Which test outcome is NOT claimed?

Measured zero-loss physical failover.

CCNA Day 33: IPv6 Static Routes and Link-Local Next Hops

First, the essential ideas

IPv6 prefix: An IPv6 address range written with a slash and count of fixed leading bits, such as 2001:db8:20::/64. IPv6 addresses contain 128 bits.
Network, default and host routes: A network route covers a prefix such as /64. The default ::/0 covers otherwise unmatched destinations. A host /128 identifies one exact IPv6 address.
Link-local scope: A link-local address such as fe80::2 identifies an interface on its own link and is not forwarded between links. The output interface disambiguates its scope.
Neighbor Discovery: IPv6's ICMPv6-based mechanisms for discovering local neighbors and their link-layer addresses. IPv6 does not use IPv4 ARP.
Floating static and distance: A higher-distance candidate for the same prefix. Here the /64 at distance 200 becomes selected after the preferred /64 at distance 1 is removed.
Fully specified IPv6 route: A route containing the destination prefix, local exit interface and next-hop address. A link-local next hop needs the interface because its scope is local to a link.
Source-qualified probe: A ping whose source address is explicitly selected. It tests the reverse route to that source rather than relying on the router's default source choice.
Null0 and /128 exception: Null0 intentionally discards selected traffic. A /128 discard is more specific than matching /64 and /0 routes, even when its distance is higher.

An IPv6 destination and its next hop have different jobs. The packet remains addressed to the remote loopback; fe80::2 identifies only the adjacent router receiving its next Ethernet frame. Writing the correct interface is part of the next-hop identity, not optional decoration.

See the idea

Scope the next hop, preserve the destination

fe80::2 identifies an adjacent neighbor on Gi0/0; the packet still targets R2's loopback.

Step 1 of 3

Select the destination

R1 looks up 2001:db8:20::10 and normally selects its network /64.

What this model leaves out: Primary path only; Gi0/1 backup is described in the lab, not drawn here. Interface and neighbor nodes are logical forwarding roles, not extra routers. No packet capture is claimed.

Full visual explanation and references

R1's loopback sends through its Gi0/0 interface to R2's fe80::2 neighbor, then terminates at R2 Loopback0 2001:db8:20::10.

Primary path only; Gi0/1 backup is described in the lab, not drawn here. Interface and neighbor nodes are logical forwarding roles, not extra routers. No packet capture is claimed.

Select the destination
R1 looks up 2001:db8:20::10 and normally selects its network /64.
www.cisco.com
Resolve the scoped hop
Gi0/0 and fe80::2 identify the adjacent receiving interface; the destination is unchanged.
www.rfc-editor.org www.rfc-editor.org
Terminate at R2
R2 delivers to its local loopback; the reply needs a route to R1's explicit source.
www.cisco.com

One idea at a time

Start with something familiar.

Write the destination 2001:db8:20::10 separately from neighbor fe80::2 on R1 Gi0/0. Then add the alternate neighbor fe80::22 on Gi0/1. Each neighbor belongs to a specific local link.

Why learn this?

A valid-looking IPv6 address does not identify the right Ethernet exit by itself. A missing interface or a more-specific discard can break the exchange while other IPv6 tests remain healthy.

An everyday comparison

A building can have the same room number on different floors. A local room reference needs the floor as well as the number, while a full delivery destination names the final office.

What happens in a network

The floor resembles interface scope for a link-local next hop. The final office resembles the remote destination prefix, which is not replaced by the local neighbor's address.

Where the comparison stops

An interface is a network scope, not a physical floor. The comparison does not model Neighbor Discovery, prefix lookup or actual packet timing.

Follow one worked example

Enable and address both routers.
Each transit gets a documentation-prefix address and a deliberate link-local value.
Why: This makes the actual neighbors inspectable without deriving unknown addresses.
Compare the /64 candidates.
The preferred /64 uses Gi0/0 and the floating /64 uses Gi0/1.
Why: The equal prefix lengths permit distance to select the preferred candidate.
Remove only the preferred route.
The backup /64 becomes selected even though the default still points at the primary link.
Why: Longer prefix specificity and same-prefix distance remain separate decisions.

Your first small task

Write the fully specified next-hop identity for R1's backup route and state whether packets to the remote loopback become addressed to fe80::22.

Show a hint

Separate the IP destination from the next Ethernet delivery.

Check the expected result

The next hop is fe80::22 on GigabitEthernet0/1. The packet destination remains 2001:db8:20::10; the link-local neighbor is the immediate forwarding recipient.

Pause and explain it in your own words

Can fe80::2 alone identify a remote routed destination?

Show a hint

Link-local addresses cannot be forwarded beyond their link.

Show the explanation

No. It can name an adjacent next hop only with the correct link scope. The remote destination uses its own routable prefix, and the router needs an output interface for a link-local static next hop.

Short answer

What should you understand today?

IPv6 static routing uses ipv6 route with a prefix length rather than an IPv4 mask. Enable IPv6 forwarding, identify the next hop and specify an interface for a link-local neighbor. Verify /64 network, ::/0 default, /128 host and higher-distance floating behavior separately, keeping the packet's source and return route explicit.

Before you begin

Learning targets

Prerequisites

Complete IPv4 route selection, default and floating-route lessons.
Read compressed IPv6 notation, including the difference between ::/0 and ::1.
Use a fresh isolated CML 2.10 IOSv 15.9M pair with two Ethernet links and console access.

By the end

Configure network, default, host and floating IPv6 routes.
Associate each link-local next hop with the correct local interface.
Verify neighbor, selected route and fixed-source traffic evidence.
Distinguish route withdrawal from physical link failure.
Restore the precise baseline after two controlled routing changes.

Words you will meet

Open a word when you need its meaning. You do not need to memorize the whole list before reading.

IPv6 prefix

An IPv6 address range written with a slash and count of fixed leading bits, such as 2001:db8:20::/64. IPv6 addresses contain 128 bits.

Network, default and host routes

A network route covers a prefix such as /64. The default ::/0 covers otherwise unmatched destinations. A host /128 identifies one exact IPv6 address.

Link-local scope

A link-local address such as fe80::2 identifies an interface on its own link and is not forwarded between links. The output interface disambiguates its scope.

Neighbor Discovery

IPv6's ICMPv6-based mechanisms for discovering local neighbors and their link-layer addresses. IPv6 does not use IPv4 ARP.

Floating static and distance

A higher-distance candidate for the same prefix. Here the /64 at distance 200 becomes selected after the preferred /64 at distance 1 is removed.

Fully specified IPv6 route

A route containing the destination prefix, local exit interface and next-hop address. A link-local next hop needs the interface because its scope is local to a link.

Source-qualified probe

A ping whose source address is explicitly selected. It tests the reverse route to that source rather than relying on the router's default source choice.

Null0 and /128 exception

Null0 intentionally discards selected traffic. A /128 discard is more specific than matching /64 and /0 routes, even when its distance is higher.

Start with forwarding and exact address scope

An IPv6 interface address does not by itself state that the router should forward IPv6 between interfaces. This lab enables ipv6 unicast-routing, assigns addresses and checks the relevant interface states. The special address ::1 means the local node itself; it is not the same thing as the router's named Loopback0 interface carrying a documentation-prefix address. Avoid using ::1 as a remote lab target.

Documentation prefixes are safe examples, not external reachable services.
Record the source address with every end-to-end test.

ReferencesCisco IPv6 static routing IPv6 addressing architecture

A link-local next hop needs an interface

Link-local addresses are valid only on their local link. The address may appear on separate links without identifying a single global location, so an IPv6 static route using such a next hop must also state the outgoing interface. Neighbor Discovery resolves the neighbor's link-layer address on that interface. Do not diagnose this using show ip arp, which displays IPv4 information.

Use show ipv6 neighbors for the local neighbor evidence.
Do not forward link-local source or destination packets to a different link.

ReferencesIPv6 addressing architecture IPv6 Neighbor Discovery Cisco IPv6 static routing

Exercise network, default and host scopes

A /64 route covers the entire named destination network. The default ::/0 covers any destination lacking a more-specific installed match. A /128 matches one address. The commands in the lab deliberately test these scopes with the same final endpoint: the network route normally wins, the default carries that exchange when both network candidates are removed, and the temporary host discard overrides both.

Keep the final endpoint constant while changing route scope.
A /128 Null0 route is an intentional drop, not failed neighbor discovery.

ReferencesCisco IPv6 static routing Cisco route selection

Floating selection does not require a broken cable

This exercise removes R1's preferred /64 configuration line while leaving both cables up. The higher-distance /64 should then install through the backup link. The lower-distance default does not override it because /64 is more specific. R2's replies still prefer its primary /64 return route, producing an explicitly asymmetric exchange. That is permitted here because no stateful firewall or policy is present.

Asymmetric routing is not automatically a fault.
Describe the actual withdrawn dependency instead of overstating failover coverage.

ReferencesCisco IPv6 static routing Cisco administrative distance

Read three independent kinds of evidence

The configuration shows what was requested, the selected routing table shows usable route choices, and neighbor plus traffic tests show more about forwarding. A route can exist while the neighbor cannot answer. A neighbor can answer while the source lacks a return route. Compare these observations at each stage and preserve real output only after execution.

Do not paste predicted success markers into a runtime report.
Unexpected behavior should retain its original output for review.

ReferencesCisco IPv6 static routing Cisco IOSv support

Real-world walkthrough

The neighbor address was right on the wrong link

In a hypothetical training network, a learner copies a link-local next hop but pairs it with the backup port instead of the primary. The router has IPv6 enabled and the destination prefix looks correct, yet the frame cannot reach the intended neighbor on that interface.

Separate destination prefix from next-hop address in the route command.
Compare the named output interface with the physical connection map.
Inspect the actual link-local neighbor on that interface using IPv6 evidence.
Correct only the interface and next-hop pairing under the lab change plan.
Repeat the original source-qualified probe and retain the selected route.

For a link-local next hop, the address and interface form one scoped forwarding choice. Correct address text on the wrong port is still wrong.

Guided GNS3 lab

Four IPv6 route types on one complete endpoint pair

Configure both links, verify a network route, withdraw its primary candidate, exercise the default, demonstrate a host discard and restore the baseline. No IOSv runtime was executed during authoring.

R1 Gi0/0 connects directly to R2 Gi0/0; Gi0/1 connects directly to Gi0/1 as backup. R1 and R2 Loopback0 are the source and destination. No switch, host, upstream provider or additional routed hop is omitted.

R1: fresh CML 2.10 IOSv 15.9M, entitled bundled image.R2: matching IOSv node with independent Console access.

Addressing plan

Device	Interface	Address	Purpose
R1	Gi0/0	`2001:db8:12::1/64; fe80::1`	Primary transit
R2	Gi0/0	`2001:db8:12::2/64; fe80::2`	Primary next hop
R1	Gi0/1	`2001:db8:13::1/64; fe80::11`	Backup transit
R2	Gi0/1	`2001:db8:13::2/64; fe80::22`	Backup next hop
R1	Loopback0	`2001:db8:10::10/64`	Fixed source
R2	Loopback0	`2001:db8:20::10/64`	Final target

Set up the lab

Use Cisco Modeling Labs 2.10 with its legally entitled bundled IOSv 15.9M reference image. Record the complete show version release; do not export that image to GNS3. No paid API is involved.
Create fresh nodes and only the listed links. No External Connector, NAT node, bridge, public uplink or production network is permitted. Leave console access independent of the test data links.
IOSv examples use GigabitEthernet0/0 and GigabitEthernet0/1. Confirm those exact ports in the workbench and show ip interface brief before configuring; if your image differs, document an explicit port substitution first.
Decline the initial configuration dialog on a new authorized node, press Enter to reach the prompt, then use enable. Do not bypass unknown credentials. All displayed expected results are predictions, not author runtime.
Without an entitled compatible image, complete the address and route-selection worksheet only. Mark command execution, packet forwarding and recovery unmeasured.

Address R1's primary link

In the isolated CML workbench, select R1, open its Console tab and connect before typing. Start at > or #; use end first if a prior task left configuration mode. Record show version, confirm the two cable endpoints, and enable the primary IPv6 interface.

enable
configure terminal
ipv6 unicast-routing
interface GigabitEthernet0/0
ipv6 address 2001:db8:12::1/64
ipv6 address fe80::1 link-local
no shutdown
end

What each command means

enable: Enter authorized privileged EXEC mode; this does not change the network configuration.
configure terminal: Enter global configuration mode on this named router; accepted changes affect its running state.
ipv6 unicast-routing: Enable IPv6 forwarding on this isolated router.
interface GigabitEthernet0/0: Select the exact declared transit interface.
ipv6 address 2001:db8:12::1/64: Assign the supplied documentation-prefix IPv6 address.
ipv6 address fe80::1 link-local: Assign the explicit link-local address for neighbor identification.
no shutdown: Enable this interface without changing any other port.
end: Return to privileged EXEC; this neither saves nor reverses accepted configuration changes.

What you should see: R1 has forwarding enabled and the primary pair of IPv6 addresses.
Why this step matters: Global forwarding and correct link scope are prerequisites for remote delivery.

Address R1's backup link

In the isolated CML workbench, select R1, open its Console tab and connect before typing. Start at > or #; use end first if a prior task left configuration mode. Use the distinct backup prefix and its explicit link-local identity.

enable
configure terminal
ipv6 unicast-routing
interface GigabitEthernet0/1
ipv6 address 2001:db8:13::1/64
ipv6 address fe80::11 link-local
no shutdown
end

What each command means

enable: Enter authorized privileged EXEC mode; this does not change the network configuration.
configure terminal: Enter global configuration mode on this named router; accepted changes affect its running state.
ipv6 unicast-routing: Enable IPv6 forwarding on this isolated router.
interface GigabitEthernet0/1: Select the exact declared transit interface.
ipv6 address 2001:db8:13::1/64: Assign the supplied documentation-prefix IPv6 address.
ipv6 address fe80::11 link-local: Assign the explicit link-local address for neighbor identification.
no shutdown: Enable this interface without changing any other port.
end: Return to privileged EXEC; this neither saves nor reverses accepted configuration changes.

What you should see: R1's second interface is independently addressed.
Why this step matters: A backup requires a declared working link, not just a second route line.

Address R2's primary link

In the isolated CML workbench, select R2, open its Console tab and connect before typing. Start at > or #; use end first if a prior task left configuration mode. Record show version and configure the peer of R1 Gi0/0.

enable
configure terminal
ipv6 unicast-routing
interface GigabitEthernet0/0
ipv6 address 2001:db8:12::2/64
ipv6 address fe80::2 link-local
no shutdown
end

What each command means

enable: Enter authorized privileged EXEC mode; this does not change the network configuration.
configure terminal: Enter global configuration mode on this named router; accepted changes affect its running state.
ipv6 unicast-routing: Enable IPv6 forwarding on this isolated router.
interface GigabitEthernet0/0: Select the exact declared transit interface.
ipv6 address 2001:db8:12::2/64: Assign the supplied documentation-prefix IPv6 address.
ipv6 address fe80::2 link-local: Assign the explicit link-local address for neighbor identification.
no shutdown: Enable this interface without changing any other port.
end: Return to privileged EXEC; this neither saves nor reverses accepted configuration changes.

What you should see: The primary neighbors should share the same /64 transit.
Why this step matters: This address is the next hop, not the remote loopback destination.

Address R2's backup link

In the isolated CML workbench, select R2, open its Console tab and connect before typing. Start at > or #; use end first if a prior task left configuration mode. Configure the peer of R1 Gi0/1 without changing the primary.

enable
configure terminal
ipv6 unicast-routing
interface GigabitEthernet0/1
ipv6 address 2001:db8:13::2/64
ipv6 address fe80::22 link-local
no shutdown
end

What each command means

enable: Enter authorized privileged EXEC mode; this does not change the network configuration.
configure terminal: Enter global configuration mode on this named router; accepted changes affect its running state.
ipv6 unicast-routing: Enable IPv6 forwarding on this isolated router.
interface GigabitEthernet0/1: Select the exact declared transit interface.
ipv6 address 2001:db8:13::2/64: Assign the supplied documentation-prefix IPv6 address.
ipv6 address fe80::22 link-local: Assign the explicit link-local address for neighbor identification.
no shutdown: Enable this interface without changing any other port.
end: Return to privileged EXEC; this neither saves nor reverses accepted configuration changes.

What you should see: Both explicit transit pairs should be usable.
Why this step matters: The different link-local labels make mistakes easier to spot in neighbor output.

Create R1 source and three routes

In the isolated CML workbench, select R1, open its Console tab and connect before typing. Start at > or #; use end first if a prior task left configuration mode. Install primary and floating /64 candidates plus a default through the primary.

enable
configure terminal
interface Loopback0
ipv6 address 2001:db8:10::10/64
exit
ipv6 route 2001:db8:20::/64 GigabitEthernet0/0 fe80::2
ipv6 route 2001:db8:20::/64 GigabitEthernet0/1 fe80::22 200
ipv6 route ::/0 GigabitEthernet0/0 fe80::2
end

What each command means

enable: Enter authorized privileged EXEC mode; this does not change the network configuration.
configure terminal: Enter global configuration mode on this named router; accepted changes affect its running state.
interface Loopback0: Create the source software interface.
ipv6 address 2001:db8:10::10/64: Assign the fixed end-to-end source.
exit: Return to global route configuration.
ipv6 route 2001:db8:20::/64 GigabitEthernet0/0 fe80::2: Add the preferred network route with scoped primary neighbor.
ipv6 route 2001:db8:20::/64 GigabitEthernet0/1 fe80::22 200: Add the same /64 through the backup at distance 200.
ipv6 route ::/0 GigabitEthernet0/0 fe80::2: Add the default that applies only without a longer matching route.
end: Return to privileged EXEC; this neither saves nor reverses accepted configuration changes.

What you should see: R1 should select the primary /64 for the remote loopback.
Why this step matters: The less-specific default is independent of the two equal-prefix /64 candidates.

Create R2 endpoint and reply routes

In the isolated CML workbench, select R2, open its Console tab and connect before typing. Start at > or #; use end first if a prior task left configuration mode. Create the final target and explicit preferred and backup routes to the source network.

enable
configure terminal
interface Loopback0
ipv6 address 2001:db8:20::10/64
exit
ipv6 route 2001:db8:10::/64 GigabitEthernet0/0 fe80::1
ipv6 route 2001:db8:10::/64 GigabitEthernet0/1 fe80::11 200
end

What each command means

enable: Enter authorized privileged EXEC mode; this does not change the network configuration.
configure terminal: Enter global configuration mode on this named router; accepted changes affect its running state.
interface Loopback0: Create the destination software interface.
ipv6 address 2001:db8:20::10/64: Assign the final target address.
exit: Return to global route configuration.
ipv6 route 2001:db8:10::/64 GigabitEthernet0/0 fe80::1: Add the preferred source-network return route.
ipv6 route 2001:db8:10::/64 GigabitEthernet0/1 fe80::11 200: Retain an available same-prefix return alternative.
end: Return to privileged EXEC; this neither saves nor reverses accepted configuration changes.

What you should see: The endpoint exists and R2 can return to the fixed source.
Why this step matters: Forward intent alone does not install a reverse route on a different router.

Collect the network-route baseline

In the isolated CML workbench, select R1, open its Console tab and connect before typing. Start at > or #; use end first if a prior task left configuration mode. Wait for address initialization and neighbor discovery, then preserve actual observations.

enable
show ipv6 interface brief
show ipv6 neighbors
show ipv6 route 2001:db8:20::10
ping ipv6 2001:db8:20::10 source 2001:db8:10::10

What each command means

enable: Enter authorized privileged EXEC mode to inspect this named router.
show ipv6 interface brief: Check configured IPv6 addresses and interface state.
show ipv6 neighbors: Inspect the IPv6 neighbors and associated interfaces.
show ipv6 route 2001:db8:20::10: Identify the primary /64 selected for this endpoint.
ping ipv6 2001:db8:20::10 source 2001:db8:10::10: Test the endpoint with the fixed source, including its return route.

What you should see: Expect a /64 through Gi0/0 and replies before proceeding.
Why this step matters: A failed baseline must be resolved before attributing later failure to the planned changes.

Withdraw only the preferred network route

In the isolated CML workbench, select R1, open its Console tab and connect before typing. Start at > or #; use end first if a prior task left configuration mode. Leave interfaces, the default and R2 unchanged; this is a configuration withdrawal.

enable
configure terminal
no ipv6 route 2001:db8:20::/64 GigabitEthernet0/0 fe80::2
end

What each command means

enable: Enter authorized privileged EXEC mode; this does not change the network configuration.
configure terminal: Enter global configuration mode on this named router; accepted changes affect its running state.
no ipv6 route 2001:db8:20::/64 GigabitEthernet0/0 fe80::2: Remove only the AD 1 /64 candidate, making the AD 200 candidate eligible.
end: Return to privileged EXEC; this neither saves nor reverses accepted configuration changes.

What you should see: The backup /64 should become the selected network route.
Why this step matters: This is not a measured cable outage or an upstream-liveness detector.

Verify the asymmetric backup exchange

In the isolated CML workbench, select R1, open its Console tab and connect before typing. Start at > or #; use end first if a prior task left configuration mode. Check the outgoing backup while recognizing that R2 still prefers its primary return.

enable
show ipv6 route 2001:db8:20::10
ping ipv6 2001:db8:20::10 source 2001:db8:10::10

What each command means

enable: Enter authorized privileged EXEC mode to inspect this named router.
show ipv6 route 2001:db8:20::10: Expect /64 at distance 200 through Gi0/1 and fe80::22.
ping ipv6 2001:db8:20::10 source 2001:db8:10::10: Test the original endpoints; replies may use R2's primary link.

What you should see: Expect the backup /64 rather than the lower-distance default.
Why this step matters: A more-specific prefix wins, and a valid exchange need not use the same link in both directions.

Restore the preferred network route

In the isolated CML workbench, select R1, open its Console tab and connect before typing. Start at > or #; use end first if a prior task left configuration mode. Recreate the exact primary candidate before testing the default separately.

enable
configure terminal
ipv6 route 2001:db8:20::/64 GigabitEthernet0/0 fe80::2
end

What each command means

enable: Enter authorized privileged EXEC mode; this does not change the network configuration.
configure terminal: Enter global configuration mode on this named router; accepted changes affect its running state.
ipv6 route 2001:db8:20::/64 GigabitEthernet0/0 fe80::2: Restore the AD 1 /64 through the original scoped neighbor.
end: Return to privileged EXEC; this neither saves nor reverses accepted configuration changes.

What you should see: The preferred /64 should become selected again.
Why this step matters: Restoration separates the floating-route experiment from the next scope experiment.

Exercise the default with the same endpoint

In the isolated CML workbench, select R1, open its Console tab and connect before typing. Start at > or #; use end first if a prior task left configuration mode. Temporarily remove both /64 candidates, retaining the default and all return routes.

enable
configure terminal
no ipv6 route 2001:db8:20::/64 GigabitEthernet0/0 fe80::2
no ipv6 route 2001:db8:20::/64 GigabitEthernet0/1 fe80::22 200
end
show ipv6 route 2001:db8:20::10
ping ipv6 2001:db8:20::10 source 2001:db8:10::10

What each command means

enable: Enter authorized privileged EXEC mode; this does not change the network configuration.
configure terminal: Enter global configuration mode on this named router; accepted changes affect its running state.
no ipv6 route 2001:db8:20::/64 GigabitEthernet0/0 fe80::2: Remove the preferred network candidate.
no ipv6 route 2001:db8:20::/64 GigabitEthernet0/1 fe80::22 200: Remove its floating network alternative.
end: Return to privileged EXEC; this neither saves nor reverses accepted configuration changes.
show ipv6 route 2001:db8:20::10: Verify the destination now matches ::/0.
ping ipv6 2001:db8:20::10 source 2001:db8:10::10: Test that the default reaches the same declared R2 endpoint.

What you should see: Expect default-based forwarding to the existing destination.
Why this step matters: This verifies a default without inventing an external destination or claiming Internet access.

Restore networks and add a host discard

In the isolated CML workbench, select R1, open its Console tab and connect before typing. Start at > or #; use end first if a prior task left configuration mode. Restore both /64 entries, then introduce only the temporary /128 fault.

enable
configure terminal
ipv6 route 2001:db8:20::/64 GigabitEthernet0/0 fe80::2
ipv6 route 2001:db8:20::/64 GigabitEthernet0/1 fe80::22 200
ipv6 route 2001:db8:20::10/128 Null0 200
end
show ipv6 route 2001:db8:20::10
ping ipv6 2001:db8:20::10 source 2001:db8:10::10

What each command means

enable: Enter authorized privileged EXEC mode; this does not change the network configuration.
configure terminal: Enter global configuration mode on this named router; accepted changes affect its running state.
ipv6 route 2001:db8:20::/64 GigabitEthernet0/0 fe80::2: Restore the original primary network route.
ipv6 route 2001:db8:20::/64 GigabitEthernet0/1 fe80::22 200: Restore the original floating network route.
ipv6 route 2001:db8:20::10/128 Null0 200: Add the exact-host Null0 exception at distance 200.
end: Return to privileged EXEC; this neither saves nor reverses accepted configuration changes.
show ipv6 route 2001:db8:20::10: Identify the /128 discard as the longest match.
ping ipv6 2001:db8:20::10 source 2001:db8:10::10: Observe the same endpoint exchange being discarded.

What you should see: Expect the /128 to override the healthy network and default routes.
Why this step matters: Host specificity is independent of the lower distance on the broader network route.

Remove the temporary host exception

In the isolated CML workbench, select R1, open its Console tab and connect before typing. Start at > or #; use end first if a prior task left configuration mode. Undo only the deliberately inserted /128 and preserve the restored network pair.

enable
configure terminal
no ipv6 route 2001:db8:20::10/128 Null0 200
end

What each command means

enable: Enter authorized privileged EXEC mode; this does not change the network configuration.
configure terminal: Enter global configuration mode on this named router; accepted changes affect its running state.
no ipv6 route 2001:db8:20::10/128 Null0 200: Remove the exact temporary IPv6 host discard.
end: Return to privileged EXEC; this neither saves nor reverses accepted configuration changes.

What you should see: The preferred /64 should again handle the destination.
Why this step matters: A precise inverse prevents unrelated routing changes from hiding the explanation.

Close with the original IPv6 test

In the isolated CML workbench, select R1, open its Console tab and connect before typing. Start at > or #; use end first if a prior task left configuration mode. Compare the same source, target and selected prefix with the first baseline.

enable
show ipv6 route 2001:db8:20::10
show ipv6 neighbors
ping ipv6 2001:db8:20::10 source 2001:db8:10::10

What each command means

enable: Enter authorized privileged EXEC mode to inspect this named router.
show ipv6 route 2001:db8:20::10: Verify the primary /64 has returned.
show ipv6 neighbors: Check the neighbor/interface evidence without using IPv4 ARP.
ping ipv6 2001:db8:20::10 source 2001:db8:10::10: Repeat the original fixed-source exchange after restoration.

What you should see: Expect restored primary selection and replies; save only actual output.
Why this step matters: A matching retest distinguishes restoration from merely accepted configuration syntax.

Verify

Show actual /64, ::/0 and /128 destination lookups from their distinct stages.
Record the interface beside every link-local next hop.
Identify the backup test as one-sided route withdrawal with possible asymmetric replies.
Compare the exact original source-target probe after all changes are reversed.

Troubleshoot

If a link-local route is rejected or unusable, check the output interface and adjacent address together.
If the route looks right but traffic fails, inspect IPv6 neighbor state and the route back to the fixed source.
If the default is ignored, look for a longer matching /64 or /128 before altering administrative distance.

Licensing and cleanup

GNS3 does not provide Cisco images. Use software only when the applicable license or entitlement permits that environment. Do not share or redistribute image files. Cisco Modeling Labs is the official alternative; its reference images are licensed for use within CML unless separately licensed. A paper exercise needs no image, but does not establish hands-on competence. Virtual routers do not reproduce physical radios, PoE delivery, hardware forwarding rates or failover timing.

Leave both network candidates, the default and the return routes restored; remove the host discard.
Retain actual outputs separately from predictions, then stop the isolated project.

Practice set

Answer before you reveal

01What is the IPv6 default prefix?

It is ::/0, not the loopback address ::1.

A zero-length prefix is a fallback for otherwise unmatched destinations. The address ::1 identifies only the local node.

02Why include an interface with fe80::2?

The next hop has link-local scope and needs a specific output link.

An address alone does not identify which local link contains that neighbor. The route must include the intended interface.

03What replaces IPv4 ARP evidence here?

IPv6 Neighbor Discovery evidence from show ipv6 neighbors.

IPv6 uses ICMPv6-based neighbor mechanisms. An IPv4 ARP table cannot verify the IPv6 next-hop mapping.

04Why can the AD 200 /64 beat an AD 1 default?

The /64 is a longer matching prefix than ::/0.

Distance compares eligible routes for an identical prefix, while forwarding uses the longest matching installed prefix.

05What does the backup test actually withdraw?

Only R1's preferred /64 configuration line, not either physical link.

The result demonstrates route selection after an administrative withdrawal. It does not measure link-down detection or two-sided convergence.

06Does a /128 Null0 exception affect the whole /64?

No; it matches only its one exact IPv6 destination.

Other addresses may continue to match the network route, though this lab supplies only one real destination endpoint.

Keep these

Technical takeaways

IPv6 prefixes use slash lengths, with /128 for one host.
A link-local next hop needs the correct interface.
Neighbor Discovery evidence is separate from route installation.
A floating /64 can outrank a less-specific default.
The backup experiment withdraws a route, not a physical link.
Restore exact commands and repeat the original source-qualified test.

First, the essential ideas

Select the destination

Start with something familiar.

Why learn this?

An everyday comparison

What happens in a network

Follow one worked example

Your first small task

Pause and explain it in your own words

What should you understand today?

Learning targets

Words you will meet

The neighbor address was right on the wrong link

Addressing plan

Set up the lab

What each command means

What each command means

What each command means

What each command means

What each command means

What each command means

What each command means

What each command means

What each command means

What each command means

What each command means

What each command means

What each command means

What each command means

Verify

Troubleshoot

Answer before you reveal

Quiz: prove the reasoning

Technical takeaways