Re: [c-nsp] load balance between to EBGP peers
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Arie, Is maximum-path eibgp a good command for this? For multipath eibgp paths to be installed AS-PATH should be the same (not only the length of path vector - [http://www.cisco.com/en/US/tech/tk365/technologies_tech_note09186a0080094431.shtml#bgpmpath]). On OP's scenario he got INTERNET into vrf so AS-PATHs are not the same almost always (unless he is dual homed to the same provider) Am I understanding this correctly? Best Regards, - -mat - -- pgp-key 0x1C655CAB -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) iD8DBQFJMA9h+BuaDRxlXKsRAilvAJ9j4Yg6UdG4k5tX4SU3+92aG+vPSgCdH+m+ 1idFt613Qx4aNN8a/S7OwH0= =06Ty -END PGP SIGNATURE- ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
[c-nsp] load balance between to EBGP peers
Hi all , i have MPLS/VPN network based on 7600's . all the Internet is going via VRF-INTERNET , my ASBR gets full route from 2 peers via OC12 POS interfaces on VRF INTERNET , also he gets 0.0.0.0 route from both peers but of course only 1 gets to FIB . my target is to get load balance between the 2 peers(on the ASBR) in the outgoing traffic and to redistribute only 0.0.0.0 to VRF INTERNET to all the IBGP peers. what i see now is that the 0.0.0.0 gets label in the FIB of the ASBR pointing on one of the POS interface, so the other POS stay empty , with the show ip route vrf INTERNET 0.0.0.0 0.0.0.0 command i see the LB0 of the ASBR which is OK but the MPLS label table on the VRF INTERNET is pointing on the POS interface . i just want that the ASBR will route the traffic according to his full route table . so actually now i redistribute to all PE's /20 subnets for getting load balance , does someone have a good ides for this one ? regards moshe ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
Re: [c-nsp] load balance between to EBGP peers
Moshe, Take a look at these command: http://www.cisco.com/en/US/docs/ios/iproute/command/reference/irp_bgp3.h tml#wp1012317 Arie -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of moshe mizrachi Sent: Thursday, November 27, 2008 15:27 PM To: cisco-nsp@puck.nether.net Subject: [c-nsp] load balance between to EBGP peers Hi all , i have MPLS/VPN network based on 7600's . all the Internet is going via VRF-INTERNET , my ASBR gets full route from 2 peers via OC12 POS interfaces on VRF INTERNET , also he gets 0.0.0.0 route from both peers but of course only 1 gets to FIB . my target is to get load balance between the 2 peers(on the ASBR) in the outgoing traffic and to redistribute only 0.0.0.0 to VRF INTERNET to all the IBGP peers. what i see now is that the 0.0.0.0 gets label in the FIB of the ASBR pointing on one of the POS interface, so the other POS stay empty , with the show ip route vrf INTERNET 0.0.0.0 0.0.0.0 command i see the LB0 of the ASBR which is OK but the MPLS label table on the VRF INTERNET is pointing on the POS interface . i just want that the ASBR will route the traffic according to his full route table . so actually now i redistribute to all PE's /20 subnets for getting load balance , does someone have a good ides for this one ? regards moshe ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/ ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
Re: [c-nsp] load balance between to EBGP peers
You don't need the default route from your providers, just the full table. Inside your VRF, originate the default route locally from you ASBR and redistribute it to the rest of your routers (I'm assuming you are not sending the full table to the rest of the routers due to several limitations you may have there). To balance between your providers use any BGP attribute that may suit the purpose. -Original Message- From: [EMAIL PROTECTED] [mailto:cisco-nsp- [EMAIL PROTECTED] On Behalf Of moshe mizrachi Sent: Thursday, November 27, 2008 11:27 AM To: cisco-nsp@puck.nether.net Subject: [c-nsp] load balance between to EBGP peers Hi all , i have MPLS/VPN network based on 7600's . all the Internet is going via VRF-INTERNET , my ASBR gets full route from 2 peers via OC12 POS interfaces on VRF INTERNET , also he gets 0.0.0.0 route from both peers but of course only 1 gets to FIB . my target is to get load balance between the 2 peers(on the ASBR) in the outgoing traffic and to redistribute only 0.0.0.0 to VRF INTERNET to all the IBGP peers. what i see now is that the 0.0.0.0 gets label in the FIB of the ASBR pointing on one of the POS interface, so the other POS stay empty , with the show ip route vrf INTERNET 0.0.0.0 0.0.0.0 command i see the LB0 of the ASBR which is OK but the MPLS label table on the VRF INTERNET is pointing on the POS interface . i just want that the ASBR will route the traffic according to his full route table . so actually now i redistribute to all PE's /20 subnets for getting load balance , does someone have a good ides for this one ? regards moshe ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/ ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
Re: [c-nsp] load balance between to EBGP peers
On Thu, 2008-11-27 at 15:26 +0200, moshe mizrachi wrote: i have MPLS/VPN network based on 7600's . all the Internet is going via VRF-INTERNET , my ASBR gets full route from 2 peers via OC12 POS interfaces on VRF INTERNET , also he gets 0.0.0.0 route from both peers but of course only 1 gets to FIB . my target is to get load balance between the 2 peers(on the ASBR) in the outgoing traffic and to redistribute only 0.0.0.0 to VRF INTERNET to all the IBGP peers. what i see now is that the 0.0.0.0 gets label in the FIB of the ASBR pointing on one of the POS interface, so the other POS stay empty , with the show ip route vrf INTERNET 0.0.0.0 0.0.0.0 command i see the LB0 of the ASBR which is OK but the MPLS label table on the VRF INTERNET is pointing on the POS interface . i just want that the ASBR will route the traffic according to his full route table . so actually now i redistribute to all PE's /20 subnets for getting load balance , does someone have a good ides for this one ? As I understand your description, the box has a specific forwarding adjacency for the traffic received on the 0/0 LSP from the core. To make it use the routing table, you need it to perform a routing lookup, which (AFAIK) an aggregate label will do for you. Haven't tested this, but if you stop redistributing a default received from upstream, and instead add a static default pointing to Null0 in the VRF, you should get an aggregate label. This might result in lower performance, since the router needs to a full IP lookup. Regards, Peter ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/