MPLS overload problem

RouterOS Forwarding Protocols
1

Hi guys, first sorry my english.
I have a problem implementing MPLS in my network. When i enable LDP my routerboard stops, the CPU up to 100%, and wireless link down.

This is my scenario:
1x CCR1072 (192.168.44.1/30) <= vlan l2mtu=1596(default) => 1x 911ac (192.168.44.2/30)
-both with OSPF in this networks
-loopback interface with unique ip address /32
-OSPF alone is working 100% fine
-transport address and lsr id is same loopback address

some images:

Thanks for assistence

2

Please share your MPLS and OSPF config and topology.

3

After many tests I found the problem…
The router with all client´s routes, over 5000 routes (/32 one per client) received by all routers on mpls cloud. This cause a overload in all routers in the mpls. I droped all client routes and the routers are running 100% fine with low processor usage.
Is my fault. MPLS cloud must have only transit routes, not customer routes.

If anyone can supplement or correct something would be interesting.

Thanks for all.

4

Hello uottoni, could you list out the steps you followed to solve this?

5

MPLS cloud must have only transit routes, not customer routes

That’s correct, you may need to use OSPF as an IGP which will be the one who can cares about convergence, load balancing, and loopback distribution, and BGP as an EGP distributing your LANs subnets or public blocks. That config is a base for a scalable network diagram.