Multiple Processes EIGRP/Multiple Loopbacks

Bl8ckr0uterBl8ckr0uter Inactive Imported Users Posts: 5,031 ■■■■■■■■□□
Greetings Netwokers:

Can someone explain me a practical (real world) reason for having multiple EIGRP Processes running on at the same time on a router? Is it for Network Segmentation (such as multi-area OSPF) or is there some other benefit? I am gearing up for the CCNA on Saturday and I cannot find the answer to this question.


In that same regard, why would you have multiple loopbacks and is it typical to advertise your loopback addresses in your routing protocols? I do this so that I can change the addressing scheme of my actual interfaces and still telnet without an issue. Does anyone else do this (at work or in the lab)?

Thanks all and wish me luck

Comments

  • networker050184networker050184 Mod Posts: 11,962 Mod
    You can use multiple instances of EIGRP if you do not want them to share routes between each other. I'll usually use a separate routing protocol to keep it simple though, but sometimes you don't have a choice really.

    As far as loopbacks, you just use as many as you need. Most people will have at least one "main" loopback to use as the router id and a single IP address on the device that never changes. You can also use loopbacks for other things like addresses for GRE, an address within a VRF for reachability testing, an interface to NAT through, the possibilities are endless. It is typical to have at least your "main" loopback advertised into your IGP so you can reach it for testing, next hop for routing update etc.
    An expert is a man who has made all the mistakes which can be made.
  • Bl8ckr0uterBl8ckr0uter Inactive Imported Users Posts: 5,031 ■■■■■■■■□□
    You can use multiple instances of EIGRP if you do not want them to share routes between each other. I'll usually use a separate routing protocol to keep it simple though, but sometimes you don't have a choice really.

    Let me ask you a question (I am going to test this later but I am going to ask you first). Suppose you have a simple 3 router network and the are all connected back to back

    R1>><<<R2>><<R3>
    Now say on router 1 I have am using eigrp process 1 and on router 3 I am using eigrp process 2. On router 2 I am using both eigrp process 1 and 2. Would router 2 have routers from processes 1 and 2 in its routing table? Would it differentiate between the processes? In router 1 the routes from router 2, would they have a higher AD (like external EIGRP)?
  • networker050184networker050184 Mod Posts: 11,962 Mod
    Yes, router 2 would have routes from both, but it would not send them to the other process automatically, so router 1 and router 3 would not have the routes from the other EIGRP process.

    You can make the router share the routes between the processes by redistribution, but that isn't really covered in the NA material.
    An expert is a man who has made all the mistakes which can be made.
  • Bl8ckr0uterBl8ckr0uter Inactive Imported Users Posts: 5,031 ■■■■■■■■□□
    Yes, router 2 would have routes from both, but it would not send them to the other process automatically, so router 1 and router 3 would not have the routes from the other EIGRP process.
    Cool. That is what I figured.

    You can make the router share the routes between the processes by redistribution, but that isn't really covered in the NA material.

    I think the word redistribution is written in my ccna book somewhere.....icon_rolleyes.gif

    In Network Warrior however it goes into much greater detail on the subject. icon_eek.gif
  • SysAdmin4066SysAdmin4066 Member Posts: 443
    You dont really start delving deeply into redistribution till the BSCI of the CCNP, and really heavily in the CCIE from what I understand. It's a hell of a concept, but really cool once you get the hang of it. But everything he said is correct, EIGRP processes are treated as seperate routing instances, so it would be like having OSPF and EIGRP running together. They wouldnt automatically share routing info with each other, you would have to redistribute between the two.
    In Progress: CCIE R&S Written Scheduled July 17th (Tentative)

    Next Up: CCIE R&S Lab
  • Daniel333Daniel333 Member Posts: 2,077 ■■■■■■□□□□
    Hm.. real world reasons to do this... all I can think of would be say if you are transitioning your network. Like you have one setup now, and are movig to a new design, but during this process you need to keep the old network running.
    -Daniel
  • jason_lundejason_lunde Member Posts: 567
    Once you start getting into MPLS VPNs you will see multiple eigrp instances running, mainly using vrf's. Actually the ISCW lab portfolio has a MPLS vpn challenge lab that uses multiple eigrp instances if you want to see it in action. I dont know, I may just be stoked from that lab....it was one of the coolest things I have done, though it took a while to wrap my head around everything that was going on there. HTH's
  • Bl8ckr0uterBl8ckr0uter Inactive Imported Users Posts: 5,031 ■■■■■■■■□□
    You dont really start delving deeply into redistribution till the BSCI of the CCNP, and really heavily in the CCIE from what I understand. It's a hell of a concept, but really cool once you get the hang of it. But everything he said is correct, EIGRP processes are treated as seperate routing instances, so it would be like having OSPF and EIGRP running together. They wouldnt automatically share routing info with each other, you would have to redistribute between the two.

    You are not lying. I was trying to redistribute EIGRP into OSPF in Packet Tracer and either I was failing or PT was failing. That was like a week ago and I haven't tried it yet. Hopefully after today when I get the CCNA I will read network Warrior Cover to Cover along with the Cisco IOS Cookbook to learn the Hows and Why.
  • Bl8ckr0uterBl8ckr0uter Inactive Imported Users Posts: 5,031 ■■■■■■■■□□
    Daniel333 wrote: »
    Hm.. real world reasons to do this... all I can think of would be say if you are transitioning your network. Like you have one setup now, and are movig to a new design, but during this process you need to keep the old network running.

    That seems logical. I was wondering if this was used as a form of security. That could be wrong because I just found out you can authenticate with eigrp


    EIGRP Message Authentication Configuration Example - Cisco Systems
  • SysAdmin4066SysAdmin4066 Member Posts: 443
    Jason is right as well, there are labs in the ISCW tract that show multiple instances of EIGRP. They are centered around VPNs. You might not want your VPN routes to mix with your internal routes. In that case, you would use one AS for your internal and a separate one for your tunnel.
    In Progress: CCIE R&S Written Scheduled July 17th (Tentative)

    Next Up: CCIE R&S Lab
  • ColbyGColbyG Member Posts: 1,264
    When running EIGRP between the PE and CE routers in an MPLS VPN you're not necessarily running different EIGRP processes, at least not in the way that the original question was posed. Here's an example:
    router eigrp 1
     address-family ipv4 vrf RED
     network 172.16.0.0 0.0.255.255
     redistribute bgp 10 metric 10000 100 255 1 1500
     autonomous-system 101
     exit-address-family
    !
    address-family ipv4 vrf BLUE
     network 192.168.0.0 0.0.255.255
     redistribute bgp 10 metric 10000 100 255 1 1500
     autonomous-system 201
     exit-address-family
    

    You're still using the same EIGRP process, but different ASes for the VPNs.
Sign In or Register to comment.