Trending:
Cloud & Infrastructure

IPv4 addressing basics still matter for cloud and Kubernetes deployments

Despite IPv6's slow rollout (40% adoption as of 2024), enterprise architects still need solid IPv4 fundamentals. Subnetting isn't legacy knowledge - it's how you design VPCs, segment Kubernetes clusters, and avoid address exhaustion in hybrid environments. The street-address analogy remains the fastest way to build intuition.

IPv4 addressing basics still matter for cloud and Kubernetes deployments

Why this still matters

IPv4 subnetting isn't exam trivia - it's active infrastructure knowledge. Cloud providers require subnet design for VPCs. Kubernetes networking depends on CIDR blocks. Multi-region deployments need careful address planning. The 4.3 billion IPv4 addresses ran out years ago, but most enterprise networks still run dual-stack (IPv4 + IPv6) or IPv4-only. Understanding how the 32-bit address space divides is non-negotiable.

The street address model works

Keith Barker's analogy holds up: an IP address like 192.168.1.25 breaks into network portion (the street, 192.168.1) and host portion (the house number, .25). The subnet mask - say /24 or 255.255.255.0 - defines the split. A /24 gives you 254 usable host addresses. Need more networks with fewer hosts each? Borrow host bits: /25 halves your hosts to 126 but doubles your networks.

Ports extend the analogy: they're doors on each house. 192.168.1.25:443 means "device 25, HTTPS service." Routers read the network portion first (find the street), then forward to the specific host (find the house), then the port handles the service (knock on the right door).

What changed (and what didn't)

The old classful system (Class A/B/C with fixed boundaries) died decades ago. Modern networks use CIDR (Classless Inter-Domain Routing) and VLSM (Variable Length Subnet Masking) to allocate addresses efficiently. A /22 might make sense for one subnet; /28 for another. This flexibility is why cloud providers let you carve up VPCs however you need.

Notably: private RFC1918 ranges (10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16) still dominate internal networks. Your laptop's 192.168.1.x address is NAT'd behind a single public IP at the router. This matters when designing hybrid cloud connectivity or VPN tunnels - overlapping private ranges break routing.

The real trade-off

IPv6 should replace this. It has 340 undecillion addresses and no NAT complexity. But enterprise adoption is slow - legacy applications, training gaps, and "if it's not broken" inertia. Until that flips, subnetting stays in the toolkit. Not as nostalgia, as necessity.