I am wondering how these really effect my networks.
Say I have the following
Network A - IP 192.168.0.0/24 Interface IP 192.168.3.2/30
|
|
O - Mikrotik Gateway to Internet, With masquerade out to internet
Interface 0 - xxx.xxx.xxx.xxx to Internet
Interface 1 192.168.3.1/30 - To Network A
Interface 2 192.168.3.5/30 - To Network B
|
|
Network B - IP 192.168.1.0/24 Interface IP 192.168.3.6/30
Now I want to be able to ping from a pc at 192.168.1.4 to a pc at 192.168.0.4
I have static routes but the ping does not make it past the mikrotik. I can see it get there but it does not relay to the next router. Both directions this is an issue.
OSPF is in place and I can see all of the routes. They pings just will not pass.
Is this a scope issue? All interfaces are the defaults:
Scope 255
Target Scope 10
This isn’t a routing issue. It’s an IP addressing issue.
e.g. Network A. The network address is 192.168.0.0/24 but from a client perspective the router’s interface is in a completely different network 192.168.3.0/24.
I must disagree with you on that one. If you have a static route that
192.168.0.0/24 gateway 192.168.3.5 then when the packet gets to the mikrotik there is a static route that 192.168.0.0/24 gateways to 192.168.3.2
That is a static route and the ping should pass with no problems. The clients on each end have default gateways so what they do not know a route for will go to the gateway, from there on routing takes over and a ping should pass.
I did however find what the issue way. It was that there are transeo backhauls hooked up and so proxy-arp had to be enabled. Now pings route perfectly.
So, still my question remains the same. I would love some clarification on how the scope and target-scope effect a network with the basics of the one above.
Proxy ARP will hide a whole multitude of mis-configurations. From a network administration point of view that’s a bad thing because a simple network change can cause unrelated stuff to break.
Maybe I mis-read your problem but IP networking requires that the clients on a network be in the same network as the gateway interface that they’re connected to. From your description, that doesn’t appear to be the case.
The network description I gave is kind of like the way the internet works. Since we are dealing with Routing it is not bound by a local networks limitations as long as the routes are in place so the routers know what way to send everything. Each network has a gateway router that is connected to the main router. Their gateway interfaces have ips that are in the same network as the main router. Since the only masquerade is from the main router out to the internet, everything inside can be routed to allow talk across networks.
The reason behind the Prox Arp is because the transeo radios (bridges… Yuck) require it because they use mac-nat and proxy-arp for bridging.
scope and target-scope control the recursive route lookup process. That essentially specifies, which routes would become active (“A” flag) in the routing table.
