Cisco CCIE Service Provider

BGP is both the provider’s business and its largest risk. You scale with route reflection, steer traffic with policy and communities, protect convergence with edge techniques, and handle large tables.

• Route reflector hierarchy and cluster design
• Communities, policy and traffic engineering
• Add-path and prefix-independent convergence
• Peering with customers, peers and upstreams
• Route origin validation
• Multiprotocol BGP for IPv6 and VPN families
• Diagnosing BGP convergence and policy faults

You build A reflector-based BGP design with customer, peer and transit policy, tested for failover.

Large providers run MPLS across many domains. You master LDP behavior and filtering, build inter-AS and seamless designs with labeled BGP, and handle carrier-supporting-carrier arrangements.

• LDP sessions, label allocation and filtering
• Labeled unicast with BGP
• Seamless MPLS across domains
• Inter-AS options and their trade-offs
• Carrier supporting carrier
• Verifying label paths end to end

You build A two-domain MPLS network joined by labeled BGP, with the end-to-end label path traced.

Segment routing is replacing signaled tunnels, but both remain in networks you will run. You build RSVP-TE tunnels with protection, deploy segment routing with engineered paths, and add fast protection.

• RSVP-TE tunnels and constraint-based paths
• Fast reroute with link and node protection
• Segment routing in the IGP
• Engineered paths with segment policies
• Topology-independent loop-free alternates
• Migrating from RSVP-TE to segment routing
• Troubleshooting a path that is not used

You build A core with one engineered segment routing path and one protected RSVP-TE tunnel.

Transport networks now carry mobile and optical services. You study how mobile backhaul and fronthaul use the IP network, what timing and synchronization demand of it, and where IP and optical layers meet.

• Mobile transport roles and traffic types
• Timing and synchronization: SyncE and PTP
• Slicing and service differentiation in transport
• Optical transport concepts for the IP engineer
• IP over DWDM and cross-layer thinking
• Designing transport for mobile services

You build A transport design for mobile services, with timing and QoS requirements met.

Layer 2 services remain a major revenue line. You deliver point-to-point and multipoint Ethernet services over MPLS, map customer service types to configurations, and operate them with Ethernet OAM.

• Carrier Ethernet service types and attributes
• Pseudowire-based point-to-point services
• Multipoint services and their scaling limits
• Ethernet VPN services over the core
• Customer VLAN handling and tag manipulation
• Ethernet OAM and fault management
• Troubleshooting a service that does not pass traffic

You build A set of Layer 2 services over one core, covering point-to-point and multipoint.

The L3VPN is the carrier’s workhorse and internet access its commodity. You build VPNs with shared services and extranets, deliver internet service in a VRF and in the global table, and support IPv6 across them.

• VRFs, route distinguishers and targets
• PE to CE routing designs
• Extranet and shared-service topologies
• Internet access in a VRF and in the global table
• IPv6 VPN service
• Inter-AS VPN options
• Troubleshooting a VPN route that does not appear

You build A multi-customer VPN network with an extranet, and internet access delivered two ways.

Provider multicast carries video and financial feeds for many customers at once. You build the core multicast service, then deliver multicast to VPN customers using GRE-based and label-based approaches.

• PIM and source-specific multicast in the core
• Rendezvous point design
• GRE-based multicast VPN
• Label-switched multicast VPN
• Signaling with BGP for multicast VPN
• Troubleshooting multicast state across the core

You build A multicast VPN service for two customers on one core, with state traced end to end.

A provider sells a service level, and quality of service enforces it. You design class models, carry markings through MPLS in the right tunneling mode, and apply hierarchical policies at the customer edge.

• Service class design
• Marking and trust at the provider edge
• Uniform, pipe and short-pipe tunneling modes
• Hierarchical QoS for customer handoffs
• Policing, shaping and congestion management
• Verifying and measuring QoS

You build A hierarchical QoS policy for a customer handoff, with each tunneling mode demonstrated.

Subscribers reach the network through broadband gateways and Layer 2 access. You configure subscriber sessions with authentication and addressing, and build the aggregation Layer 2 that supports them and business customers.

• Subscriber models: PPPoE and IP sessions
• RADIUS authentication, authorization and accounting
• Address assignment for subscribers
• Access aggregation with VLAN stacking
• Business customer Layer 2 handoffs
• Subscriber policy and QoS
• Troubleshooting a subscriber that will not come up

You build A broadband gateway serving both session types, with RADIUS-driven policy.

Customers measure a provider by how little they notice failure. You detect failures quickly, repair paths before routing converges, keep forwarding through control plane restarts, and measure the result.

• Failure detection with BFD
• IP fast reroute and loop-free alternates
• Prefix-independent convergence
• Nonstop forwarding, graceful restart and nonstop routing
• Link and node protection at the MPLS layer
• Measuring convergence time
• Designing for a stated recovery target

You build A convergence test report that compares the same failure with and without each technique.

A provider network faces attacks from customers, peers and the internet. You protect the control, management and data planes, filter at the edge, and mitigate attacks in flight.

• Control plane protection on provider routers
• Management plane access and logging
• Edge filtering and anti-spoofing
• Remotely triggered black hole filtering
• Protecting BGP sessions and prefixes
• Infrastructure ACL design

You build A security configuration for a provider edge router, with a working black hole trigger.

Provider networks are too large to run by hand. You monitor services with telemetry, probes and flows, write Ansible playbooks and Python scripts that talk to devices, and automate repetitive provisioning safely.

• Telemetry, SNMP and flow monitoring
• Service assurance with active probes
• Ansible playbooks for provider devices
• Python scripts that talk to routers
• Templated service provisioning
• Verifying changes automatically

You build A playbook and script pair that provisions a customer service and checks it afterwards.

Assemble a carrier network end to end. You build the core, deliver Layer 2 and Layer 3 services, add protection and fast recovery, secure it, automate its operation, then find faults without hints.

• Planning the core and its services
• Building transport and customer services
• Adding protection, QoS and security
• Automating provisioning and verification
• Locating faults across the whole network

You build A complete provider lab with services, protection and automation, and a log of faults found.