> This is becaused it is measured to the end of the first read from the response.
I always thought that latency should be the amount of time between the request sent and the first byte received, not the first read of an entire packet. I've seen some cases, using JMeter 2.7, that latency is greater than elapsed time. I'll try to reproduce this issue and raise a bug report ASAP. I'll also try it on JMeter 2.9. Thanks sebb
