I have four routers Ra and Rb in one location, Rc and Rd in another location. Ra and Rb are connected by UTP (crossed-cabled), Rc and Rd are also connected by UTP (crossed-cabled), Ra and Rd are connected by ISDN-PRI (direct connection), now the problem is Rb and Rc they are connected via MPLS Provider which the distand of Rc are 5 hops away from Rb.
1. Can I use IP Tunnel for iBGP between Rb and Rc?
2. Do I need to fully meshed the iBGP between four routers as what Cisco suggest?
3. Thus the fully meshed needs to be physical? Can I use logical (IP Tunnel)?
4. Please point me to a lot of example with explanations for BGP MED, BGP neighbour default-originate, BGP route-maps, BGP regular expressions. NOTE: I've been to BGP Case Studies, I need more samples.
Are you using an MPLS VPN between Rb and Rc ? If yes, what routing protocol are you using - RIP, ospf, static ? Either way all networks at Rc should be available at Rb and vice versa, as BGP route (because networks at Rc (and vice versa) are redistributed through Multiprotocol BGP via the provider.
If you need to run an IBGP session between Rb and Rc, you shouldnt be needing a tunnel. You can directly configure them. Using a logical interface such as loopback interface for IBGP peering would be the best way to implement IBGP in this case, because you have redudant connection between the two locations.
int lo 0
ip add 18.104.22.168 255.255.255.255
router bgp 10
neighbor 22.214.171.124 remote-as 20
neighbor 126.96.36.199 update-source lo 0
int lo 1
ip add 188.8.131.52 255.255.255.255
router bgp 20
neighbor 184.108.40.206 remote-as 10
neighbor 220.127.116.11 update-source lo 0
you should include the loopbacks 18.104.22.168 and 22.214.171.124 into the routing protocol you are running between the sites, or just add a static route on each router for reachability.
If you do not use a full mesh IBGP, then eBGP routes wont be exchanged from first neighbor to third neighbor ( say Ra gives a BGP route to Rb, Rb wont give the same route to Rc, if there is no iBGP session between Ra and Rc)
Since you have only 4 routers, it shouldnt be a problem to implement a full mesh iBGP session (6 sessions needed). If you plan to grow, or have more routers, shifting to a Route reflector or Confederation would be the best method. That way you can save the number of peerings needed.
Full mesh neednt be physical. It can be logical session too. Thats why I suggested to use loopback interfaces for peering. If link between Rb and Rc goes down, still the loopback interface on Rc and Ra is up, route to the loopback network will reconverge through routing protocol and the BGP connection will still remain up.
To read more on BGP, you might want to check out Cisco Press Books like, Internet Routing Architecture from Bassam Halabi, Routing TCP/IP part 2, from Jeff Doyle, etc. Also related to MPLS VPN and Multiprotocol BGP, you might want to read the related MPLS book from Sybex publications. I would recommend reading this book, coz the author has wonderfully described the concept behind MPLS.
Thanks for your help! You have answered most of them but the major problem is the connection between Rb and Rc, I'm not running MPLS VPN between the two router. I'm connecting them to an MPLS VPN provider thats why the two routers are 5 hops away with each other. I'm running OSPF as IGP, we intend to configure in such a way that the primary link between Ra/Rb to Rc/Rd is the MPLS VPN. The ISDN-PRI is a backup (OSPF link on-demand circuit). I'm using loopback interfaces in all my iBGP sessions.
Ok, in this case, you dont need to run an IP tunnel between both routers. Enable OSPF on the serial interfaces connecting to the ISP on both routers with OSPF. Even though your routers Rb and Rc might be 4 or 5 routers apart ( ISP routers in between) , these routers are only running MPLS, and hence they will switch all packets coming from the customer (you).
Configure two loopback interfaces, one each on Rb and Rc, as per my previous configuration, and include them in OSPF. Now configure iBGP between both routers, using the loopback interface.
For ibgp multihop is not needed. Its only needed in case of EBGP.
You might also want to turn synchronization off, in case you are running Full mesh IBGP.
Thanks. I will try it and if theres a problem I will post again here.
Upstream ISP is ISP1 and ISP 2 (AS 1 and AS2)
Ra and Rc is my router (AS3)
Rx and Ry is cutomer router (AS4)
Ignore the wwwww i put it there for drawing purposes.
Local traffic coming from ISP1 (my country local traffic)
International traffic coming from ISP2 (outside my country traffic)
Customer want the incoming local traffic to flow via link-a and incoming international traffic to flow via link-b. If link-a is down both traffic should flow to link-b, same if link-b is down both traffic should flow to link-a.
1. Can I enfluence my outgoing traffic to Customer from my upstream ISP? How to do it?
Here is some of what you requested. I have a similar situation. I have 1 router in location A and 3 in location B. I have a T1 between locations. I run EIGRP as the inside routing protocol and BGP outside. To keep BGP connected incase the site to site T1 fails. I have tunnels from the router in location A to all 3 routers in location B. If my T1 fails all traffic is redirected through the tunnels and no failure is noticed by end users.
We are pleased to announce availability of Beta software for 16.6.3.
16.6.3 will be the second rebuild on the 16.6 release train targeted
towards Catalyst 9500/9400/9300/3850/3650 switching platforms. We are
looking for early feedback from customers befor...
Introduction Featured Speakers Luis Espejel is the Telecommunications
Manager of IENova, an Oil & Gas company. Currently he works with Cisco
IOS® and Cisco IOS XE platforms, and NX to some extent. He has also
worked as a Senior Engineer with the Routing P...
In this session you can learn more about Layer 3 multicast and the best
practices to identify possible threats and take security measures. It
provides an overview of basic multicast, the best security practices for
use of this technology, and recommendati...