| || |
Bandwidth usage with SilkPerformer
I'm in the process of architecting two geographically separate test labs that will be connected by a dedicated point-to-point circuit (T1/T3). One question is how many virtual users can I expect to concurrently run on a T1 circuit? A T3 circuit? In the short term we want to test with 2000-5000 virtual users. There will be some PCAnywhere and ssh traffic as well, although this will be mostly nominal. Also, does anyone have experience load testing with burstable circuits? ($$$ is always a factor)
Re: Bandwidth usage with SilkPerformer
Bandwidth is always dependent on the amount of data that each virtual user will generate. The users run on the agent machines and are bound by the memory (~1M/user) on the machine.
If each user is then going to try and access a 100MB file, then there will be alot of data being sent down (one way) to the agent machine. The network card may be a bottle neck in this BIG DATA situation. If you have multiple agents on the same T1 it is possible that in a BIG DATA scenario to get close to the limits of a T1 line. I would guess that you would need a lot of machines to do the same to a T3.
Most leased burstable lines will average your data over a given period of time. You should read the fine print. If you have a relatively short test time, then the burst of data may be offset by long periods of idle time.