Your data does not travel as one solid wire â it is cut into packets, and every router it meets decides the next hop. Follow one packet across the network, table lookup by table lookup.
When you send data across a network, it is not transmitted as one continuous stream. It is broken into small units called IP packets, and each packet is forwarded separately by a chain of routers until it reaches its destination.
Routing is the process of choosing, at each step, where a packet should go next. It is what lets a message cross many different networks â each owned and run by someone else â without any of them needing to know the full route.
Because independent packets are far more flexible than one long connection. Different packets can take different paths, share links with other traffic, and route around a failure. If one path becomes unusable, later packets can simply be sent another way.
The trade-off is that the network only does its best: routers forward what they receive but do not guarantee delivery or order. Making the data reliable again is a job for the endpoints.
The sending host wraps a chunk of data in an IP packet and stamps it with a source address and a destination address. From here on, the packet is self-contained.
The packet reaches the local gateway, Router A. It reads the destination address and checks its routing table for the entry that best matches, then forwards the packet toward that next hop.
Router B does the same thing, then Router C. Each one only decides the next hop â not the whole journey â and passes the packet along to a router that is closer to the destination.
The routing table is the heart of the decision. It lists destination ranges and the next hop to use for each. The router picks the most specific entry that matches the packet's address.
Eventually a router finds the destination host on a network it is directly connected to and delivers the packet. The same journey happens, in reverse, for any reply.
Because every hop decides on its own, packets can take different paths, arrive out of order, or be dropped along the way. Reassembly and retransmission are handled by the endpoints, not the network.
It is a small, self-contained chunk of data. Each packet carries its own source and destination address, so routers can forward it independently without knowing the whole message it belongs to.
A router reads the destination address on each packet, looks it up in its routing table, and forwards the packet to the next hop that moves it closer to that destination.
It is a list of rules that maps destinations to next hops. When a packet arrives, the router finds the best matching entry and uses it to decide where the packet goes next.
Not necessarily. Each hop is a separate decision, so different packets of the same message can take different paths and may even arrive out of order.
No single controller does. Each router decides only its own next hop, and the end-to-end path emerges from all those local choices added together.
Routers simply forward what they receive; they do not promise delivery. Recovery is handled by protocols on the endpoints, which detect missing data and ask for it again.