We kick off the CCNA series from the beginning. Operation of IP data networks is weighted as 5% in the CCNA RS blueprint. The first topic is:
1.1 Recognize the purpose and functions of various network devices such as routers, switches, bridges and hubs
A router is a device that routes between different networks, meaning that it looks at the IP header and more specifically the destination IP of a packet to do forwarding. It uses a routing table which is populated by static routes and routes from dynamic protocols such as RIP, EIGRP, OSPF, ISIS and BGP. These routes are inserted into the Routing Information Base (RIB). The routes from different sources compete against each other and the best route gets inserted into the RIB. To define how trustworthy a route is, there is a metric called Administrative Distance (AD). These are some of the common AD values:
0 Connected route
1 Static route
20 External BGP
200 Internal BGP
255 Don't install
If a value of 255 is used, the route will not installed in the RIB as the route is deemed not trustworthy at all.
The goal of routing is to find an outgoing interface so that the traffic can be sent on its way. Routes may be recursive, meaning that the next-hop of a route 10.0.100.0/24 could be 10.0.0.1 and 10.0.0.1 is a connected route. The router needs to go through the recursive routes until it finds an outgoing interface.
A router also breaks up broadcast domains, meaning that broadcasts does not get sent between the ports of a router.
A switch uses the destination MAC address to forward frames and it’s a device mainly found on the Local Area Network (LAN). The switch learns where hosts are connected by looking at the source MAC address of incoming frames. It uses this information to populate the MAC address table where each Virtual LAN (VLAN) has its own forwarding table. If the destination MAC is not in the MAC address table, the switch will flood the frame on all ports in the VLAN, except the port the frame came in on. This is known as unknown unicast flooding. A switch will also flood broadcast frames such as ARP because these frames need to be received by all hosts in the VLAN.
A switch breaks up collisions domains, every host connected to a switch is in its own collision domain. There may still be errors reported on a switch as collisions if half duplex is used but this is most often seen where there is a mismatch in the auto negotiation parameters.
Most modern switches supports at least some form of routing such as static routes. The definition of a switch or a router is very blurry but generally the Catalyst series are regarded as switches even if they do routing. There are different router models such as the ISR routers, ASR routers and so on.
A switch is a multi-port bridge. A bridge is a historical device not longer in use which was used to bridge between LAN segments. It’s main advantage over hubs was that it broke up collision domains. A bridge had fewer ports than a switch, often only two or four ports and it relied on software forwarding as opposed to the hardware forwarding performed by switches in Application Specific Integrated Circuits (ASICs). I’m sure this is all debatable but from a certification view, this is what they expect you to know.
Ah, the mythical creature known as the hub. You won’t find a hub in your network today (I hope). All devices connected to a hub are in the same collison domain. Any frame received on a port was simply sent out all other ports in the hub which meant that anyone could see traffic destined for someone else. It also meant that only half duplex communication was possible. If anyone were to use a hub today, it would most likely be used as a cheap network tap, since traffic received on one port is sent out other ports, it can be used to sniff the traffic. There are of course better ways of doing this such as SPAN, RSPAN, ERSPAN, dedicated network taps and so on.
This was the first post in the CCNA series. Look out for more.