Cisco CCNP Service Provider

Customers buy treatment as well as bandwidth. You classify and mark traffic into service classes, carry markings across MPLS, and apply queuing, policing and shaping to deliver the promised behavior.

• Differentiated services and class design
• Classification and marking at the edge
• Carrying markings into MPLS
• Queuing and congestion avoidance
• Policing and shaping customer traffic
• Verifying QoS behavior

You build A four-class QoS policy for a provider edge, with verification commands.

A provider network is a high-value target. You protect routing sessions, limit management access, and filter spoofed or abusive traffic so one customer cannot damage the rest.

• Control plane policing
• Routing protocol authentication and BGP protections
• Prefix filtering and origin validation
• Management plane access control
• Unicast reverse path forwarding and anti-spoofing
• Infrastructure ACLs

You build A security baseline for a provider router covering all three planes.

Interior routing carries the core. You configure IS-IS and OSPF for IPv4 and IPv6, choose areas or levels, tune metrics and timers, and troubleshoot adjacencies at scale.

• IS-IS levels, NETs and adjacencies
• OSPF areas and LSA flooding at scale
• IPv4 and IPv6 in one IGP
• Metrics, wide metrics and path control
• Timers and convergence tuning
• Troubleshooting neighbor and database faults

You build A dual-stack IS-IS core with tuned metrics and verified adjacencies.

BGP carries every customer and peer. You configure iBGP and eBGP, scale with route reflectors, read attributes and communities, and explain why one path wins over another.

• iBGP, eBGP and session design
• Route reflectors and confederations
• Path attributes and the best-path algorithm
• Communities and how providers use them
• Multiprotocol BGP for IPv6
• Troubleshooting session and prefix problems

You build An iBGP design with route reflectors and customer routes tagged with communities.

Policy decides what is advertised and preferred, and availability decides whether customers notice a failure. You write routing policy, describe IPv6 transition methods, and use BFD and graceful restart to cut outage time.

• Prefix lists, route policy and route maps
• Applying policy to inbound and outbound routes
• Dual stack, tunneling and translation for IPv6
• Bidirectional forwarding detection
• Graceful restart and nonstop forwarding
• Equal-cost paths and load sharing
• Troubleshooting a policy that does not match

You build A routing policy set with BFD enabled and a documented failure test.

MPLS lets the core forward on labels, not addresses. You follow a packet through label push, swap and pop, run LDP, and read label tables to verify a path.

• Why MPLS exists and what it carries
• Label stacks and forwarding operations
• LDP sessions and label bindings
• Penultimate hop popping
• Reading label forwarding tables
• Troubleshooting a broken label path

You build An MPLS core in a lab where every label operation is traced hop by hop.

Shortest path is not always the best path. You steer traffic with MPLS traffic engineering and segment routing, and see how segment routing simplifies a core by carrying paths in the packet.

• Why steer traffic away from the shortest path
• RSVP-based traffic engineering tunnels
• Segment routing with the IGP
• Segment lists and engineered paths
• Fast reroute for protection
• Comparing traffic engineering with segment routing

You build A segment routing lab with one engineered path and a protected failover.

Many customers still buy Ethernet service. You describe point-to-point and multipoint Layer 2 VPNs, the Carrier Ethernet service types, and how a provider delivers them over MPLS.

• Pseudowires and point-to-point services
• Multipoint Layer 2 services
• Carrier Ethernet service types
• Attachment circuits and VLAN handling
• Ethernet OAM at a working level
• Verifying a Layer 2 service end to end

You build A configured point-to-point Ethernet service between two customer sites.

L3VPN is the provider’s core business service. You build VRFs, route distinguishers and targets, run multiprotocol BGP between provider edges, and exchange routes with customers.

• VRFs and customer separation
• Route distinguishers and route targets
• MP-BGP VPN address families
• PE to CE routing options
• Import and export policy for shared services
• Troubleshooting a VPN that does not carry routes

You build A two-customer L3VPN with a shared service and verified route leaking.

Customers buy wide distribution and assured treatment as services. You implement multicast with PIM and carry it across the provider core, and apply customer QoS agreements at the edge.

• PIM modes and rendezvous points
• Multicast routing in the provider core
• Delivering multicast to VPN customers
• Customer QoS agreements at the edge
• Verifying multicast state
• Troubleshooting a missing stream

You build A multicast service for one customer VPN, with its QoS treatment verified.

Provider scale requires machines to configure routers. You read YANG models, use NETCONF and RESTCONF to push configuration, and use gNMI to read state and stream telemetry.

• YANG models as the shared language
• NETCONF operations and transactions
• RESTCONF for simple integrations
• gNMI get, set and subscribe
• Choosing a protocol for a task
• Testing calls against a router

You build A script that reads interface state with gNMI and changes a setting with NETCONF.

Services need to be provisioned consistently and watched afterwards. You describe how Cisco NSO models and deploys services across devices, and configure NetFlow and IPFIX to see who is using the network.

• Service models and device models in NSO
• Service deployment and rollback
• Flow monitoring: NetFlow and IPFIX
• Exporters, records and collectors
• Using flow data for capacity and anomalies
• Linking provisioning to assurance

You build A flow export configuration and a written service model for one customer service.

Build a provider service from core to customer. You choose the routing and transport, create the VPN, protect the planes, add QoS and multicast, and describe how it is automated and observed.

• Reading a service requirement
• Designing core routing and transport
• Delivering the VPN and customer edge
• Adding protection, QoS and multicast
• Automating and monitoring the service

You build A documented provider service design, with configurations and a verification plan.