Chapter 6: Advanced Features
Discover advanced MPLS capabilities including Traffic Engineering for optimal path selection, QoS implementation, Fast Reroute for rapid failover, Segment Routing for modern networks, and Anycast RP for multicast optimization.
Traffic Engineering
MPLS Traffic Engineering (TE) enables explicit path control, allowing network operators to optimize bandwidth utilization and provide guaranteed service levels through constraint-based routing.
TE Benefits
Traffic Engineering provides path optimization, bandwidth guarantees, congestion avoidance, and improved network utilization through intelligent path selection.
RSVP-TE Signaling
- Bandwidth reservation
- Explicit path specification
- Admission control
- Constraint-based routing
- Loop-free paths
OSPF-TE Extensions
- Traffic Engineering Database (TED)
- Link bandwidth advertisement
- Administrative weights
- Resource availability
- Constraint-based SPF
TE Component | Function | Protocol | Information Distributed |
---|---|---|---|
Information Distribution | Advertise link attributes | OSPF-TE, IS-IS-TE | Bandwidth, administrative weight |
Path Calculation | Compute optimal paths | CSPF algorithm | Constrained shortest path |
Path Signaling | Establish TE tunnels | RSVP-TE | Bandwidth reservation, labels |
Forwarding | Traffic steering | MPLS forwarding | Label-switched paths |
MPLS QoS
MPLS Quality of Service provides differentiated treatment of traffic classes, ensuring service level guarantees through traffic classification, marking, queuing, and scheduling mechanisms.
EXP Bit Marking
3-bit EXP field in MPLS header for QoS marking
- 0: Best Effort
- 5: Expedited Forwarding (EF)
- 3,4: Assured Forwarding (AF)
- 6,7: Network Control
Traffic Classification
Identify and mark traffic at ingress
- DSCP-to-EXP mapping
- Access-list based
- Protocol classification
- Port-based marking
Queuing & Scheduling
Differentiated packet treatment
- Priority Queuing (PQ)
- Weighted Fair Queuing (WFQ)
- Class-Based WFQ (CBWFQ)
- Low Latency Queuing (LLQ)
QoS Models
MPLS supports multiple QoS models: Pipe Model (preserve customer markings), Uniform Model (modify markings), and Short Pipe Model (hybrid approach).
Fast Reroute
MPLS Fast Reroute provides sub-50ms failover by pre-computing backup paths and installing backup forwarding entries, enabling carrier-grade network availability.
Link Protection
Protects against link failures
- Pre-computed backup tunnels
- Local repair at failure point
- Fast detection mechanisms
- Automatic failover
Node Protection
Protects against node failures
- Next-next-hop backup
- Bypass entire failed node
- More comprehensive protection
- Higher complexity
Protection Type | Failure Detection | Recovery Time | Implementation |
---|---|---|---|
FRR Link Protection | Layer 1/2 alarms, BFD | < 50ms | Backup tunnels to next-hop |
FRR Node Protection | Layer 1/2 alarms, BFD | < 50ms | Backup tunnels to next-next-hop |
Global Repair | IGP convergence | 1-3 seconds | New path computation |
Segment Routing
Segment Routing (SR) simplifies network operations by encoding paths as sequences of segments, eliminating the need for per-flow state and enabling centralized path computation.
SR Advantages
Segment Routing provides simplified operations, centralized control, stateless forwarding, and improved scalability compared to traditional MPLS-TE.
SR Segments
- Node Segment: Identifies a specific router
- Adjacency Segment: Identifies a specific link
- Prefix Segment: Shortest path to prefix
- Anycast Segment: Path to anycast address
SR Implementation
- SR-MPLS: Uses MPLS data plane
- SRv6: Uses IPv6 extension headers
- SDN Integration: PCE-based control
- IGP Extensions: IS-IS, OSPF support
Feature | Traditional MPLS-TE | Segment Routing | Benefit |
---|---|---|---|
State Requirements | Per-tunnel state | Stateless forwarding | Improved scalability |
Path Signaling | RSVP-TE required | No signaling needed | Simplified operations |
Path Control | Distributed | Centralized (SDN) | Better optimization |
Convergence | RSVP re-signaling | IGP convergence only | Faster recovery |
Anycast RP
Anycast RP (Rendezvous Point) provides redundancy and load balancing for PIM-SM multicast deployments in MPLS VPN environments, ensuring high availability for multicast services.
Traditional RP Issues
- Single point of failure
- Sub-optimal source paths
- Load concentration
- Manual failover required
Anycast RP Benefits
- Automatic failover
- Load distribution
- Optimal path selection
- Improved availability
Anycast RP Implementation Methods
RFC 3446 (Anycast RP)
- Multiple RPs with same anycast address
- MSDP between RPs for state synchronization
- IGP advertises anycast RP address
- Sources register with closest RP
RFC 4610 (Anycast RP using PIM)
- No MSDP required
- PIM Register messages handle synchronization
- Simpler configuration
- Better for single-domain deployments
Next Steps
Now that you understand advanced MPLS features, continue to Chapter 7: Troubleshooting MPLS-IP to learn how to diagnose and resolve common MPLS issues.