We have 2 CCR1036 with OSPF and a handful of subnets that are advertising to each other. Out of the blue starting about 2 days ago we lost access to some subnets on the other side of the network (aka: the other CCR1036) But if I got behind the CCR1036 that we were not able to access I could route to all subnets fine.
After researching and troubleshooting I hypothesized that the routing tables were frozen or stuck. Other users reported that there is no way to flush or clear the routing cache.
Did you check the routing tables during the outage to confirm that all IP routes pointed in the proper direction, and on both routers?
I’m having a little trouble visualizing your issue because you speak of two routers, but when you say you checked from “the router” it’s unclear which one you mean.
So let’s say you have routers A and B. Could you re-explain the failure referring to the two routers independently?
e.g.: I was on host x.x.x.x behind router A, and could not reach IP address y.y.y.y which is behind router B. When I logged into router A, A could ping y.y.y.y but host x.x.x.x still could not. (etc)
Are there any other features in play such as RP filter (loose / strict mode), stateful firewall, etc?
Router A could ping Router B but could not ping devices behind it.
Router B could ping Router A and could ping all devices behind it.
No firewalls were in place. I checked routing tables and all was working.
More to add about OSPF connectivity between both routers…(both are connected by a /30 network)
Router A could ping it’s own /30 address but could not ping the Router B’s /30 address
However…
Router B could ping both /30 addresses.
Very weird situation I was in. But a reboot on Router A solved the issue/bug I was experiencing.
