[Xorp-users] Routes redistribution between OSPF and BGP
Atanu Ghosh
atanu at ICSI.Berkeley.EDU
Mon Mar 31 19:15:16 PDT 2008
Hi,
When redistributing routes from OSPF to BGP only the routes that OSPF
has learned from its neighbours will be redistributed. The connected
routes will not be redistributed, so you are correct you also need to
redistribute connected.
Atanu.
>>>>> "Yue" == Yue Li <yueli.m at gmail.com> writes:
Yue> Hi, Atanu
Yue> I have a topology like below. I run OSPF on all routers and BGP on R2,
Yue> R5, which forms a eBGP peering.
Yue> The problem is when I redistribute OSPF into BGP, I can only see the
Yue> routes to "net13" and "net34" on R5.
Yue> I wonder whether I should export "connected" to BGP, so I may see
Yue> "net12" and "net24" on R5.
Yue> net12
Yue> (R1)---------(R2)------------(R5)----------(R6)
Yue> net13 | | | |
Yue> | AS1 | net24 | AS2 |
Yue> (R3)---------(R4) (R7)-----------(R8)
Yue> net34
Yue> The config of R2 is:
Yue> interfaces {
Yue> interface tun0 {
Yue> disable: false
Yue> vif tun0 {
Yue> disable: false
Yue> address 10.0.0.1 {
Yue> prefix-length: 30
Yue> disable: false
Yue> }
Yue> }
Yue> }
Yue> interface tun1 {
Yue> disable: false
Yue> vif tun1 {
Yue> disable: false
Yue> address 10.0.0.17 {
Yue> prefix-length: 30
Yue> disable: false
Yue> }
Yue> }
Yue> }
Yue> interface tun2 {
Yue> disable: false
Yue> vif tun2 {
Yue> disable: false
Yue> address 10.0.0.65 {
Yue> prefix-length: 30
Yue> disable: false
Yue> }
Yue> }
Yue> }
Yue> }
Yue> policy {
Yue> policy-statement connected {
Yue> term export {
Yue> from {
Yue> protocol: "connected"
Yue> }
Yue> }
Yue> }
Yue> policy-statement ospf-to-bgp {
Yue> term ospf2bgp {
Yue> from {
Yue> protocol: "ospf4"
Yue> }
Yue> }
Yue> }
Yue> }
Yue> protocols {
Yue> ospf4 {
Yue> router-id: 10.0.0.1
Yue> export: "connected"
Yue> area 0.0.0.0 {
Yue> interface tun0 {
Yue> link-type: "p2p"
Yue> vif tun0 {
Yue> address 10.0.0.1 {
Yue> hello-interval: 10
Yue> interface-cost: 100
Yue> neighbor 10.0.0.2 {
Yue> router-id: 10.0.0.2
Yue> }
Yue> }
Yue> }
Yue> }
Yue> interface tun1 {
Yue> link-type: "p2p"
Yue> vif tun1 {
Yue> address 10.0.0.17 {
Yue> hello-interval: 10
Yue> interface-cost: 100
Yue> neighbor 10.0.0.18 {
Yue> router-id: 10.0.0.14
Yue> }
Yue> }
Yue> }
Yue> }
Yue> }
Yue> }
Yue> bgp {
Yue> bgp-id: 10.0.0.1
Yue> export: "ospf-to-bgp"
Yue> local-as: 65010
Yue> peer 10.0.0.66 {
Yue> local-ip: 10.0.0.65
Yue> as: 65011
Yue> next-hop: 10.0.0.65
Yue> ipv4-multicast: false
Yue> ipv6-unicast: false
Yue> ipv6-multicast: false
Yue> }
Yue> }
Yue> }
Yue> On Mon, Mar 24, 2008 at 12:20 PM, Atanu Ghosh <atanu at icsi.berkeley.edu> wrote:
>> Hi,
>>
>> All the networks that you wish to advertise to an external AS need to be
>> redistributed into BGP. You therefore need to redistribute OSPF into
>> BGP.
>>
>> Atanu.
>>
>> >>>>> "Yue" == Yue Li <yueli.m at gmail.com> writes:
>>
Yue> Hi, All I have several XORP routers which form 2 Autonomous
Yue> Systems. OSPFv2 is used for intra-AS protocol and BGP is
Yue> running on the AS border routers.
>>
Yue> I found no BGP routes when monitoring BGP behaviors, while
Yue> their peering relations are correct. And routers in different
Yue> AS can't ping each other.
>>
Yue> So I guess I need to add routes redistribution to the
Yue> configuration of the AS border routers, right? Do I need to
Yue> add ospf-to-bgp policy, or bgp-to-ospf policy, or both of them?
>>
Yue> Any help would be greatly appreciated.
>>
Yue> _______________________________________________ Xorp-users
Yue> mailing list Xorp-users at xorp.org
Yue> http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users
>>
More information about the Xorp-users
mailing list