local subnet issue with v4.17 (async routing)

Hi there,

Apparently I am missing something here. Running into an issue where the mikrotik doesn’t want to use it’s own interface on a subnet to send the traffic back towards said subnet …

The problem … I have a subnet hanging off of a RB493G, let’s say it’s 192.168.100.0/24. The routerboard is 192.168.100.1.

Devices on this subnet cannot ping 192.168.100.1, from which .1 is their gateway. They see the ARP entry for .1, but cannot initiate traffic to it. However, I can initiate a ping from the mikrotik and I get replies. I also see ARP entries for the other devices on the network. I see that there is entries in the local routing table for the 192.168.100.1 interface for 192.168.100.0/24. These appear to be created when I created the interface … So it appears that layer2 is fine …

So you would think that the mikrotik would shove return traffic back to this subnet on it’s own interface - but that is not the case.

Now for the fun part …Return traffic only occurs if I take a spare interface of the mikrotik’s gateway (a cisco router), and put it on this 192.168.100.0 network as .2. This tells me that the Mikrotik is shoving traffic from 192.168.100.0/24 back out it’s default route, and not learning the more specific route form it’s own interface, which obviously is .1.

While I am not new to routing and switching, I am learning about the mikrotik platform. What kind of output should I post here for some guidance?

Thanks all,

Post the output of “/interface print detail”, “/ip address print detail”, “/ip route print detail”, and “/ip firewall export”.

Thanks fewi. I think I see the problem. The route table has the destination address being .0/32 … not .0/24.
I looked at the IP addresses again, and now see that the address doesn’t have a /24 after it, hence it’s putting it in the route table as a /32. Booger.

I removed that test interface from the upstream router, re-did the addresses and it’s working as it should now.
Thanks for the reply and the pointers …