What qualifies a “bad” broker/vps combination? After 9 years of FX trading I would like to be educated. I have a vps with winserver 2012 with latency under 1ms with my broker. I am testing an EA at the moment and the vendor claims that I don’t meet his numbers due to a “bad” broker/vps combination. Decisive for me always was the quality of the data feed and the rest is up to MT4 and the EA. Please correct my line of reasoning/thinking.

bsternfeld 

What qualifies a “bad” broker/vps combination? After 9 years of FX trading I would like to be educated. I have a vps with winserver 2012 with latency under 1ms with my broker. I am testing an EA at the moment and the vendor claims that I don’t meet his numbers due to a “bad” broker/vps combination. Decisive for me always was the quality of the data feed and the rest is up to MT4 and the EA. Please correct my line of reasoning/thinking.

Buying a system, that's the problem.

bsternfeld 

What qualifies a “bad” broker/vps combination? After 9 years of FX trading I would like to be educated. I have a vps with winserver 2012 with latency under 1ms with my broker. I am testing an EA at the moment and the vendor claims that I don’t meet his numbers due to a “bad” broker/vps combination. Decisive for me always was the quality of the data feed and the rest is up to MT4 and the EA. Please correct my line of reasoning/thinking.

You should write a smal EA which is simply sending a limit order far from the current price (to not being filled), cancel it when it's acknowledged by the broker, and measure the time spent between the two.
That will give you a realistic latency to your broker.

The 1 ms you mention might be only the network latency itself, not the "order in the brokers/exchange book" latency which is higher (10 to 1000 higher depending on the broker's back-office procedure).
I won't say it's the problem here but at least you'll have real number to give to your EA vendor, like "order sent and cancelled in 8 ms".