > should not really be driving large volumes of requests out of your internal > network to get to it. Doing this risks not only inaccurate results but you > could also cause problems for your company's network. In this case, a direct > internet connection would be best, better yet, you should use an externally > based test rig to drive the test. > > If the application is located internally then you should consider which > internal network components you want to include in the test. Traditionally, > a lab setup with the minimum number of network variables is preferred - run > JMeter from machines plugged into the same switch as the application. You > certainly should not be sending requests all the way out of you network only > for them to turn around and come right back in again - that's messy, you're > setting yourself up for trouble later.
Good points. But if JMeter is just being used for functional testing, rather than load testing, it's not as important to avoid the proxy. > ----- > http://www.http503.com/ > -- > View this message in context: > http://jmeter.512774.n5.nabble.com/Newbie-Not-able-to-record-the-script-under-corporate-proxy-tp5139668p5148396.html > Sent from the JMeter - User mailing list archive at Nabble.com. > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] > --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
