4. Routing Algorithms and Internet Control Plane

Introduction and Context

Routing is the process by which a network determines how packets should travel from a source to a destination. Forwarding uses an already selected next hop; routing is the control-plane activity that discovers, evaluates, and distributes the information used to make that decision. In a small network, routes may be configured manually. In an Internet-scale environment, however, topology changes, failures, administrative boundaries, policy constraints, and traffic-engineering objectives make static configuration insufficient.

The Internet therefore relies on a control plane composed of routing protocols, routing information bases, policy rules, and operational mechanisms. These components collectively determine reachability. The data plane then applies the resulting forwarding table to individual packets. A useful distinction is that the control plane answers, “Which paths should be used?” whereas the data plane answers, “How should this packet be transmitted now?”

This chapter examines the three principal algorithmic families used in modern networks: distance-vector, link-state, and path-vector routing. It then develops the operation of OSPF, IS-IS, and BGP, with particular attention to convergence, route policy, multicast, traffic engineering, and inter-domain governance. These subjects are central to advanced network design because a technically shortest path is not always the operationally preferred path. Routing is simultaneously an algorithmic, architectural, economic, and governance problem.

Learning Objectives

After studying this chapter, the reader should be able to:

1. Explain the principles and trade-offs of distance-vector, link-state, and path-vector routing.

2. Analyze convergence, loop prevention, scalability, and stability in routing systems.

3. Compare OSPF and IS-IS as interior gateway protocols.

4. Explain BGP route selection, policy control, inter-domain routing, and governance.

5. Evaluate multicast routing and traffic-engineering approaches.

6. Diagnose common routing failures and distinguish control-plane from data-plane problems.

7. Apply routing concepts to enterprise, service-provider, cloud, and multi-homed network scenarios.

8. Assess emerging developments such as segment routing, programmable control planes, and route-security mechanisms.

Routing as a Distributed Decision Process

A routing protocol enables multiple routers to construct a consistent, or at least sufficiently useful, view of reachability. Each router maintains local information, exchanges protocol messages with selected peers, calculates routes, and installs preferred entries in its routing table. The calculation is distributed: no single router normally possesses complete authority over the entire Internet.

Several terms are fundamental. A destination prefix is an address block, such as 203.0.113.0/24, rather than a single host address. A next hop is the neighboring router to which a packet is sent. A metric is a value used to compare paths. A routing domain is a set of routers operating under a common administrative policy. An autonomous system (AS) is a separately administered routing domain identified by an Autonomous System Number (ASN).

A router may learn multiple routes to the same prefix. The route-selection process typically considers protocol preference, route validity, metric, policy attributes, and tie-breaking rules. The selected route is installed in the routing information base and may be transformed into a forwarding information base optimized for rapid packet processing. This separation permits sophisticated control-plane computation without requiring every forwarding decision to repeat the computation.

Three Algorithmic Traditions

Distance-vector routing

Distance-vector protocols advertise a distance, or cost, to destinations. Each router tells its neighbors what it believes the distance to be, and a neighbor calculates whether routing through that router is advantageous. The classical update rule is represented by the Bellman–Ford recurrence:

D_x(y) = min_v { c(x,v) + D_v(y) }

Here, D_x(y) is router x’s estimated cost to destination y, c(x,v) is the cost of the link from x to neighbor v, and D_v(y) is the neighbor’s advertised cost to y.

Distance-vector routing is conceptually simple and requires routers to exchange relatively compact information. However, routers do not possess a complete topology map. This limited knowledge can produce slow convergence and routing loops. The classic “count-to-infinity” problem occurs when neighboring routers continue increasing their estimates after a destination becomes unreachable.

Practical distance-vector protocols use techniques such as split horizon, poisoned reverse, route poisoning, triggered updates, and hold-down timers. These mechanisms reduce, but do not completely eliminate, instability. Enhanced Interior Gateway Routing Protocol (EIGRP) uses a more sophisticated distance-vector approach with the Diffusing Update Algorithm, enabling improved convergence while preserving aspects of distributed-vector computation.

Link-state routing

In link-state routing, each router describes its directly connected links, their states, and their costs. These advertisements are distributed throughout an area or routing domain. Every participating router constructs a link-state database representing the topology and independently computes shortest paths using Dijkstra’s algorithm.

If the graph is G = (V,E), with non-negative edge costs, Dijkstra’s algorithm maintains a set of nodes whose shortest distance from the source is known. It repeatedly selects the unsettled node with the smallest tentative distance and relaxes its adjacent edges. The resulting shortest-path tree provides the next hop for each reachable destination.

Link-state routing generally converges faster and offers a more predictable view of topology than basic distance-vector routing. Its costs include database storage, flooding overhead, and computational requirements. A large flat link-state domain can create excessive control-plane work, so protocols introduce hierarchy, summarization, and scoped flooding.

Link-state knowledge is not the same as perfect knowledge. Databases may be temporarily inconsistent during convergence, advertisements may be filtered or summarized, and equal-cost paths may create multiple valid forwarding choices. Consequently, operators must consider both the algorithm and the information-distribution process.

Path-vector routing

Path-vector protocols advertise reachability together with a path description and policy attributes. BGP, the dominant inter-domain routing protocol, carries an AS_PATH attribute that records the sequence of autonomous systems through which a route has been learned. If a router sees its own ASN in the path, it can reject the route, providing a basic loop-prevention mechanism.

Path-vector routing is not designed to calculate a globally shortest path. Instead, it supports policy-based selection. An organization may prefer a customer route over a provider route, avoid a particular jurisdiction, restrict transit, or influence inbound traffic. BGP therefore combines reachability information with administrative intent.

The distinction among the three families is important. Distance-vector routing distributes destination costs; link-state routing distributes topology; path-vector routing distributes reachability together with a policy-relevant path. Modern protocols sometimes combine features, so these categories describe dominant design principles rather than rigid boundaries.

Convergence, Stability, and Failure Behavior

Convergence is the process by which routers detect a change and reach a new consistent forwarding state. It includes failure detection, information propagation, route calculation, and forwarding-table installation. Fast convergence is desirable, but excessive sensitivity can cause route flapping: repeated changes that consume control-plane resources and disrupt traffic.

Failure detection may use protocol hellos, hold timers, physical-layer signals, Bidirectional Forwarding Detection (BFD), or data-plane probes. Detection speed alone does not guarantee fast recovery. The network must still propagate the event, recalculate routes, and update forwarding hardware.

Convergence has several dimensions:

• Detection time: how quickly a failure is recognized.
• Propagation time: how quickly other routers learn of it.
• Computation time: how quickly alternative paths are calculated.
• Installation time: how quickly forwarding entries are updated.
• Traffic recovery: how quickly packets use a valid alternative.

A network can be highly available while still experiencing transient loops, black holes, or micro-loops during reconvergence. Fast-reroute mechanisms, loop-free alternates, ordered updates, and carefully engineered metric designs reduce these effects. Graceful restart can preserve forwarding during a control-process restart, but it must be used carefully: preserving stale routes for too long can hide genuine failures.

OSPF and IS-IS: Interior Link-State Routing

OSPF architecture and operation

Open Shortest Path First (OSPF) is a link-state interior gateway protocol. Routers form adjacencies, exchange link-state advertisements (LSAs), synchronize a link-state database, and compute shortest paths. OSPF supports hierarchical organization through areas. Area 0 is the backbone, while other areas connect to it through area border routers.

OSPF uses different LSA types to represent router links, network segments, summaries, external routes, and specialized area behavior. Designated routers and backup designated routers reduce adjacency overhead on multi-access networks. OSPF supports equal-cost multipath, authentication options, route summarization, stub areas, and external route redistribution.

An OSPF interface cost is normally derived from bandwidth, although administrators can configure it explicitly. The metric should reflect operational intent rather than merely physical speed. Inconsistent costs can cause asymmetric paths, congestion, or unexpected use of low-capacity links.

OSPF’s area hierarchy limits flooding and computation, but it introduces design obligations. Summarization can improve scalability but may conceal more-specific failures or create less precise path choices. Redistribution between OSPF and other protocols can produce feedback loops, route instability, and ambiguous ownership of prefixes.

IS-IS architecture and operation

Intermediate System to Intermediate System (IS-IS) is also a link-state protocol. It uses a two-level hierarchy: Level 1 routers operate within an area, while Level 2 routers provide inter-area connectivity. A router can operate at both levels. IS-IS transports routing information using a flexible type-length-value (TLV) encoding, which has helped it accommodate extensions such as IPv6 and traffic-engineering information.

Unlike OSPF, IS-IS runs directly over the data-link layer rather than over IP. This gives it a distinctive operational model and avoids dependence on IP address configuration for protocol communication. IS-IS is widely used in large service-provider networks because of its scalable flooding behavior, straightforward two-level architecture, and extensibility.

OSPF and IS-IS can both provide robust interior routing. Selection between them is usually based less on algorithmic capability than on organizational experience, existing tooling, vendor support, operational culture, and integration requirements. OSPF is common in enterprise environments; IS-IS has a strong presence in carrier backbones. Either protocol can be poorly designed if hierarchy, metrics, addressing, and failure behavior are not carefully managed.

BGP and the Logic of Inter-Domain Routing

BGP is a path-vector protocol used to exchange routes between autonomous systems and, frequently, within large autonomous systems. External BGP sessions normally connect routers in different ASes, while internal BGP distributes externally learned routes within one AS. BGP uses a reliable transport connection, commonly TCP port 179, and exchanges incremental updates rather than repeatedly transmitting a complete table.

A BGP route consists of a prefix and attributes. Important attributes include:

• AS_PATH: the sequence of ASes traversed by the route.
• NEXT_HOP: the address to which traffic should be forwarded.
• LOCAL_PREF: an attribute used within an AS to express preferred outbound exit points.
• MED: a suggestion to neighboring entry points, generally meaningful between the same neighboring ASes.
• COMMUNITY: a tag used to apply or communicate policy.
• ORIGIN: information about how the route entered BGP.

Exact decision order varies by implementation and configuration, but a typical process prefers a higher local preference, shorter AS path, better origin type, lower MED in relevant comparisons, an eBGP-learned route over an iBGP-learned route, lower internal cost to the next hop, and deterministic tie-breakers. The central principle is that BGP selects the best route according to policy, not simply the route with the lowest numerical network distance.

BGP’s separation of reachability from policy makes it powerful but operationally complex. Importing a route from a customer, peer, or transit provider should not be treated identically. Export policies determine whether an AS provides transit, advertises only its own prefixes, or selectively announces routes. Prefix filters, maximum-prefix limits, route validation, and explicit community policies are essential controls.

Internal BGP scaling

A full-mesh of iBGP sessions grows quadratically with the number of routers. Route reflectors reduce this requirement by allowing selected routers to redistribute routes within clusters. Confederations divide an AS into sub-autonomous systems for internal management. Both mechanisms improve scalability but introduce additional troubleshooting complexity and can affect path visibility.

BGP convergence and routing stability

BGP intentionally favors policy stability over rapid reaction. Its timers, path exploration, route damping mechanisms, and incremental updates can produce slower convergence than an IGP. When a route disappears, routers may temporarily explore multiple alternatives before settling. BGP PIC, selective route preinstallation, add-path extensions, and carefully designed redundancy can improve recovery.

Route flapping, accidental route leaks, and overly broad announcements can affect networks far beyond the originating organization. A small configuration error can propagate globally if filtering and validation are inadequate. Thus, BGP engineering is inseparable from operational governance.

Route Policy and Inter-Domain Governance

Routing policy translates business, security, and engineering objectives into protocol behavior. A provider may prefer customer-learned routes because customers generate revenue, use peers for settlement-free traffic, and use transit providers only when necessary. An enterprise may select one provider for primary outbound traffic and another for resilience. Policy may also reflect regulatory restrictions, data-sovereignty requirements, or security classifications.

Inbound and outbound traffic require different techniques. Outbound path selection is often controlled by local preference. Inbound traffic can be influenced through selective advertisement, AS-path prepending, more-specific announcements where appropriate, communities, and provider-specific controls. These methods are imperfect because the remote AS retains the final route-selection decision.

Inter-domain governance includes the allocation of IP address space and ASNs, coordination among registries and operators, contractual relationships, and incident response. The Internet is not governed by one central routing authority. It is an ecosystem of technical standards, regional Internet registries, network-operator groups, transit agreements, and bilateral policy decisions.

Security mechanisms are increasingly important. Resource Public Key Infrastructure (RPKI) allows a prefix holder to publish a Route Origin Authorization stating which AS is authorized to originate a prefix. Route Origin Validation classifies announcements as valid, invalid, or unknown. RPKI primarily addresses origin authorization; it does not prove the entire AS path or eliminate all forms of route leakage. Operators must combine validation with filtering, monitoring, and coordinated response.

Multicast Routing

Unicast routing selects a path from one source to one destination. Multicast delivers traffic from one source to multiple receivers identified by a group address. Efficient multicast avoids creating an independent unicast stream for every receiver.

Multicast routing generally relies on a reverse-path check: a router accepts multicast traffic on the interface it would use to reach the source. This helps ensure that forwarding follows a loop-free distribution tree. Protocol Independent Multicast (PIM) uses the underlying unicast table for reachability and builds multicast distribution structures. PIM-SM uses a rendezvous point during group establishment and may later switch to source-specific shortest-path trees. Source-Specific Multicast (SSM) identifies both source and group, reducing some rendezvous and security complexities.

Multicast is useful for live media distribution, financial feeds, collaborative systems, and controlled enterprise applications. It introduces challenges involving receiver authorization, state scaling, interoperability across administrative domains, and troubleshooting. In many public Internet applications, content-distribution networks and application-layer replication have reduced dependence on native inter-domain multicast.

Traffic Engineering and Intent-Based Control

Shortest-path routing optimizes a metric, but network operators often need to control capacity utilization, latency, resilience, or service placement. Traffic engineering modifies path selection to achieve these objectives. Traditional approaches adjust link metrics, manipulate BGP attributes, or establish label-switched paths. Such methods can be effective but may become difficult to maintain as the network grows.

Multiprotocol Label Switching (MPLS) can steer traffic along explicitly selected label-switched paths and support virtual private networks and fast reroute. Segment routing encodes a path as a sequence of segments, reducing dependence on per-flow signaling state. Segment Routing over IPv6 and Segment Routing with MPLS can support policy-driven paths, service chaining, and disjoint-path requirements.

Centralized and software-defined controllers can compute paths using a broader network view. Intent-based networking expresses desired outcomes—such as low latency, isolation, or guaranteed survivability—and automation translates these intents into device configurations. The principal risks are incorrect models, stale telemetry, automation at excessive scope, and lack of safe rollback. Human review remains important for high-impact changes.

Useful traffic-engineering measurements include link utilization, delay, loss, jitter, path diversity, queue occupancy, and failure-recovery time. Optimization should not focus on utilization alone: a lightly loaded route with high delay may be worse for an interactive application than a moderately loaded low-latency route.

Applied Case Study: Multi-Homed Research Cloud Network

Background

Consider an illustrative research organization operating a private cloud across two data centers. It connects to two Internet service providers and maintains an internal backbone with redundant links. The organization hosts research portals, datasets, and real-time laboratory services. Availability is important, but some datasets are subject to access controls and regional handling requirements.

Situation and problem

The organization initially used static routes at the edge and OSPF inside the backbone. During an ISP outage, engineers manually changed default routes. A later configuration error caused one provider to receive a broad internal prefix, creating a potential route leak. At the same time, the primary backbone link became congested because normal shortest-path routing ignored application priorities.

Analysis

Static edge routing did not provide acceptable failure response. OSPF was suitable for the internal topology but could not express inter-provider business policy. The network also lacked explicit prefix filters, RPKI-based origin checks, and telemetry correlating utilization with route changes. A single metric scheme treated bulk data replication and latency-sensitive laboratory traffic alike.

Strategy and solution

The organization introduced eBGP sessions with both providers and used explicit import and export policies. Only approved public prefixes could be exported. Customer, peer, and transit classifications were represented through controlled communities. Local preference selected the preferred outbound provider, while selective announcements and provider communities influenced inbound traffic.

OSPF remained the internal IGP, with carefully reviewed costs and summarization at appropriate boundaries. BFD improved failure detection on critical links. The organization deployed RPKI validation, maximum-prefix limits, route-change monitoring, and automated configuration checks in a staging environment.

For traffic engineering, latency-sensitive services received a policy-based path across the lower-delay backbone link, while bulk replication used the alternate path during peak periods. Segment-routing policies were introduced in a limited domain after validating telemetry and rollback procedures.

Outcome

The organization reduced dependence on manual failover and obtained clearer separation between internal reachability and external policy. The route-leak risk was reduced through filtering and validation. Congestion was not eliminated, but it became measurable and controllable through application-aware path policies. Operational procedures also improved because every change had an owner, a validation test, and a rollback plan.

Learning points

The case illustrates that no single routing protocol solves every problem. The IGP provided internal topology convergence; BGP expressed external policy; traffic engineering addressed resource allocation; and governance mechanisms reduced the impact of configuration error. Reliability depended on the interaction of algorithms, policy, monitoring, security, and organizational discipline.

Challenges, Risks, and Common Mistakes

A frequent mistake is confusing reachability with service availability. A route may exist while the application is unavailable because of firewall rules, asymmetric return paths, DNS failure, or overloaded servers. Another mistake is treating route metrics as universal measures of quality. An OSPF cost, an MPLS administrative metric, and a BGP local preference serve different purposes and should not be interpreted interchangeably.

Over-redistribution between protocols can create feedback loops and obscure route ownership. Excessive summarization may hide failures; insufficient summarization may increase table size and instability. Aggressive timers can cause false failures, while conservative timers delay recovery. Route reflectors can hide alternate paths, and undocumented communities can make policy behavior difficult to predict.

Operators should also avoid using AS-path prepending as a guaranteed inbound-engineering mechanism. It is only an influence, not an instruction. Similarly, adding redundant links without checking shared physical infrastructure may create the appearance of resilience without genuine failure independence.

Professional and Ethical Considerations

Routing engineers control infrastructure that may carry sensitive, safety-critical, or legally regulated information. Unauthorized route changes can redirect traffic, expose data, or cause large-scale denial of service. Change management, least privilege, peer review, audit logging, and separation of duties are therefore professional responsibilities rather than administrative inconveniences.

Network operators should communicate major incidents responsibly, protect customer confidentiality, and avoid using routing mechanisms to bypass lawful restrictions or impose unjustified discrimination. Automated systems require especially careful authorization because a faulty policy can propagate at machine speed. Security validation must be combined with respect for privacy and proportionality in telemetry collection.

Current Developments and Best Practices

Contemporary routing practice increasingly combines traditional distributed protocols with programmable systems, streaming telemetry, and automated validation. Segment routing reduces signaling state and supports policy-based forwarding. RPKI adoption strengthens origin security. BGP monitoring services, route collectors, and anomaly-detection systems help identify leaks and hijacks. Machine-learning techniques may assist in detecting unusual route changes or predicting congestion, but they should support—not replace—protocol reasoning and human accountability.

Recommended practices include:

• Define route ownership, import policy, export policy, and failure behavior explicitly.
• Use prefix filtering, maximum-prefix limits, authentication, and appropriate RPKI validation.
• Separate the roles of the IGP, BGP, and traffic-engineering system.
• Measure convergence, packet loss, latency, and control-plane stability rather than relying on configuration intent.
• Test failure scenarios and rollback procedures before production deployment.
• Prefer hierarchical design and summarization where they do not conceal necessary detail.
• Document communities, local-preference conventions, route-reflector topology, and redistribution boundaries.
• Automate repetitive validation while retaining approval controls for high-impact changes.

Chapter Summary

Routing is a distributed control-plane function that produces the information used by the data plane to forward packets. Distance-vector protocols exchange destination costs, link-state protocols distribute topology information and compute shortest paths, and path-vector protocols exchange reachability together with policy-relevant path attributes. Their differences explain their convergence, scalability, and policy behavior.

OSPF and IS-IS provide interior link-state routing with hierarchical mechanisms for controlling flooding and computation. BGP provides policy-driven inter-domain routing and is central to Internet governance, but its flexibility creates risks involving leaks, hijacks, instability, and slow convergence. Multicast requires distribution-tree construction and reverse-path validation. Traffic engineering extends routing beyond shortest-path selection through MPLS, segment routing, metrics, and controller-based policies.

Effective routing design is not merely a matter of selecting a protocol. It requires coordinated algorithms, metrics, policy, security, monitoring, automation, governance, and operational discipline. Robust networks are designed around failure behavior and verified through measurement rather than assumed to be reliable because redundant links or multiple protocols exist.

Key Terms

Autonomous System (AS): A separately administered routing domain identified by an ASN.

BGP: A path-vector protocol used primarily for inter-domain routing and policy control.

Convergence: The process through which routers reach a new stable forwarding state after a change.

Distance-vector routing: Routing based on exchanging destination distances with neighboring routers.

Forwarding Information Base (FIB): The optimized table used by the data plane to forward packets.

Link-state routing: Routing in which routers distribute link information and independently calculate paths.

Local preference: A BGP attribute used within an AS to select preferred outbound exits.

Metric: A value used to compare candidate routes.

Multicast: Delivery of traffic from one source to multiple receivers using a group address.

Next hop: The neighboring router to which a packet is sent on its selected path.

OSPF: A link-state interior gateway protocol using areas and shortest-path computation.

Path vector: A routing representation containing reachability and the sequence or attributes of domains traversed.

Route reflector: An iBGP router that reduces the need for a full internal BGP mesh.

RPKI: A public-key infrastructure supporting authorization of ASes to originate IP prefixes.

Segment routing: A forwarding method that represents a desired path as an ordered set of segments.

Traffic engineering: Deliberate control of paths to meet utilization, latency, resilience, or policy objectives.

IS-IS: A link-state interior protocol using Level 1 and Level 2 hierarchy and TLV extensions.

Route leak: An unintended advertisement of routes to an inappropriate neighboring network.

Review Questions

Short Answer Questions

1. Distinguish between routing and forwarding.

2. What problem does split horizon address in distance-vector routing?

3. Why does link-state routing require flooding and a link-state database?

4. What is the purpose of the BGP AS_PATH attribute?

5. Define convergence and identify two of its component times.

Long Answer Questions

1. Compare distance-vector, link-state, and path-vector routing with respect to information exchanged, loop prevention, convergence, scalability, and policy support.

2. Explain the architecture and operation of OSPF, including areas, LSAs, designated routers, and route summarization.

3. Compare OSPF and IS-IS and discuss the circumstances in which an organization might prefer one over the other.

4. Explain BGP route selection and analyze the roles of local preference, AS_PATH, MED, communities, and route filtering.

5. Discuss the relationship among routing policy, inter-domain governance, RPKI, and operational security.

Analytical / Application Questions

1. An enterprise has two Internet providers, but all outbound traffic uses the more expensive provider. Propose a BGP policy design and explain how you would verify its effect.

2. A backbone experiences transient forwarding loops after a link failure even though the IGP eventually converges. Analyze possible causes and recommend mitigation techniques.

3. Design a routing and traffic-engineering approach for a cloud network carrying latency-sensitive transactions and large backup transfers over partially redundant links. Justify the roles of the IGP, BGP, telemetry, and automation.

22. Multiple Choice Questions with Answers

Question 1

A router using a basic distance-vector algorithm learns that its route to a destination has failed. Which problem is most directly associated with neighboring routers continuing to advertise increasingly larger costs to that destination?

A. Count-to-infinity

B. LSA fragmentation

C. BGP path hiding

D. Reverse-path multicast filtering

Correct Answer: A. Count-to-infinity

Question 2

What is the principal information distributed by a link-state routing protocol?

A. Only the preferred next hop for every destination

B. A description of local links and their states

C. Only the business relationship between ASes

D. An application-layer service map

Correct Answer: B. A description of local links and their states

Question 3

An AS wants all internal routers to prefer routes learned through Provider X for outbound traffic. Which BGP attribute is most directly suited to this internal policy?

A. MED

B. AS_PATH

C. LOCAL_PREF

D. ORIGIN

Correct Answer: C. LOCAL_PREF

Question 4

Why does BGP reject a route when its own ASN appears in the AS_PATH?

A. To prevent a routing loop

B. To select the route with the lowest delay

C. To force multicast tree construction

D. To enable equal-cost multipath

Correct Answer: A. To prevent a routing loop

Question 5

Which OSPF design mechanism primarily limits the scope of link-state flooding and computation?

A. Autonomous-system prepending

B. Area hierarchy

C. BGP communities

D. RPKI validation

Correct Answer: B. Area hierarchy

Question 6

Which statement best distinguishes BGP from a conventional shortest-path IGP?

A. BGP cannot advertise prefixes

B. BGP selects routes exclusively by physical distance

C. BGP supports policy-based selection across administrative boundaries

D. BGP operates only inside a single broadcast LAN

Correct Answer: C. BGP supports policy-based selection across administrative boundaries

Question 7

A network operator wants to verify whether an AS is authorized to originate a particular prefix. Which technology provides this form of validation?

A. RPKI and Route Origin Validation

B. Dijkstra’s algorithm

C. BFD alone

D. PIM reverse-path forwarding

Correct Answer: A. RPKI and Route Origin Validation

Question 8

In multicast routing, the reverse-path check primarily helps a router to do what?

A. Choose the cheapest BGP provider

B. Prevent forwarding loops by validating the source-facing interface

C. Calculate an OSPF area border

D. Encrypt multicast payloads

Correct Answer: B. Prevent forwarding loops by validating the source-facing interface

Question 9

Which situation is an example of traffic engineering rather than ordinary shortest-path routing?

A. Learning a directly connected subnet

B. Using a policy-based path to keep bulk transfers off a latency-sensitive link

C. Accepting a TCP connection on port 179

D. Assigning an ASN to a router

Correct Answer: B. Using a policy-based path to keep bulk transfers off a latency-sensitive link

Question 10

Why are route filters and maximum-prefix limits important at an eBGP boundary?

A. They prevent all internal link failures

B. They restrict unintended advertisements and limit the impact of routing errors

C. They replace the need for an IGP

D. They guarantee that inbound traffic uses a chosen provider

Correct Answer: B. They restrict unintended advertisements and limit the impact of routing errors