Switching, Routing and Network Protocols
This domain tests how connections are established and how packets find their way: circuit versus packet switching, Ethernet switching, IP addressing and subnetting, routing and the transport protocols that carry services.
Core rules
- Circuit switching reserves an end-to-end path; packet switching shares links and suits bursty data.
- A switch learns source MAC addresses and forwards frames; a router forwards IP packets between networks. Hosts use a default gateway for off-subnet traffic.
- IPv4 addresses have a network and host part. The network and broadcast addresses cannot be assigned to hosts. Private ranges are 10.0.0.0/8, 172.16.0.0/12 and 192.168.0.0/16; NAT translates them.
- Subnetting: block size = 256 − mask octet. /26 is 255.255.255.192, giving 64 addresses and 62 usable hosts.
- Routing: static routes, RIP as distance vector with a 15-hop limit, OSPF as link state.
- TCP is connection-oriented with a three-way handshake and retransmission; UDP has low overhead. Common ports: 80 HTTP, 443 HTTPS, 53 DNS, 67/68 DHCP, 22 SSH, 25 SMTP. ICMP carries ping.
Worked example
Host 192.168.10.70 with mask /26: block size 64, so subnets begin at .0, .64, .128 and .192. The host lies in 192.168.10.64/26. Network address .64, broadcast .127, usable range .65–.126. The default gateway must be one of those usable addresses.
Traps
Calling a switch a router; forgetting network and broadcast addresses; thinking 192.168.x.x is routable on the internet; using 255.255.255.0 for /26; confusing TCP and UDP; saying ping uses a TCP port; treating RIP as link state.