OSPF Implemention

Got a little network I am working on, but OSPF is not working the way I want it too, or at all for that matter.

Here is the config..

main site (192.168.1.x) – MT 1gig – ATT Router – NET – ATT ROUTER →
MT 532 – remote site (192.168.14.x)

Tried doing a L2TP connection from Remote site to the main site. I just picked numbers on the PPP secret on the main site MT. In this case, 192.1.1.1 for the main site, and 192.1.1.2..

OSPF is configured, but I don’t get the 192.168.14.x newtork nor any of hte statics from the main site.

Suggestions?

Everything works fine for me :slight_smile:

What is your ospf configuration?

When I try to change the OSPF area, add a new one, delete it, it sits there forever and eventually says it has trouble issuing hte command contact MT.

This is on a 2.9.41 box now, 1 gig cPU, also, the CPU meter shows 100% usage after that but traffic still flows.

I have now enabled the routing TEST package, and the CPU no longer goes to 100% in this configuration. NOW I have another issue.

My routers have TWO internet providers, cable and T1s. Both with routes, no biggy. I am creating one L2TP tunnel to my primary MT from my second, and one PPTP tunning over the cable provider! What I want is an automatic failover using OSPF.

The issue I am running into now, is that when OSPF sends out, it also sends the connected “internet publics” that we use to “CONNECT” the tunnels, hence, the tunnels stop working, give it a bit, and it starts working s the tunnels drop and the routes are removed!

How can i make OSPF only send out addresses in my private range of 192.x.x.x?

Thanx,

Dennis

it is simple enough, so, you have 2 different boxes with possible default GW, on one set OSPF to redistribute-static and set value a as link cost, and set up netwatch that will disable that static GW if link goes down, and enable if up.

on second router set up also redistribute-static, but set up cost value to at least a+1 so, traffic all the time will choose route with less cost, hence your primary gw, if it goes down you will OSPFmaticaly start to use second GW as first will not be there anymore.

I don’t think the issue is doing the OSPF, the issue now is when the ospf sends out on the backbone area, it also sends its directly connected routes. On the remote router, this would include the directly connected subnet that the remote router uses to estblish the tunnels. So, it thinks that subnet, is accross the virutal link and not out to the net.

Right now, i need a way to not send any other routes other than 192.x.x.x networks.

You can filter out unneeded external routes.
/routing filter add chain=ospf-in prefix=“192.x.x.x” prefix-
length=0-32 action=discard invert-match=yes

What does the prefix length referr too.

Right now I have ospf-in prefix=192.0.0.0/8 action=discard invert match

I’m assuming it is doing:

match OSPF packets coming in with a prefix of 192.x.x.x, and discard them, then the inverrt match, then basically says (NOT) 192.x.x.x.

Don’t know what the prefix length is for though, don’t say a whole lot in the docs.