Hello everyone,
I don't quite understand the behaviour of the root dispersion of Windows NTP servers.
I would like to set up a network of Windows-based and non-Windows-based devices.
I would like to define a computer with Windows 10 (Enterprise LTSC version 1809) as the NTP server that is to supply non-Windows-based devices with Chronyd time service with a synchronised time. However, the computer is only a secondary reference and receives its time as a client from an external Windows NTP server. Unfortunately, this secondary server receives the time with too high a root dispersion (~10s fluctuation). According to the manufacturer, the devices with Chronyd time service that are supposed to connect to this time server only accept a root dispersion of less than 3s.
How does the root dispersion generally behave in such an architecture? Is the root dispersion passed through over several levels or does it add up over several levels? Is it possible to forward the root dispersion on the secondary server with a value less than 3s without adjusting the root dispersion of the primary server?
Maybe you can enlighten me about the behaviour or have an idea how I could implement this scenario.
Best regards
Tobias