Navigation – Contents – – 1 Moving to Performance Routing – 1.1 Better Use of Wide area network Links – 1.2 Better Application performance and accessibility – 1.3 Application Routing – 1.4 Performance Routing – two Performance Routing Phases – 3 PfR Technical Overview – 3.1 General – 3.2 Enterprise Domain, MC Peering – 3.3 PfR Target Discovery – 3.4 Performance Routing Coverage Engine – 3.5 Reacquisitability must be confirmed – 4 Provisioning MC and BRs – 4.1 Provisioning the BRs – 4.2 Provisioning that the MC – 5 Learning of traffic classes – 5.1 Definition – 5.2 Manual Configuration – 5.3 Automated Configuration – 5.4 The way to use the master list traffic class standards – 5.4.1 Example1 – if you wish to learn only 10.1.0.0\/16 prefixes – 5.4.2 Example2 – Proper use of ACL with master list – 5.5 Performance of Groups, aka Service Courses – 6 Measuring Performance and Bandwidth – 6.1 Mode monitor passive – 6.2 Mode monitor active – 6.3 Hybrid Modes – 6.3.1 Mode monitor both – 6.3.2 Mode monitor Fast – 6.4 Mixing Modes – Enterprise Wide area network Use Case – 7 Policy Performance – 7.1 PfR Features that allow load balancing – 7.1.1 Link Utilization – 7.1.2 Range – 7.2 Traffic Class Performance – 7.3 PfR Policy Thresholds – 7.4 Resolvers – 7.5 PfR Traffic Class States – 7.6 Configuration Example – Enterprise Wide area network Use Case – 8 Enforce Period – 9 Reporting – 10 Best Practices – 10.1 Load Period and Bandwidth – 10.2 Passive mode – Better Utilization of Wide area network Links – Many clients have multiple links to the Wide area network, being privately held or handled by a Service Provider.

Routing protocols aren’t aware of such events. This could lead to a bad user experience, to application performance degradation. But still, that the routing protocol is up. Blackout: it’s 100% loss of data. Brownout: it’s subjective – It’s a high loss and very high delay condition. Challenges: Program performance degraded – Unaware of software errors – New application requirements – Yet No routing issues reported – Where is the problem? Solution: Avoidance of network drops and soft Errors – Intelligent Routing based on functionality – Moving to adaptive routing based upon application delay, loss, jitter, bandwidth – Program Routing – Among the key requirement is to be capable to route not just based on the prefix, but additionally based on the kind of application.

Routing should take that the best path for an application as opposed to a prefix. Challenges: Critical application routed on main link for best SLA – Utilization backup connections for best effort applications – But still. Utilization backup in case of performance degradation – Current Deployment: Complex configuration based on EEM scripts and Coverage Based Routing. Static routing – Solution: Responsiveness to critical application performance requirements – Time\/delay sensitive: voice, vertical apps – Loss very sensitive! Video, circuit emulation – Data centre traffic: SAN expansion, Internet ISP load balancing – Transactional traffic: ecommerce transactions, automated Business-to-business, ERP – Performance Routing – Cisco Performance Routing enhances routing in order to select that the best path based on user defined policy.