NTP (Network Time Protocol) is one of the oldest protocols still in use today. It was developed in the 1980’s when the internet was still in its infancy and was designed to help computers synchronise together, preventing drift and ensuring devices can communicate with unreliable time causing errors.
NTP is now packaged in most operating systems and forms the basis for time synchronisation in computers, networks and other technologies. Most technologies and networks use a network time server (commonly called an NTP time server) for this task.
These time servers are external devices that receive the time from a radio frequency or GPS signal (both generated by atomic clocks). This time signal is then distributed across the network using NTP ensuring all devices are using the exact same time.
As NTP is ubiquitous in most operating systems and the internet is awash with sources of atomic clock time, this begs the question of whether NTP time servers are still necessary for modern computer networks and technology.
There are two reasons why networks should always use a NTP time server and not rely on the internet as a source of time for synchronisation. Firstly, internet time can never be guaranteed. Even if the source of time is 100% accurate and kept true (incidentally most sources of internet time are derived using an NTP time server at the host’s end) the distance from the host can lead to discrepancies.
Secondly, and perhaps fundamentally more important to most business networks is security. NTP time servers work externally to the network. The source of time either radio of GPS, is secure, accurate and reliable and as it is external to the network it can’t be tampered with en-route, or used to disguise malicious software and bots.
NTP servers don’t require an open port in the firewall, unlike internet sources of time which can be used as an entry point by malicious users and software.