Latency, how do you come up with the number?

binarysoul

How do network monitoring tools derive latency and delay on a link?

When I hear that there was 500ms delay, I wonder how it was derived. I want to know the math behind it. Is it 500ms for each KB of data, for a packet to get from source to destinatin or what?

Please show me the numbers and be my guest to be as detailed as you want to be

dtlokee

It's usually measured on single packets of a set size to gather the information. In some cases you can specify how you want the packets created to measure latency with and without QoS applied, some packets could look like RTP packtes and some could look like FTP packets to see the difference.

garv221

dtlokee wrote:

It's usually measured on single packets of a set size to gather the information. In some cases you can specify how you want the packets created to measure latency with and without QoS applied, some packets could look like RTP packtes and some could look like FTP packets to see the difference.

Correct. There is a million different ways to determine latency.

In my mind, I just picture data packets doing suicides like in basketball practice, there and back.

JDMurray

garv221 wrote:

Correct. There is a million different ways to determine latency.

Well, there are a million different reasons for latency, not a million ways to calculate it. Latency itself is simply the interval of time between initiation and action.