1.2 - Cloud networking (VPCs, subnets, gateways)
Cloud+ CV0-004 objective 1.2 covers designing networks in the cloud. Dividing a VPC's address range into smaller routable segments creates subnets. A subnet becomes public when a route-table entry sends 0.0.0.0/0 to an internet gateway, while a private subnet has no such route and reaches the internet only through NAT. A transit gateway is a hub that many spoke VPCs and on-premises sites connect through for central routing at scale. Fog computing processes data on an intermediate layer between edge devices and the central cloud, reducing latency for IoT. You should also know CIDR blocks, route tables, NAT gateways, peering, DNS and network ACLs versus security groups. Expect scenario questions that describe routing, making a subnet reachable, central connectivity or edge processing and ask which cloud networking construct - subnet, internet gateway, transit gateway or fog - applies.
Split a VPC range into routable segments = subnets. 0.0.0.0/0 to an internet gateway = public subnet. Hub for many VPCs + on-prem = transit gateway. Processing between edge and cloud = fog computing.
Practice questions
1. Which cloud characteristic lets resources grow and shrink automatically as demand changes?
- Elasticity (correct answer)
- Multitenancy
- Portability
- Redundancy
Rapid elasticity scales capacity up or down automatically to match load. Multitenancy is sharing infrastructure; redundancy is duplicate components for HA.
2. Adding more identical instances behind a load balancer to handle load is an example of:
- Vertical scaling
- Horizontal scaling (correct answer)
- Bursting to disk
- Cold standby
Horizontal scaling (scale out) adds more nodes. Vertical scaling (scale up) adds CPU/RAM to one node. Horizontal is more resilient and near-unlimited.
3. Deploying an application across multiple isolated datacenters in one region protects against what?
- A region outage
- A zone outage (correct answer)
- A DNS change
- A billing error
Availability Zones are isolated datacenters within a region; spreading instances across zones survives a single-zone failure. Surviving a whole-region outage needs multi-region.
4. Which trait describes a system designed to grow to handle much larger load over time by adding capacity?
- Elasticity
- Scalability (correct answer)
- Portability
- Durability
Scalability is the planned ability to handle growth by adding resources. Elasticity is automatic short-term expansion and contraction to match live demand.
5. An app must stay available even if one whole datacenter fails, but stays inside a single region. What design achieves this?
- Single-AZ deployment
- Multi-AZ deployment (correct answer)
- Multi-region active-active
- Edge caching only
Spreading instances across multiple Availability Zones survives a single datacenter failure while staying in one region. Multi-region would add cross-region resilience but is broader.
6. Under CAP theorem, a network partition forces a distributed datastore to sacrifice which one?
- Network partition tolerance
- Storage durability guarantee
- Consistency or availability (correct answer)
- Rapid elasticity behavior
CAP says during a partition you must choose between consistency and availability, since partition tolerance is unavoidable on real networks. Durability is not a CAP dimension.