We have a weird problem where connectivity is lost between 2 routers. The problem seemingly comes and goes randomly, which makes it very difficult to diagnose.
Setup is as follows:
Router2 is connected via a switch to a Mimosa Radio, on the other side Mimosa radio connects directly to Router3. All routers on ROS V7.9
(The wireless link is not the problem, both radios are accessible at all times)
Gateway(R1) → RB3011(R2) → CRS328(Switch) → MimosaC5c (( — )) MimosaC5c → RB960PGS(R3)
There are other Links and AP’s connected to R2 via the switch, and we have no problems on those links.
The problem started after we replaced the two basebox5’s which made the wireless link between R2 and R3 with Mimosa radio’s.
Due to some legacy setup the radio on R2 side was setup as a router. With the new setup both radio’s are only bridges.
For the first few days the new setup worked, but the problem appeared shortly after a lighting storm in the area. We suspected damage on R3’s side and replaced R3 and the radio on the far side, but that didn’t solve the problem.
When the problem starts to manifest both radios are accessible, showing the wireless link is fine.
I can ping both radios from R2, and weirdly, I can ping R3 from the radio directly connected to it. Sadly I haven’t tried to ping R3 from the near side radio yet.
R2 see R3 in its neighbour list, but I cannot MAC ping from R2 to R3.( Mac ping is fine when the link works)
One time I was able to access R3 via Romon, on another occasion I was not.
When I was able to access R3 via Romon, rebooting R2 and R3 didn’t resolve the issue
The last time the issue was resolved after rebooting both radio’s. At other times it resolves by itslef without any action on my part.
Clients behind R3 report very slow connectivity at times and total loss at other times, so the link isn’t completely dead all the time.
When I was able to access R3 via Romon I found something very strange. Running a traceroute from R3 to some other routers on the network(via R2) the traceroute would always fail on the hop previous to the router being tested to.
There are no NAT rules or mangle rules that can cause these problems. R3 have only static routes, no OSPF setup.
If there is a L3 setup problem, how come MAC ping fails between the routers and if it is hardware related how is the weird traceroute results explained?
I have include the relevant setup on the devices in question below.
RB3011(R2)
/interface ethernet
set [ find default-name=ether1 ] name=ether1-R1
set [ find default-name=ether2 ] name=ether2-CRS328
/interface vlan
add interface=ether2-CRS328 name=vlan200-Link_to_R3 vlan-id=200
/routing ospf instance
add disabled=no name=ospf-instance-1 redistribute=connected,static
/routing ospf area
add disabled=no instance=ospf-instance-1 name=ospf-area-1
/ip address
add address=10.0.0.254/24 interface=ether1-R1 network=10.0.0.0
add address=10.0.1.253/24 interface=vlan200-Link_to_R3 network=10.0.1.0
/ip route
add disabled=no dst-address=0.0.0.0/0 gateway=10.0.0.253 routing-table=main suppress-hw-offload=no
/routing ospf interface-template
add area=ospf-area-1 disabled=no interfaces=ether1-R1
add area=ospf-area-1 disabled=no interfaces=vlan200-Link_to_R3
/tool romon
set enabled=yes
CRS328
/interface bridge
add ingress-filtering=no name=bridge1 vlan-filtering=yes
/interface ethernet
set [ find default-name=ether1 ] name=ether1-R2
set [ find default-name=ether2 ] name=ether2-Link_to_R3
/interface bridge port
add bridge=bridge1 interface=ether1-R2
add bridge=bridge1 interface=ether2-Link_to_R3 pvid=200
/interface bridge vlan
add bridge=bridge1 tagged=ether1-R2 untagged=ether2-Link_to_R3 vlan-ids=200
/tool romon
set enabled=yes
RB960PGS(R3)
/interface ethernet
set [ find default-name=ether1 ] name=ether1-Link_to_R2
/ip address
add address=10.0.1.254/24 interface=ether1-Link_to_R2 network=10.0.1.0
/ip route
add gateway=10.0.1.253
/tool romon
set enabled=yes