Im using routerOS 6.35.4 stable with CCR1036 for all packages. Before resetting the router which would give a 10ms ping for no reason, it now gives <1ms ping but the NTP server doesnt work. I have gone through my firewall and NAT rules.
On windows it says the peer is unreachable when i try to sync with router directly. When i only use unicast it still doesnt work.
The time shown on winbox for the router and my computer time is different by 1 minute. On my router i use cloud to update time.
You know that NTP server is an optional package that you need to upload and install, then configure?
The basic NTP function is only a client.
For me the NTP server works, but I think it is too limited.
(only two time sources, no monitoring interface so ntpq -p or ntpq -c rv is not working)
The result is when you misconfigure it you have no way of knowing what is going wrong.
What I found is that the last properly working NTP server version was in ROS 6.32.4
The later ones are lingering in the init state and that’s it.
My setup is only an USB GPS with NTP server package on a hAP.
Does the NTP package work for you with external servers as a reference?
For me it does, also with 6.35.4
It is not clear if the topic starter uses external servers or a local GPS receiver.
I never tried that since I have a local Stratum 2 NTP server running in the network, and I only use the hAP + GPS in the attic as a backup time source for that server in case of uplink failure.
Let’s see what SystemErrorMessage uses…
I use IP cloud for time sync and it seems to be 1 minute ahead.
NTP server stopped working after i reset router.
6.37.rc11 has a fix for the NTP server using the system clock as reference. Maybe this will solve your issue.
But you could always add a public NTP time source using the NTP client (e.g. pool.ntp.org), which will get your NTP server working.
It works now thanks. used ntp client instead,