Cloud Networking Guides
Cloud networking borrows the vocabulary of traditional networking — subnets, routing tables, load balancers, gateways — and reuses most of the concepts, but with different defaults, different pricing models, and different operational concerns. These guides explain how the major cloud network primitives actually work, where the costly traps are (egress bandwidth is the biggest), and how to choose between the connectivity options.
Where to start
If you are new to cloud networking, begin with what is a VPC — every other concept in cloud networking builds on the Virtual Private Cloud model. Once that is concrete, the next decisions are how to connect on-premises networks to the cloud (VPN or dedicated connectivity), and how to connect multiple VPCs together (transit gateway vs peering).
If you are debugging cloud networking issues or trying to control costs, jump to cloud egress costs — bandwidth pricing is where most cloud spend surprises come from.
Fundamentals
The core building blocks every cloud network uses, regardless of provider.
Fundamentals
What is a VPC?
Virtual Private Cloud basics — AWS VPC, Azure VNet, and GCP VPC compared.
Subnets and CIDR
VPC address planning and subnet layouts that scale.
NAT Gateway vs Internet Gateway
When private resources need outbound internet — patterns that work and patterns that break.
Cloud Load Balancer Types
L4 vs L7, ALB vs NLB vs GLB — when each one is the right tool.
IPv6 in the Cloud
Dual-stack VPCs, egress-only gateways, and the end of NAT.
Connectivity
Direct Connect vs ExpressRoute vs Interconnect
Dedicated cloud connectivity across AWS, Azure, and GCP compared.
Site-to-Site VPN to Cloud
The cheaper alternative to Direct Connect — when VPN-based connectivity is the right call.
Transit Gateway vs VPC Peering
Connecting multiple VPCs — when to use peering and when to use a transit hub.
Private Endpoints
PrivateLink, VPC endpoints, and private access to cloud services.
Internet Egress Paths
Six ways outbound traffic can leave a VPC and how to pick.