Cloud concept · Updated

Network Peering - Network peering connects independently managed networks so their resources can communicate through explicitly configured routes and access controls.

In short

Peering provides direct network-to-network connectivity without combining the networks into a single administrative domain. In cloud architectures, it commonly enables private communication between virtual networks, but its routing limitations, address requirements, and traffic charges influence whether direct peering or a transit topology is the better design.

Cloud Concepts Hub →

6 key aspects of Network Peering

Effective peering depends on compatible addressing, complete routing, explicit security controls, and a topology chosen for both traffic cost and operational scale.

01
Independent networks with explicit connectivity

Peering connects networks without merging their ownership or security administration. Treat the connection as a routing relationship, not as authorization to access every workload.

02
Address compatibility and return routing

Non-overlapping address space is generally required for cloud virtual-network peering. Verify destination routes and return paths, accounting for whether the provider creates routes automatically.

03
Non-transitive routing boundaries

Two peering connections do not generally create an indirect route between their outer networks. Use a supported transit architecture when workloads require communication through a central network.

04
Topology and operational scale

Direct peering is straightforward for a small number of communicating networks, but a full mesh grows quadratically. Selective peering or a managed transit hub can reduce connection and route-management complexity.

05
Traffic-aware cost optimization

Compare connection charges, transit processing, and regional or zonal data transfer for the actual traffic pattern. A connection with no hourly fee can still generate significant transfer costs.

06
Separate security, DNS, and encryption controls

Peering does not replace firewall policies, private DNS configuration, or application encryption. Validate these controls independently and restrict permitted sources, destinations, and ports.

Cloud services for Network Peering

Amazon VPC, Azure Virtual Network, and Google Cloud VPC implement direct peering within their respective platforms; these mechanisms do not directly peer virtual networks across different cloud providers. For an AWS estate with a few communicating VPCs, direct VPC peering can avoid transit attachment and processing charges. As connectivity requirements grow, AWS Transit Gateway can replace selected mesh connections with centrally controlled transit routing. These approaches can coexist, but routes must deliberately select the intended path. Across all platforms, address planning, security policies, DNS, and traffic-based cost analysis remain necessary.

Amazon VPCAmazon VPC

Amazon VPC implements direct connectivity through VPC peering connections, including supported cross-account and inter-Region configurations. VPC peering requires non-overlapping address space, explicit subnet routing, and appropriate security rules, and does not support transitive routing.

AWS Transit GatewayAWS Transit Gateway

AWS Transit Gateway provides a managed transit hub when many VPCs or hybrid networks need routed connectivity, offering an alternative to a large peering mesh. Transit gateways can also use peering attachments to connect to other transit gateways; assess attachment, processing, and transfer charges for the selected topology.

Network Peering exam questions

Real practice exam questions on Network Peering, with the correct answer and its explanation. Answer them with every option in the exam simulator.

SAA-C03

A company needs to connect two VPCs in different AWS accounts in the same Region. The VPCs have the following IPv4 CIDR blocks:

• Application VPC: primary CIDR 10.10.0.0/16 and secondary CIDR 10.30.0.0/16. The secondary CIDR has no subnets or resources and is no longer needed.
• Analytics VPC: primary CIDR 10.20.0.0/16 and secondary CIDR 10.30.0.0/16. Both CIDR blocks contain active workloads.

Instances in each VPC must initiate TCP and UDP connections to instances throughout the other VPC by using their existing private IP addresses. Existing workloads cannot be moved or readdressed. Security rules already allow the required traffic.

Which solution meets these requirements with the LOWEST recurring network cost?

Show answerCORRECT ANSWER

Disassociate the unused secondary CIDR block from the Application VPC. Create a VPC peering connection, accept it in the other account, and configure routes in both VPCs for the remote workload CIDR blocks.

Removing the unused secondary CIDR eliminates the overlap without changing any workload IP addresses. Cross-account VPC peering then provides the required private connectivity. It has no connection-hour or per-GB processing charges, although applicable data transfer charges still apply.

SAP-C03

Following an acquisition, a company must connect two VPCs in separate AWS accounts in the same AWS Region. The engineering VPC has a primary IPv4 CIDR block of 10.40.0.0/16 and a secondary IPv4 CIDR block of 172.20.0.0/16. The acquired company's VPC has a primary IPv4 CIDR block of 10.60.0.0/16 and a secondary IPv4 CIDR block of 172.20.128.0/17.

All existing workloads use the primary CIDR blocks. Neither secondary CIDR block contains subnets or other resources. Both companies need secondary address space for planned expansion, but those ranges can change.

Workloads in either VPC must initiate TCP and UDP connections directly to private IPv4 addresses in the other VPC, with original source addresses preserved. The solution must support existing workloads and future workloads in the secondary address space. The company wants to avoid intermediary appliances and minimize changes to existing workloads.

Which solution meets these requirements?

Show answerCORRECT ANSWER

Disassociate the acquired company's unused secondary CIDR block and replace it with 172.21.0.0/16. Establish and accept a cross-account VPC peering connection. Configure routes in both VPCs for the remote address ranges and allow the required traffic through security groups and network ACLs.

Replacing the unused overlapping secondary block makes all VPC address ranges nonoverlapping without renumbering existing workloads. Cross-account VPC peering then provides direct private IPv4 connectivity while preserving source addresses. Bidirectional routes and appropriate security rules are required, including routes for future subnets that use the secondary ranges.

ANS-C01

A company connects 30 VPCs in one AWS Region through a transit gateway. Each VPC has a unique /16 CIDR block within 10.0.0.0/8. Application subnet route tables use a 10.0.0.0/8 route to the transit gateway for connectivity to the other VPCs.

Two VPCs exchange 200 TB of private IPv4 traffic each month, approximately equally in both directions. The communicating workloads are in the same Availability Zone, verified by Availability Zone ID. This traffic does not require centralized inspection or access to on-premises networks. The company must retain existing connectivity between all VPCs and cannot move workloads or introduce routing appliances.

Which change will minimize recurring network charges for the high-volume traffic while meeting these requirements?

Show answerCORRECT ANSWER

Create a VPC peering connection between the two VPCs. In both VPCs' application subnet route tables, add a route for the other VPC's /16 CIDR block through the peering connection. Retain the 10.0.0.0/8 routes to the transit gateway.

The more-specific /16 routes select the direct peering connection in both directions, bypassing transit gateway data processing for this traffic. Same-Availability-Zone private traffic over same-Region VPC peering has no data transfer charge, and peering has no hourly connection charge. The existing summary routes continue to provide connectivity to the other VPCs through the transit gateway.

Practice Network Peering in a full exam

Questions on Network Peering in each practice exam, with every answer option and an explanation for each:

Start a free practice exam

What is Network Peering

2 min read

Network peering establishes connectivity between separate networks while preserving their administrative boundaries. In cloud environments, this usually means connecting virtual networks so workloads can communicate using private IP addresses rather than public endpoints. Connections may span accounts, subscriptions, projects, or regions when the provider supports those combinations. Peering provides network reachability; it does not merge identity policies, automatically authorize application access, or make every resource reachable.

A working connection requires compatible address ranges, appropriate routes, and security rules that permit the intended traffic. For example, an application network using 10.10.0.0/16 might peer with a shared-services network using 10.20.0.0/16. The application subnet needs a route toward the shared-services destination, and the shared-services subnet needs a return route. Security groups, network security groups, firewalls, and network ACLs must also allow the relevant flows. Some providers install peering routes automatically, while others require explicit route-table updates.

Overlapping address ranges are a common obstacle because the routing system cannot reliably distinguish local destinations from remote destinations using the same addresses. AWS VPC peering rejects overlapping IPv4 CIDR blocks. Plan address allocation before connecting environments, especially after acquisitions or when teams create networks independently. When overlap cannot be removed, consider a deliberately designed translation solution or service-level private connectivity instead of assuming that peering will resolve it.

A critical distinction is direct connectivity versus transit connectivity. Many cloud peering implementations are non-transitive. If network A peers with B and B peers with C, A does not automatically gain access to C through B. With AWS VPC peering, adding routes does not turn B into a transit network. A peer also cannot simply use another VPC's internet gateway, NAT gateway, VPN connection, or Direct Connect connection as its gateway to external networks. Azure supports specific gateway-transit configurations, illustrating why these details must be checked for each platform.

Topology strongly affects cost and operational effort. Two networks with sustained private traffic may benefit from direct peering because it avoids a central transit service's attachment and processing charges. However, a full mesh of n networks requires n(n−1)/2 connections. Ten networks therefore require 45 peerings if every pair needs direct connectivity. A managed transit hub can simplify routing and centralized inspection at scale, but introduces its own service charges and routing configuration.

For AWS cost comparisons, VPC peering has no hourly connection charge, but applicable data-transfer charges still matter. Same-Availability-Zone traffic over same-Region VPC peering has no data-transfer charge; cross-Availability-Zone and inter-Region traffic can incur charges. AWS Transit Gateway commonly adds attachment-hour and data-processing charges. Compare actual traffic paths, volumes, regions, and Availability Zones rather than assuming either design is always cheaper.

Finally, private connectivity is not equivalent to application encryption or automatic DNS integration. Encryption characteristics and DNS resolution behavior depend on the platform and configuration. Apply TLS where required, configure private-name resolution separately, and test both forward and return paths. Route tables, flow logs, and reachability diagnostics help distinguish routing failures from firewall, DNS, or application failures.

Learn Network Peering in lessons

5 short lessons on Network Peering in 2 certification courses.

Open the lessons