route aggregation

Suppose four neighboring streets all share the same postal region. Rather than printing four separate entries on the sorting machine, the post office writes one rule covering the whole region. Route aggregation is the same compression for routers: combining several specific network prefixes that share a common path into a single, shorter prefix that covers them all.

It rides on the structure of CIDR addressing. A prefix like 192.0.2.0/24 names a block of addresses; if a router has four adjacent blocks — 192.0.2.0/24, 192.0.3.0/24, 192.0.4.0/24, 192.0.5.0/24 — that all reach the same next hop, it can sometimes advertise a single covering prefix such as 192.0.0.0/22 instead of four. The shorter prefix length (smaller number after the slash) means fewer bits are fixed, so the one entry spans a larger range. Routers receiving the aggregate store one route where they would have stored several, and longest-prefix match still works correctly because more specific routes, if present elsewhere, are preferred.

It matters as the practical engine of hierarchical routing and Internet scalability. Aggregation is why the global routing table holds on the order of a million prefixes rather than billions: providers summarize their customers' many small networks into a few large advertised blocks. Without it, routing tables and update churn would explode.

An honest caveat: aggregation only works when the addresses are allocated contiguously and actually share the same path — you cannot summarize randomly scattered blocks. And like all summarization, it can hide detail: if one network inside an aggregate fails or needs a different path, the coarse advertised prefix may not reflect that, occasionally sending traffic the wrong way until a more specific route is announced. It is a scalability win bought with a little loss of granularity.