I posted a few days ago in regards to an issue redistributing routes from MPLS/BGP to OSPF and have returned with more concise info and the issue as i see. Hoping for some more valued input!
Simplified topology is below:
MPLS/L3VPN running on the core routers, with a corp vrf we use. Between the CLL routers and R1 i want to use OSPF to redistribute the routes learnt from all core routers. The export policy on the core routers is as follows:
term 1 {
then accept;
}
Pretty standard and simple. Redistribute all the routes it knows into OSPF. These routes are in the OSPF database on the core, but they are not going into the routing table of R1. Upon looking at the OSPF database on R1 i see the following:
Type-5 AS External Link States
Link ID ADV Router Age Seq# Checksum Tag
0.0.0.010.2.50.22359 0x80000005 0x0044F8 3489681806
10.0.0.010.2.50.2580 0x800000D0 0x00F6A4 0
10.2.50.010.2.50.2545 0x800000D0 0x00C89B 0
10.2.60.010.2.50.2510 0x800000D0 0x005AFF 0
10.2.70.010.2.50.2475 0x800000D0 0x00EB64 0
10.2.80.010.2.50.2440 0x800000D0 0x007DC8 0
10.20.0.010.2.50.22324 0x80000005 0x00D04E 3489681806
10.20.2.010.2.50.2406 0x800000D0 0x000280 0
10.44.50.010.2.50.22290 0x80000005 0x00874D 3489681806
10.44.60.010.2.50.22255 0x80000005 0x0019B1 3489681806
The routes marked with a tag of 3489681806 are routes that are originating from a router which is not directly attached, and it is these routes that are not populating the routing table.
I have googled this and it seems to be an MPLS PE-CE OSPF loop prevention method as linked here:
The thing is i want these routes in R1's table! R1 is running vrf-lite, and the corp RD is 20878:10, on the core routers the RD is 20878:101. I think im right in saying that the whole string should match for it to be recognised as a loop, not really sure what is going on
Anyone able to shed some light on this please?
No comments:
Post a Comment