Cisco Updated 300-420 Exam Questions and Answers by amirah

The correct IPv6 address type is anycast. Cisco defines an IPv6 anycast address as an address assigned to a set of interfaces, typically on different nodes, where a packet sent to the address is delivered to the nearest interface according to the routing protocol. This makes anycast useful for distributing client requests across multiple servers or service nodes that provide the same function. In a request/response application, each server can advertise the same anycast address, and routing selects the closest or best path based on the current topology and metrics. If a server or path fails, routing can withdraw or change the route and send new requests to another service instance. Unique local addresses are private unicast addresses; they do not provide service distribution. Aggregatable global addresses are routable unicast prefixes, not a multi-instance service mechanism. Multicast sends one stream to multiple receivers and is not appropriate for selecting one server to respond to an individual request. Reference topics: IPv6 anycast, service distribution, routing-based server selection, IPv6 address types.

Weighted Random Early Detection is the QoS congestion-avoidance feature that responds to early congestion by selectively dropping packets before a queue becomes completely full. WRED can use packet markings, such as IP precedence or DSCP, to determine drop probability, so lower-priority traffic can be dropped earlier than higher-priority traffic. This helps TCP-based applications slow down before tail drop causes large synchronized retransmissions. CBWFQ is a queuing and bandwidth-allocation mechanism; it classifies traffic into queues and assigns bandwidth, but it is not specifically the feature that randomly drops lower-priority packets during congestion. Tail drop drops packets only when the queue is full and does not intelligently prefer lower-priority traffic. Strict priority queuing protects delay-sensitive traffic by servicing a priority queue first, but it does not provide weighted random early discard behavior. In Cisco QoS design, WRED is used as a congestion-avoidance mechanism, especially for TCP traffic classes where early packet drops can reduce global synchronization and protect higher-priority classes.

Large EIGRP networks require careful hierarchy and query-boundary design. A structured hierarchical topology with route summarization is one of the most important scaling techniques because summarization reduces routing table size, hides topology changes, and limits the EIGRP query scope when failures occur. This improves convergence and reduces CPU and bandwidth consumption on routers that do not need detailed route visibility. Implementing multiple EIGRP autonomous systems can also be used as a scaling technique in very large designs because it creates administrative and query boundaries, although it introduces redistribution and policy complexity that must be controlled. Sub-second timers are not a primary scaling method; aggressive timers can increase control-plane load and do not reduce the size of the routing domain. Distribute lists can filter routes, but broad filtering alone is not the core scaling architecture for a network with more than 1000 routers. Modifying delay values influences metric calculation and path selection, not EIGRP domain scale. Therefore, the valid scaling techniques are structured hierarchy with summarization and, where justified, multiple EIGRP autonomous systems.

DiffServ is the correct scalable QoS architecture for classifying traffic according to business requirements and using DSCP values as the priority descriptor. Cisco QoS design distinguishes DiffServ from IntServ. IntServ uses per-flow signaling with RSVP and can reserve resources, but that model is not operationally scalable across large enterprise networks. DiffServ is class-based. Traffic is classified and marked at the network edge, then each device applies per-hop behavior to those classes using local QoS policy. DSCP provides a six-bit field in the IP header, which supports more than ten classification values and allows standardized treatment classes such as EF, AF, class selector, and default. Best effort provides no differentiated treatment, and RSVP is a signaling mechanism associated with IntServ rather than the requested scalable class-based architecture. “Interserv” is not the correct Cisco QoS model term. Therefore, the engineer should design around DiffServ, with a clear trust boundary, marking policy, queuing model, and consistent treatment across WAN and campus boundaries. Reference topics: DiffServ, DSCP, per-hop behavior, class-based QoS, scalable QoS architecture.