Hi,
We'd like to force chronyd to verify that all NTP servers are currently available in a deterministic fashion. Presently, we have a computer physically connected via Ethernet to a local NTP server.
The /etc/chrony/chrony.conf contains:
server 10.10.10.10 minpoll 0 maxpoll 0 ... etc.
makestep 1 3
Chrony queries the server at a regular 1-second interval, as expected. When running libchrony's examples-report, we see:
State: Selectable
Mode: Client
This is expected. For test purposes, we want to drop the server connection and verify the result. For example:
iptables -A OUTPUT -dport 123 -d 10.10.10.10 -j DROP
It seems chrony takes non-deterministic time of approximately 1 second to determine that the NTP server is offline.
Can we force chronyd to refresh its state immediately, such that upon returning from the request (or command) the State is guaranteed Selectable for a valid connection and Nonselectable when severed?
At present, we need to introduce a 1.1-second delay in our tests to give chronyd time to detect the NTP server disconnection. While this works, it slows down the unit tests and introduces knock-on issues.
Do we need to call something similar to the following:
chronyc -m 'burst 3/3' 'makestep 0.1 3'
Thank you!