One thing that is essential to understand: when one does ping (ICMP echo request/reply) or traceroute (which triggers sending ICMP time exceeded), it’s a particular device’s L3 that has to get involved. So reported RTT does not include only delay on link from origin to that point, it also includes L3 processing time on responding device.
That being told, RB4911 has a much faster CPU than hAP ac and hence it’s expected to reply faster. At the same time device, which is running test, also has to process data (both data to be sent as well as return data). So basically it’s similar to running bandwidth test … which should never be executed on device being benchmarked itself. It’d be interesting to see if same phenomenon can be seen when running traceroute from a machine on LAN side of router … and where the difference becomes biggest.
Tracert on different routers shows some variation on the third hop (modem to isp) on my end. Sadly, my bad rural connection shows a lot more than 2 ms - jitter is terrible with my provider. Router to modem is always 1 ms though, and my ping times/jitter to Google are more or less stable.
Which ROS version are you using ? Run your tests from outside the router to make sure the results are repeatable.
Does anything change if you change CPU frequency on RB4011? Default since 6.48 (or something) is auto, what happens if you change it to some static value (I’d try something around 1700MHz just to be on safe side, I don’t know what is nominal frequency)?
i tried @ 1400 and 1700 but same results. i have routeros v7.8
this is my config, export compact, i’m not an expert but i think there’s nothing wrong or at least nothing that can cause this problem. 4011-compact.txt (4.02 KB)
The configuration shows that your firewall filter is incorrectly configured. As a recommendation, use the default firewall rules at first and then supplement them with the configuration you need.
If you configure according to this tutorial, you will also have proper traffic flow and security. https://forum.mikrotik.com/viewtopic.php?t=180838
thanks for the link, i reconfigured the firewall part following the guide.
anyway i found that the problem was the ethernet port 1 of the router, connecting the ONT to other ports the ping came back normal.
it is a test that I hadn’t done because the line went perfectly apart from those extra 3ms.