CCNA Routing Fundamentals

How Routers Choose Routes

A router's decision has two stages, and the exam tests the order relentlessly. First, longest prefix match: among all routes to a destination, the one with the longest (most specific) subnet mask wins, full stop. A /26 beats a /24 beats a default route, regardless of how each was learned. Only when two routes have identical prefix lengths does the second stage apply: lowest administrative distance (AD) wins, and if the AD ties too, the lowest metric breaks it (with equal cost, the router load-balances).

Learn to read show ip route output fluently. The code letter tells you the source: C connected, L local (the router's own interface address, a /32), S static, O OSPF, D EIGRP, R RIP, B BGP. The numbers in brackets are [AD/metric] — for example O 192.168.2.0/24 [110/20] via 10.0.0.2 means an OSPF route, AD 110, cost 20, next hop 10.0.0.2. A * marks a candidate default route.

The classic exam question gives you a routing table and asks which path a packet takes: apply longest match first, then AD. A second classic shows two equal routes and asks about load balancing — Cisco routers load-balance up to four equal-cost paths by default (maximum-paths can raise it). Per-destination load balancing (the default with CEF) keeps each flow on one path; per-packet can reorder packets and is rarely wanted.

Static Routes and the Default Route

Static routes are hand-configured and never change unless you change them — ideal for stub networks, simple topologies and default gateways. The syntax is ip route <destination> <mask> <next-hop | exit-interface> [AD]. For a point-to-point serial link either form works, but on Ethernet always prefer the next-hop IP (or a fully specified route with both): an exit-interface-only static on Ethernet forces the router to ARP for every destination address, which is inefficient and breaks with some topologies.

The default route — ip route 0.0.0.0 0.0.0.0 <next-hop> — matches everything not matched more specifically, and is how stub networks reach the internet. Verify with show ip route static and show ip route 0.0.0.0. In IPv6 the equivalent is ipv6 route ::/0 <next-hop>.

Watch for the exam's static-route traps: a missing return route (traffic goes out fine but replies have no path back — ping fails one way), a typo'd next-hop in a different subnet (route never installs), and an exit-interface static on Ethernet causing proxy-ARP dependence. When a static route should be in the table but isn't, check the next-hop is reachable — a static route only installs if its next hop is resolvable.

Floating Static Routes

A floating static route is a backup path: a static route configured with a higher AD than the primary routing protocol, so it sits unused in the configuration until the primary route disappears. The classic example backs up an OSPF-learned route (AD 110) with ip route 0.0.0.0 0.0.0.0 192.0.2.1 200 — the trailing 200 is the AD. While OSPF advertises its route, the floating static stays out of the routing table; if the OSPF route is lost, the static "floats" down into the table and takes over.

The exam tests the AD arithmetic directly: any AD higher than the primary's works, but the convention is a clearly-bogus value like 200 or 250 so its purpose is obvious. Remember the AD table cold — connected 0, static 1, eBGP 20, EIGRP internal 90, OSPF 110, RIP 120, iBGP 200 — because questions ask things like "which AD makes this static a backup for the OSPF route?" (anything above 110).

Note that EIGRP appears in the AD table for completeness, but it is not in the current CCNA blueprint — don't spend study time on its configuration. Verify floating statics with show ip route before and after shutting the primary interface: the backup should appear only when the primary vanishes.

First-Hop Redundancy: HSRP, VRRP, GLBP

Hosts need a default gateway, but a single gateway router is a single point of failure. FHRPs (First Hop Redundancy Protocols) let two or more routers share one virtual IP (and virtual MAC) that hosts use as their gateway; if the active router dies, a standby takes over in seconds, transparently to the hosts. All three protocols follow the same pattern — the exam tests the differences.

HSRP is the Cisco default to know: standby 1 ip 192.168.1.254, standby 1 priority 110, standby 1 preempt (needed if you want the better router to retake control), and standby 1 track g0/1 to decrement priority when an uplink fails. VRRP is the multivendor equivalent — terminology changes to master/backup, and preemption is on by default. GLBP adds load balancing: the Active Virtual Gateway hands out different virtual MACs to different hosts so traffic spreads across up to four forwarders.

Verify with show standby brief (HSRP), show vrrp brief, show glbp brief — confirm one router is Active/Master and the other is Standby/Backup, both see the same virtual IP, and priorities are as designed. The exam's favourite trap: preempt configured on only one side, or a priority set without preempt so it never takes effect.

Related CCNA study guides

• CCNA OSPF Explained
• CCNA Subnetting Cheat Sheet
• CCNA NAT and PAT
• CCNA Access Lists (ACLs)

Frequently asked questions