mschuerig (1) [Avatar] Offline
I think you're using the term "latency" much too frequently and a bit too loose.

Section 2.3.3 starts with a sentence that to me appears to be plain wrong: "Latency is the maximum time it takes from the time a request is received until it has been handled and the response sent."

The time you're referring to is the service time.

The book "Systems Performance" by Brendan Gregg (Prentice-Hall 2014) has some nice and crisp definitions (sec. 2.1) that you might want to adopt:

"Response time: the time for an operation to complete. This includes any time spent waiting and time spent being serviced (service time), including the time to transfer the result."

"Latency: Latency is a measure of time an operation spends waiting to be serviced. In some contexts, it can refer to the entire time for an operation, equivalent to response time."

By those definitions, most of the usages of the word "latency" in your book refer to its second meaning, ie response time. I for one would much prefer you use "response time" in all those places. It is clearer, less pompous and leaves "latency" for when you're really talking about waiting time.
roland.kuhn (39) [Avatar] Offline
Re: Latency, service time, response time
Thanks for the feedback, quick research has confirmed my understanding that latency is a rather generic term, denoting the time passing between the application of a stimulus and the onset of the associated effect (be that a viral infection or the transmission of light waves in a fiber). While I do not agree that use of such a technical term qualifies as pompous, it is indeed a good idea to pick more concrete terms where applicable; service time and response time are good candidates. We’ll take this into account for the next update.