MPLS in the SDN Era

This blog keeps the book MPLS in the SDN Era alive between its first and its (hopefully) second edition.

Check the configuration repository in this page, it keeps growing!


19 thoughts on “MPLS in the SDN Era

  1. “In this example, CEs and BRs advertise their own loopacks to one single eBGP peer:

    CE1 advertises to PE1 only.

    CE2 advertises to PE2 only.”

    Figure 1-4 has eBGP Single-Hop Sessions between CE1 & PE2, and CE2 and PE1, so I’m a bit confused.


  2. Hi there,

    Good point !! As far as I have found it is only available for ISIS.

    Slide 82 (2015) of the following ppt
    says in fact the following:
    XE: OSPF prefix suppression
    XR/XE: ISIS advertise passive-only

    I guess when you guys mentioned the OSPF lack of flexibility you meant the fact that suppressing a prefix in ISIS is easier to enforce than in OSPF as it involves manipulating prefixes within a TLVs as opposed to manipulating Link-types within LSA1 and LSA2 ?


  3. We have been looking at Junos and IOS XR only. And neither in Junos, nor in IOS XR we didn’t found the way to suppress link prefixes for OSPF. Therefore the book states lack of flexibility for OSPF (we didn’t checked any other OSes, like IOS, for example). For ISIS it is possible to create route-policy, to specify what should and what shouldn’t be advertised. So, suppression of link prefixes is possible.


  4. First of all I want to say you guys did a great job with writing this book.
    I have a question and hopefully you can help me with that.
    I am interested in the configuration snippets of chapter 8. Specially the combination of a a MC-LAG configuration on Juniper PE nodes with an EVPN configuration. It would be great if this can be shared


  5. Another errata: on p.132, ex.3-9 there’s output from “show bgp ipv6 unicast neighbors routes” (or similar) rather than “show bgp ipv6 unicast fc00::10:2:34:0/112” as stated in the book.
    Correct me if I’m wrong.
    The book is absolutely great, BTW, thank you.


  6. Hy all,
    first of all I have to say that MPLS in SDN Era is a great book, because it joins both theory and both practical examples.

    I have a couple of technical questions that I submit you in the following:

    Chapter 2 Section BGP LU in Data Center.
    At page 108 is written:


    I don’t udenstand why we must assure to not readvisertise labeled routes as unlabeled

    Can you provide an example to better understand what would happen in IOS XR in this case?

    –Question#2– (to get confirm of my understanding)
    Chapter 2 Section BGP LU in Data Center.
    Page 115 -(Juniper MPLS enabled Server device)

    In S1 we must set up the two vanilla BGPv4 sessions towards the RRs/Spines.

    In order to let Srv nodes to do that, Srvs nodes must known how to reach them.
    For this reason we perform a local copy, thanks to rib group, of the Spines LB from inet.3 towards inet.0.

    My question is:

    in the output of the following commands:

    juniper@Srv1> show route detail

    There will be as next-hop for the label towards the leaf L1? (learned from the inet.3)


  7. At page 108 is written:

    This careful community scheme is due to the fact that IOS XR keeps labeled and
    unlabeled IP routes in the same global table, so it is important not to readvertise
    labeled routes as unlabeled routes, or vice versa. Said differently, you need to pay spe‐
    cial attention so that the SAFI=1 and SAFI=4 worlds remain independent.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s