Imagine you're a postal worker in a massive city. Every day, millions of letters arrive, each with a different destination address. You don't memorize the path to every house—instead, you consult a reference that maps address ranges to delivery routes and next-stop sorting facilities.

A router faces a similar challenge, and its solution is the routing table—a comprehensive database that maps network destinations to the paths that can reach them. The routing table is the brain of the router, the repository where all routing knowledge is stored and organized for decision-making.

A routing table (also called a Routing Information Base or RIB) is a data structure stored in a router that contains the routes to particular network destinations. It serves as the authoritative source of routing knowledge for the router.

Formal Definition:

A routing table is a database that maps destination network prefixes to routing information including the next-hop address, outgoing interface, metric, and attributes sufficient to determine the best path to each destination.

The routing table is populated by three sources:

1. Directly connected networks — Automatically added when interfaces are configured
2. Static routes — Manually configured by administrators
3. Dynamic routing protocols — Learned from neighboring routers via OSPF, RIP, BGP, etc.

A complete routing table entry contains all the information necessary to make routing decisions. Let's examine each component in detail:

Destination Prefix: The network address and subnet mask (or prefix length) that this entry covers. This is the primary lookup key. Examples:

• 10.0.0.0/8 — All addresses from 10.0.0.0 to 10.255.255.255
• 192.168.1.0/24 — Addresses from 192.168.1.0 to 192.168.1.255
• 0.0.0.0/0 — The default route, matches any address

Next-Hop Address: The IP address of the next router that should receive packets destined for this prefix. For directly connected networks, this may be marked as 'connected' or 'directly attached' rather than specifying an IP.

Outgoing Interface: The local interface (physical or logical) through which packets should be sent. Even with a next-hop IP, the router needs to know which interface connects to that next-hop.

Understanding the Output:

• Code letters indicate route source (C=Connected, S=Static, O=OSPF, B=BGP)
• [110/20] means Administrative Distance 110, Metric 20
• via 10.0.0.2 indicates the next-hop router
• 00:05:32 shows route age (how long since last update)
• GigabitEthernet0/0 is the outgoing interface
• S* indicates this static route is the candidate default route

When a router learns multiple routes to the same destination from different sources, it must select which route to install in the routing table and ultimately use for forwarding. This selection follows a strict hierarchy:

Step 1: Longest Prefix Match First, consider only routes that match the destination. More specific prefixes (longer masks) take priority over less specific ones.

Step 2: Administrative Distance (AD) Among routes from different sources for the same prefix, prefer the source with the lowest Administrative Distance. This measures 'trust' in the routing source.

Step 3: Metric Comparison Among routes from the same source for the same prefix, prefer the route with the lowest metric. Metrics are protocol-specific (hop count for RIP, cost for OSPF).

Step 4: Load Balancing (ECMP) If multiple routes have equal AD and equal metric, the router may install all of them for Equal-Cost Multi-Path (ECMP) load balancing.

Routes in the routing table have a lifecycle—they're installed, maintained, and eventually removed. Understanding this lifecycle is crucial for troubleshooting routing issues.

Route Installation:

1. Connected routes are installed immediately when an interface is configured with an IP address and comes up operationally.

2. Static routes are installed when configured, provided the next-hop is reachable and the outgoing interface is up.

3. Dynamic routes are installed when received from routing protocols, passing validity checks (no loops, valid metrics, not filtered by policy).

Route Maintenance:

Dynamic routes require ongoing maintenance:

• Periodic updates: Protocols send regular updates to confirm routes are still valid
• Triggered updates: Immediate updates sent when topology changes
• Holddown timers: Stabilize routes by preventing rapid oscillation
• Route aging: Tracks how long since last confirmation

Modern routers maintain two distinct tables for routing information. Understanding this separation is fundamental to grasping how routers actually work:

Routing Information Base (RIB): The RIB is the control plane's master routing database. It contains:

• All routes learned from all sources
• Complete routing attributes and policies
• Routes that might not be 'best' but are kept for backup
• Administrative and operational metadata

Forwarding Information Base (FIB): The FIB is the data plane's lookup table. It contains:

• Only the 'best' routes (after selection process)
• Optimized for fast lookup (may use tries, TCAMs)
• Pre-resolved next-hop information (MAC addresses)
• Minimal metadata (only what's needed for forwarding)

The Internet's routing table has grown dramatically over decades. In 1989, there were approximately 5,000 routes. Today, the global BGP table exceeds 950,000 IPv4 prefixes and 200,000 IPv6 prefixes. This growth presents significant challenges:

Memory Requirements: Each routing table entry consumes memory. With nearly a million entries and overhead for data structures, the RIB alone can require several gigabytes. The FIB, stored in expensive TCAM or SRAM, is even more constrained.

CPU Requirements: Every route update requires processing—recalculating best paths, updating data structures, synchronizing FIBs. During convergence events, routers may process thousands of updates per second.

Convergence Time: More routes mean longer SPF calculations (for link-state protocols) and more updates to propagate (for distance-vector). Large routing tables slow network convergence after failures.

Scalability Techniques:

1. Route Aggregation: Summarize multiple specific prefixes into a single aggregate (supernet). Reduces table size at the cost of path optimality.

2. Route Filtering: Don't accept routes you don't need. Edge networks can use default routes instead of the full BGP table.

3. Hierarchical Routing: Divide networks into areas (OSPF) or tiers. Routers only need full knowledge of their area.

4. Hardware Advances: More TCAM capacity, faster processors, compressed data structures.

5. Software Optimizations: Incremental SPF, batched FIB updates, optimized lookup algorithms.

Network engineers regularly examine routing tables to diagnose connectivity issues, verify routing policy, and optimize network performance. Key commands and techniques:

Cisco IOS/IOS-XE:

Linux/Unix:

We've comprehensively explored the routing table—the central data structure of router control plane operation. Let's consolidate the key concepts: