In networking, what does 'latency' refer to?

Latency is primarily defined as the delay experienced in a system before a transfer of data begins after an instruction is given. This encompasses the total time taken for data to start being processed, which can involve various aspects such as the processing time of the system, the transmission time across a network, and any queuing delays.

While it's true that latency includes the time from when a request is issued to the time the first byte of data starts to arrive, its focus is specifically on that initial delay before any actual transfer occurs. This makes it a crucial metric in networking, as high latency can result in noticeable delays in applications, particularly those that rely on real-time data and interactions, such as video conferencing or online gaming.

Understanding latency is vital for network optimization and performance troubleshooting. It's important to differentiate this from other concepts such as bandwidth or throughput, which refer to the amount of data transmitted in a given time, rather than the delay in data transmission itself.