You are previewing MPLS in the SDN Era.
O'Reilly logo
MPLS in the SDN Era

Book Description

How can you make multivendor services work smoothly on today’s complex networks? This practical book shows you how to deploy a large portfolio of multivendor Multiprotocol Label Switching (MPLS) services on networks, down to the configuration level. You’ll learn where Juniper Network's Junos, Cisco's IOS XR, and OpenContrail, interoperate and where they don’t.

Two network and cloud professionals from Juniper describe how MPLS technologies and applications have rapidly evolved through services and architectures such as Ethernet VPNs, Network Function Virtualization, Seamless MPLS, Egress Protection, External Path Computation, and more. This book contains no vendor bias or corporate messages, just solid information on how to get a multivendor network to function optimally.

Table of Contents

  1. Preface
    1. About This Book
      1. Interoperability
      2. MPLS in the SDN Era
      3. Live Book
    2. Contents of This Book
    3. Disclaimer
    4. Conventions Used in This Book
    5. Safari® Books Online
    6. How to Contact Us
    7. Acknowledgments
  2. 1. Introduction to MPLS and SDN
    1. The Internet
    2. ISP Example Topology
      1. Router Types in a Service Provider
      2. BGP Configuration
      3. BGP Route Signaling and Redundancy
      4. Packet Forwarding in a BGP-Less Core
    3. MPLS
      1. MPLS in Action
      2. The MPLS Header
      3. MPLS Configuration and Forwarding Plane
      4. Forwarding Equivalence Class
      5. Again, What Is MPLS?
    4. OpenFlow
      1. OpenFlow—Flow-Based Forwarding
      2. OpenFlow—Openness and P4
    5. SDN
      1. Separation of the Control and Forwarding Planes
      2. SDN and the Protocols
    6. The SDN Era
      1. SDN-Era Use Cases
  3. 2. The Four MPLS Builders
    1. LDP
      1. LDP Discovery and LDP Sessions
      2. LDP Label Mapping
      3. LDP and Equal-Cost Multipath
      4. LDP Implementation Details
      5. LDP Inter-Area
      6. Protecting LDP Networks from Traffic Blackholing
    2. RSVP-TE
      1. RSVP-TE LSP Fundamentals
      2. RSVP-TE in Action
      3. RSVP-Constrained Paths and ECMP
      4. Inter-Area RSVP-TE LSPs
      5. RSVP Auto Tunnel
    3. IGP and SPRING
      1. SPRING in Action
      2. SPRING Concepts
      3. SPRING Adjacency Segments
      4. A Comparison of LDP, RSVP-TE, and SPRING
    4. BGP-Labeled Unicast
      1. IGP-Free Large-Scale Data Centers
      2. BGP-LU Configuration
      3. Service Configuration in an IGP-Less Topology
      4. BGP-LU—Signaling and Forwarding Plane
      5. BGP-LU—SPRING Extensions
  4. 3. Layer 3 Unicast MPLS Services
    1. 6PE: IPv6 Transport in an IPv4/MPLS Core
      1. 6PE—Backbone Configuration at the PEs
      2. 6PE—RR Configuration
      3. 6PE—Access Configuration at the PEs
      4. 6PE—Signaling
      5. 6PE—Forwarding Plane
    2. BGP/MPLS IP Virtual Private Networks
      1. Attachment Circuits and Access Virtualization
      2. L3VPN in a Nutshell
      3. L3VPN—Signaling
      4. L3VPN—Forwarding Plane
      5. L3VPN—Backbone Configuration at the PEs
      6. L3VPN—RR Configuration
      7. L3VPN—VRF Configuration at the PEs
      8. L3VPN—Routing Tables in Junos
      9. L3VPN—Service Label Allocation
      10. L3VPN—Topologies
      11. L3VPN—Loop Avoidance
      12. Internet Access from a VRF
    3. Route Target Constraint
      1. RTC—Signaling
      2. RTC—RR Configuration
      3. RTC—PE Configuration
    4. Coupling MPLS Services to Transport Planes
      1. Configuring Several Loopbacks in the Default Instance
      2. Signaling LSPs to Different Loopback Addresses
      3. Changing the Service Routes’ BGP Next Hop
  5. 4. Internet Multicast Over MPLS
    1. IP Multicast
      1. IP Multicast Protocols
      2. IP Multicast Modes
    2. Classic Internet Multicast
      1. Starting Multicast Sources and Receivers
      2. Signaling the Multicast Tree
      3. Classic Internet Multicast—Connecting Multicast Islands Across the Core
    3. Signaling Join State Between Remote PEs
      1. Carrier IP Multicast Flavors
      2. Direct Inter-PE Model—PE-to-PE PIM Adjacencies over Unicast IP Tunnels
      3. Direct Inter-PE Model—PE-to-PE PIM Adjacencies over Multicast IP Tunnels
      4. Direct Inter-PE Model—PE-PE PIM Adjacencies over MPLS Label-Switched Paths
      5. Beyond the Direct Inter-PE Model—Not Establishing PE-PE PIM Adjacencies
    4. Internet Multicast over MPLS with In-Band Multipoint LDP Signaling
      1. Multipoint LDP
      2. In-Band Signaling
      3. Life of a C-Multicast Packet in an mLDP P2MP LSP
      4. CE Multihoming
      5. mLDP In-Band and PIM ASM
      6. Other Internet Multicast over MPLS Flavors
  6. 5. Multicast VPN
    1. BGP Multicast VPN with mLDP Transport
      1. MVPN Address Family
      2. Configuring BGP MVPN
      3. MVPN Site AD
      4. Signaling C-Multicast (S, G) Join State with BGP
      5. Signaling Provider Tunnels—BGP and the PMSI Attribute
      6. Signaling Provider Tunnels—Multipoint LDP for Transport
    2. BGP Multicast VPN with RSVP-TE P2MP Transport
      1. Advertising the Inclusive PMSI—RSVP-TE P2MP
      2. Advertising Selective PMSIs—RSVP-TE P2MP
      3. Signaling P- Tunnels with RSVP-TE P2MP
    3. BGP Multicast VPN with Ingress Replication
      1. Inclusive PMSI—IR
      2. Selective PMSI—IR
      3. BGP Multicast VPN with Other P- Tunnel Flavors
    4. CE Multihoming in BGP Multicast VPN
      1. Egress PE Redundancy
      2. Ingress PE Redundancy
      3. Choosing the Best RD Scheme
    5. BGP Multicast VPN with C-PIM ASM
      1. ASM Mode
      2. C-Rendezvous Point—PE and CE Configuration
      3. C-Multicast Signaling—ASM Mode with C-RP at the PEs
    6. Noncongruent C-Unicast and C-Multicast
  7. 6. Point-to-Point Layer 2 VPNs
    1. L2VPN in a Nutshell
      1. L2VPN Use Cases
      2. L2VPN Topological Classification
      3. L2VPN Signaling and Transport
      4. P2P L2VPN—Varied Access Technologies
      5. L2VPN Flavors Covered in This Book
    2. VPWS Signaled with BGP
      1. BGP L2VPN Address Family
      2. BGP VPWS Configuration at the PEs
      3. BGP VPWS Signaling
      4. L2VPN Forwarding Plane
      5. BGP VPWS—CE Multihoming to Several PEs
      6. Ethernet OAM (802.3ah, 802.1ag)
      7. BGP VPWS—VLAN Tag Multiplexing
      8. BGP VPWS—VLAN Tag Translation and Manipulation
      9. BGP VPWS—PW Head-End (PWHE)
      10. BGP VPWS—Load Balancing
    3. VPWS Signaled with LDP
      1. LDP VPWS Configuration at the PEs
      2. LDP VPWS Signaling and Forwarding Planes
      3. LDP VPWS—CE Multihoming and PW Redundancy
      4. LDP VPWS—VLAN Tag Multiplexing
      5. LDP VPWS—VLAN Tag Translation and Manipulation
      6. LDP VPWS—PWHE
      7. LDP VPWS—FAT
  8. 7. Virtual Private LAN Service
    1. Introduction to VPLS
    2. VPLS Signaled with BGP
      1. BGP VPLS Configuration
      2. BGP VPLS Signaling
      3. BGP VPLS—Efficient BUM Replication
    3. VPLS Signaled with LDP
      1. LDP VPLS Configuration
      2. LDP VPLS Signaling
      3. LDP VPLS—Autodiscovery via BGP
    4. VLANs and Learning Domains in VPLS
      1. VPLS in default VLAN mode
      2. Junos VPLS Instances—Normalized VLAN Mode
      3. Junos VPLS Instances—VLAN-Free Mode
      4. Junos VPLS Instances—VLAN-Aware Mode
      5. Junos Virtual Switches
    5. Integrated Routing and Bridging in VPLS
      1. IRB Configuration in Junos VPLS Instances
      2. IRB Configuration in Junos Virtual Switches
      3. IRB Configuration in IOS XR
      4. VPLS—IRB Redundancy and Traffic Tromboning
    6. Hierarchical VPLS
      1. H-VPLS Model with LDP Signaling
      2. H-VPLS Models with BGP for Autodiscovery and Signaling
  9. 8. Ethernet VPN
    1. EVPN with MPLS Transport
      1. EVPN Versus VPLS
      2. EVPN Implementations
      3. EVPN—This Book’s Topology
      4. BGP EVPN Address Family
      5. EVPN with MPLS Transport—Junos Configuration
      6. EVPN MPLS—Inclusive Tunnel and Autodiscovery
      7. EVPN with MPLS Transport—Advertising MACs
      8. EVPN with MPLS Transport—Intra-VLAN Bridging
      9. EVPN with MPLS Transport—Inter-VLAN Forwarding
      10. EVPN with MPLS Transport—All-Active Multihoming
    2. Ethernet VPN with VXLAN Transport
      1. Data Center Challenges
      2. VXLAN
      3. EVPN with VXLAN Transport—Motivation
      4. EVPN with VXLAN Transport—Forwarding Plane
      5. EVPN with VXLAN Transport—Junos Configuration
      6. EVPN with VXLAN Transport—Signaling
    3. Provider Backbone Bridging EVPN
      1. Introduction to PBB
      2. PBB EVPN in a Nutshell
      3. PBB EVPN Implementations
      4. PBB EVPN in Action
      5. PBB EVPN Configuration
      6. PBB EVPN Signaling
  10. 9. Inter-Domain MPLS Services
    1. Inter-Domain Architectures
      1. This Chapter’s Example Topology
    2. Inter-AS Flavors
    3. Inter-AS Option A
    4. Inter-AS Option B
      1. Inter-AS Option B—Signaling and Forwarding
      2. Inter-AS Option B—Junos Configuration
      3. Inter-AS Option B—IOS XR Configuration
      4. Inter-AS Option B with Local VRF
    5. Inter-AS Option C
      1. BGP Sessions in Inter-AS Option C
      2. Inter-AS Option C—Signaling and Forwarding
      3. Inter-AS Option C—Configuration
    6. Carrier Supporting Carrier
    7. Inter-Domain RSVP-TE LSPs
  11. 10. Underlay and Overlay Architectures
    1. Overlays and Underlays
      1. Overlay and Underlay Are Relative Concepts
      2. Other Fundamental Concepts
    2. Multiforwarder Network Devices
      1. Single-Chassis Network Devices—Forwarding Plane
      2. Single-Chassis Network Devices—Control Plane
    3. Multichassis Network Devices
    4. Legacy Data Center Networking
      1. The Challenges of L2 Bridged Networks
      2. Underlays in Modern Data Centers
      3. Overlays in Modern Data Centers
    5. Data Center Underlays—Fabrics
      1. IP Fabrics—Forwarding Plane
      2. IP Fabrics with Distributed-Only Control Plane
      3. IP Fabrics with Hybrid Control Plane
    6. Network Virtualization Overlay
      1. Compute Controllers
      2. Virtual Network Controllers
      3. NVO—Transport of Control Packets
      4. NVO—Agents
  12. 11. Network Virtualization Overlays
    1. OpenContrail in a Nutshell
      1. OpenContrail Controllers
      2. Compute, Gateway, and Service Nodes
    2. Case Study: A Private Cloud
      1. vRouter-VM Link Addressing
      2. Initializing vNICs—XMPP as a DHCP-Like Protocol
      3. Interconnecting VMs—XMPP as a BGP-Like Protocol
      4. Interconnecting Subscribers to Cloud VMs
    3. Communication Between Virtual Networks
    4. Network Virtualization Overlay: L2_L3 Mode
      1. VXLAN Refresher
      2. Intrasubnet (L2) and Intersubnet (L3) Traffic
      3. Interconnecting VMs—IntraSubnet Traffic with VXLAN
      4. vRouter and Gateway Nodes—L2_L3 Mode
    5. Integrating Legacy L2 World into the NVO
      1. L2 Gateways and OVSDB
      2. ToR Service Nodes
      3. Binding a Bare-Metal Server to the Overlay
      4. MAC Learning with OVSDB
      5. Bare-Metal Servers and OVSDB—the Forwarding Plane
  13. 12. Network Function Virtualization
    1. NFV in the Software-Defined Networking Era
      1. Virtual or Physical?
      2. Applicability of NFV to Service Providers
    2. NFV Practical Use Case
    3. NFV Forwarding Plane
    4. NFV—VRF Layout Models
      1. Legacy VRF Layout—Transit VN Model
      2. Modern VRF Layout—Two-VN Model
    5. NFV—Long Version of the Life of a Packet
    6. NFV Control Plane
    7. NFV Scaling and Redundancy
      1. NFV Scaling and Redundancy—Load Balancing
    8. Service Instance Flavors
      1. In-Network Service Instances
      2. In-Network-NAT Service Instances
      3. Transparent Service Instances
      4. Network Service Function Outside a VM or Container
  14. 13. Introduction to Traffic Engineering
    1. TE Protocols
      1. TE LSP Types
    2. TE Information Distribution
      1. TE Distribution via OSPF
      2. TE Distribution via IS-IS
      3. The TED
    3. TE Static Constraints
      1. TE Metric
      2. Link Coloring—Administrative Group
      3. Extended Administrative Groups
      4. Shared Risk Link Group
    4. Egress Peer Engineering
      1. EPE Based on BGP-LU
  15. 14. TE Bandwidth Reservations
    1. TE Static Bandwidth Constraints
      1. TE Bandwidth Attributes
      2. Default TE Interface Bandwidth
      3. Basic RSVP-TE Bandwidth Reservation
      4. LSP Priorities and Preemption
      5. Traffic Metering and Policing
    2. TE Auto-Bandwidth
      1. Introduction to Auto-Bandwidth
      2. Auto-Bandwidth in Action
      3. Auto-Bandwidth Configuration
      4. Auto-Bandwidth Deployment Considerations
    3. Dynamic Ingress LSP Splitting/Merging
      1. Dynamic Ingress LSP Splitting/Merging—Configuration
      2. Dynamic Ingress LSP Splitting/Merging in Action
  16. 15. Centralized Traffic Engineering
    1. BGP Link-State
    2. PCEP
      1. PCE Implementations
      2. Interaction Between PCE and PCC
      3. PCE-Initiated RSVP-TE LSPs
      4. PCC-Initiated RSVP-TE LSPs
    3. PCC Label-Switched Path Signaling
      1. RSVP-TE LSPs
      2. SPRING (IGP) TE LSPs
      3. BGP LSPs
    4. PCC Configuration
      1. PCC Templates for PCE-Initiated LSPs
      2. Delegating PCC-Initiated LSPs to the PCE
    5. PCE Use Cases
      1. Extending the Link Attributes Palette
      2. Enhanced LSP Preemption Logic
      3. Diverse Paths
  17. 16. Scaling MPLS Transport and Seamless MPLS
    1. Scaling an IGP Domain
      1. Scaling an IGP—OSPF
      2. Scaling an IGP—IS-IS
      3. Scaling an IGP—MPLS Protocols
    2. Scaling RSVP-TE
      1. RSVP-TE Protocol Best Practices
    3. Intradomain LSP Hierarchy
      1. Tunneling RSVP-TE LSPs Inside RSVP-TE LSPs
      2. Tunneling LDP LSPs Inside RSVP-TE LSPs
      3. Tunneling SPRING LSPs Inside RSVP-TE LSPs
    4. Interdomain Transport Scaling
      1. Nonhierarchical Interdomain Tunnels
      2. Hierarchical Interdomain Tunnels (Seamless MPLS)
    5. IGP-Less Transport Scaling
      1. BGP-LU Hierarchy
      2. MPLS-Capable Servers and Static Labels
  18. 17. Scaling MPLS Services
    1. Hierarchical L3VPN
      1. Default Route L3VPN Model
      2. Default Route with Local Routes L3VPN Model
      3. Pseudowire Head-End Termination L3VPN Model
  19. 18. Transit Fast Restoration Based on the IGP
    1. Fast Restoration Concepts
      1. Ingress/Transit/Egress Transport Protection Concepts
      2. Global Repair Concepts
      3. Local Repair Concepts
    2. Loop-Free Alternates
      1. Per-Link LFA
      2. Per-Prefix LFA
    3. Extending LFA Backup Coverage
      1. LFA with LDP Backup Tunnels (Remote LFA)
      2. RLFA with RSVP-TE Backup Tunnels
      3. Topology Independent Fast ReRoute
      4. Modifying the default LFA selection algorithm
      5. Topology-Independent LFA
    4. Maximally Redundant Trees
  20. 19. Transit Fast Restoration Based on RSVP-TE
    1. RSVP-TE Path Protection
    2. RSVP-TE Facility (Node-Link) Protection
      1. Manual Link Protection Bypass
      2. Manual Node-Link Protection Bypass
      3. Facility Protection in Action
      4. Automatic Protection Bypass
    3. RSVP-TE One-to-One Protection
    4. Transit Fast-Restoration Summary
  21. 20. FIB Optimization for Fast Restoration
    1. Next-Hop Hierarchy
      1. Topology used in Chapter 20 and in Chapter 21
        1. Flat Next-Hop Structures
          1. Indirect Next Hop (Junos)
      2. Chained Composite Next Hop (Junos)
      3. BGP PIC Core (IOS XR)
    2. Preinstalled Next Hops to Multiple Egress PEs (PIC Edge)
      1. Active/Standby Next Hops to Egress PEs
      2. Active/Active Next Hops to Egress PEs
      3. BGP Best External Failover
  22. 21. Egress Service Fast Restoration
    1. Service Mirroring Protection Concepts
    2. Combined Protector/Backup Egress PE Model
    3. Separate (Centralized) Protector and Backup Egress PE Model
    4. Context-ID Advertisement Methods
      1. Stub-Alias
      2. Stub-Proxy
    5. L3VPN PE→CE Egress Link Protection
    6. Layer 2 VPN Service Mirroring
      1. BGP-Based L2VPN Service Mirroring
      2. LDP-Based L2VPN Service Mirroring
    7. Egress Peer Engineering Protection
    8. Protection in Seamless MPLS Architecture
      1. Border Link (ASBR-ASBR) Protection
      2. Border Node (ABR or ASBR) Protection
    9. Summary
  23. Index