# Hub-and-Spoke Topology

Hub-and-spoke topology is a network design in which multiple peripheral networks connect to a central hub that controls connectivity between them and shared destinations.

Cloud concept. Updated 2026-10-29. Page: https://blowstack.com/tools/cloud-concepts-hub/hub-and-spoke-topology

> The hub provides a common routing point for otherwise separate networks, reducing the number of connections needed compared with a full mesh. In cloud architectures, this simplifies connectivity across accounts, regions, and on-premises environments, but introduces cost, resilience, and performance trade-offs around the shared transit layer.

Practice exams covering Hub-and-Spoke Topology: [ANS-C01 AWS Certified Advanced Networking - Specialty](https://blowstack.com/practice-exams/aws-certified-advanced-networking-specialty), [SAA-C03 AWS Certified Solutions Architect - Associate](https://blowstack.com/practice-exams/aws-certified-solutions-architect-associate), [SAP-C03 AWS Certified Solutions Architect - Professional](https://blowstack.com/practice-exams/aws-certified-solutions-architect-professional).

## 6 key aspects of Hub-and-Spoke Topology

Hub-and-spoke networking simplifies connectivity through shared transit while requiring deliberate segmentation, routing, capacity planning, resilience, and cost analysis.

1. **Centralized transit**: Spokes use a shared routing hub to reach permitted networks and shared destinations. Distinguish a true transit-capable hub from a central network connected only through nontransitive peering.
2. **Scalable connectivity management**: Adding a spoke usually requires a connection to the hub rather than a new connection to every existing network. Route limits, attachment quotas, and address planning still constrain growth.
3. **Explicit segmentation**: Separate routing domains can allow access to shared services while preventing production-to-development connectivity. Apply security controls in addition to routing policy, and verify both forward and return paths.
4. **Shared hybrid and network services**: Multiple spokes can use common connectivity to data centers, inspection infrastructure, or internet egress. Design these shared dependencies for sufficient capacity and an acceptable failure impact.
5. **Traffic-dependent economics**: Evaluate attachment, processing, transfer, NAT, and appliance costs using expected traffic volumes. A hub reduces connection-management complexity but can cost more than direct peering for a small or high-volume network.
6. **Regional placement and resilience**: Place hubs near workloads and connect regional hubs when global reach is needed. Test failover and preserve symmetric paths through stateful appliances rather than assuming centralization guarantees availability.

## Cloud services for Hub-and-Spoke Topology

Amazon VPC supplies workload spokes, while AWS Transit Gateway provides regional transit and routing segmentation. Direct Connect and Site-to-Site VPN extend that topology to data centers and branches. AWS Network Firewall can inspect selected flows when routing deliberately sends them through an inspection VPC. For a global architecture, AWS Cloud WAN provides a managed core with regional connectivity and centralized network policies. These services do not remove the need for address planning, return routes, security rules, or resilient hybrid links. Compare the combined attachment, processing, transfer, and inspection costs against simpler direct-peering designs before centralizing traffic.

- **[AWS Transit Gateway](https://blowstack.com/tools/cloud-services-hub/aws-transit-gateway)**: Provides a regional transit hub for VPC and hybrid network connectivity. Transit gateway route tables support segmentation, while peering can connect regional hubs; attachment routes and VPC routes must still be configured correctly.
- **[Amazon VPC](https://blowstack.com/tools/cloud-services-hub/amazon-vpc)**: Provides the isolated networks commonly used as spokes or as shared-services and inspection networks. VPC peering supports direct network connections but is nontransitive, so a central peered VPC does not automatically route between spokes.
- **[AWS Direct Connect](https://blowstack.com/tools/cloud-services-hub/aws-direct-connect)**: Supplies dedicated connectivity between customer networks and AWS. A Direct Connect gateway associated with transit gateways can extend shared hybrid connectivity to multiple spokes; resilient designs require redundant connectivity.
- **[AWS Site-to-Site VPN](https://blowstack.com/tools/cloud-services-hub/aws-site-to-site-vpn)**: Connects on-premises or branch networks to a cloud transit hub using IPsec tunnels. It can provide primary hybrid connectivity or a backup path, with routing and failover configured for the intended behavior.
- **[AWS Network Firewall](https://blowstack.com/tools/cloud-services-hub/aws-network-firewall)**: Can provide managed traffic inspection in a dedicated inspection VPC connected to the hub. Routes must explicitly steer traffic through firewall endpoints, and stateful inspection requires symmetric traffic paths.

## Hub-and-Spoke Topology exam questions

Real practice exam questions on Hub-and-Spoke Topology, with the correct answer and its explanation.

### Question 1 (SAA-C03)

**Question:**

A company has six business units in one AWS Region. Each business unit operates four application VPCs connected to its own AWS Transit Gateway. A shared-services VPC is attached to all six transit gateways. All VPC CIDR blocks are nonoverlapping.

Applications must communicate with other applications in the same business unit and with the shared-services VPC. Routing must prevent direct communication between different business units. The shared-services VPC does not forward traffic between application VPCs. The company wants to reduce recurring attachment charges and simplify connection management without operating routing appliances. Traffic volumes and applicable data transfer rates will remain unchanged.

Which architecture meets these requirements with the LOWEST recurring attachment cost?

**Correct answer:** Consolidate onto one transit gateway with one attachment per VPC. Use a separate transit gateway route table for each business unit, containing routes only to that unit's VPCs and shared services. Associate the shared-services attachment with a route table containing all application VPC routes. Configure corresponding VPC routes.

**Explanation:** A single managed routing hub can maintain separate routing domains. Each business unit's attachment-associated route table permits communication within that unit and with shared services, but has no routes to other units. The shared-services route table provides return connectivity. Consolidation reduces the shared-services VPC from six attachments to one while retaining one attachment per application VPC.

### Question 2 (SAP-C03)

**Question:**

A company operates 40 VPCs with nonoverlapping CIDR ranges in one AWS Region. Each VPC currently has two AWS Site-to-Site VPN connections attached to its virtual private gateway: one connection to each of two independent on-premises routers. The routers use separate internet providers. The company must retain connectivity through either router if the other router or provider fails.

On-premises systems and VPC workloads must be able to initiate connections to each other by using arbitrary private IP protocols. VPC-to-VPC connectivity is not required. The company wants to reduce recurring network charges and the number of hybrid connections it manages without deploying routing appliances or application proxies.

Use the following pricing assumptions:
• Each Site-to-Site VPN connection costs $0.05 per hour.
• Each Transit Gateway attachment, including a VPN attachment, costs $0.05 per hour.
• Transit Gateway data processing costs $0.02 per billable GB. Any proposed transit gateway design would process a total of 2,000 billable GB per month across all its gateways.
• Use 730 hours per month. All other charges are identical or can be ignored, and each proposed VPN connection has sufficient capacity.

Which solution meets the connectivity and resilience requirements at the LOWEST recurring network cost?

**Correct answer:** Create one transit gateway and attach all 40 VPCs. Establish one Site-to-Site VPN connection from the transit gateway to each on-premises router. Configure VPC, transit gateway, and on-premises routing for both paths, and remove the existing per-VPC VPN connections.

**Explanation:** A transit gateway provides a shared routing hub for the VPCs and both VPN connections, replacing 80 separately managed hybrid connections with two while retaining independent on-premises paths. The design requires 40 VPC attachments, two VPN attachments, and two VPN connections. Monthly cost is (44 × $0.05 × 730) + (2,000 × $0.02) = $1,646, the lowest cost among the designs that meet all requirements.

### Question 3 (ANS-C01)

**Question:**

A company operates 28 application VPCs and one operations VPC in a single AWS Region. All VPC CIDR blocks are nonoverlapping. Monitoring collectors in the operations VPC initiate TCP and UDP connections to agents in each application VPC. The agents also initiate connections to ingestion servers in the operations VPC. Security controls require the receiving workloads to see the originating workload's private IP address.

The application VPCs must not have routes to one another. No VPC requires connectivity to on-premises networks or other Regions, and these requirements are not expected to change. Existing service quotas support the proposed connections. The company wants the lowest recurring connectivity cost without introducing customer-managed forwarding infrastructure. Applicable inter-Availability Zone data transfer charges would be the same for all proposed designs.

Which architecture best meets these requirements?

**Correct answer:** Create a VPC peering connection between each application VPC and the operations VPC. Configure reciprocal routes only for the operations VPC and the corresponding application VPC CIDR blocks.

**Explanation:** A peering-based star supports direct, bidirectional connectivity to workloads in the central VPC while preserving source private IP addresses. Peering is nontransitive, so it does not provide connectivity between application VPCs through the operations VPC. VPC peering has no attachment-hour or data-processing charges, making it less expensive than a transit gateway for these requirements, given the stated data-transfer assumptions.

## Practice Hub-and-Spoke Topology in a full exam

Questions on Hub-and-Spoke Topology in each practice exam, with every answer option and an explanation for each:

- [ANS-C01](https://blowstack.com/practice-exams/aws-certified-advanced-networking-specialty): 12 questions
- [SAP-C03](https://blowstack.com/practice-exams/aws-certified-solutions-architect-professional): 3 questions
- [SAA-C03](https://blowstack.com/practice-exams/aws-certified-solutions-architect-associate): 2 questions

## What is Hub-and-Spoke Topology

A hub-and-spoke topology connects peripheral networks, called spokes, through a central networking hub. Spokes commonly represent application networks, business units, environments, or remote sites. The hub can be a managed transit router or a routing infrastructure that provides shared connectivity. Its purpose is to simplify network relationships and apply routing policy centrally; it does not require every spoke to communicate with every other spoke.

The scaling advantage becomes clear when comparing it with a full mesh. Connecting n networks directly to every other network requires n(n−1)/2 connections. Connecting those networks to a separate hub requires approximately n spoke connections. This reduces connection management as the environment grows, although route tables, security rules, address planning, and operational ownership still require careful design.

In the cloud, an organization might connect production, development, and shared-services virtual networks to a managed transit hub. Production and development can each reach shared DNS or identity services while remaining isolated from one another. Hub route-table associations, route propagation, and explicit routes determine which destinations are reachable. Virtual-network and subnet routes must also direct traffic toward the hub, and return routes must exist. Security groups, network access controls, and firewalls remain separate controls; routing connectivity alone does not authorize application access.

A common misconception is that connecting virtual networks to a central virtual network using peering automatically creates transit routing. AWS VPC peering is nontransitive: if network A peers with B and B peers with C, A cannot reach C through B using those peering connections. A peering-based star can support independent access to services hosted in the central VPC, but it is not a general-purpose transit hub. For a few networks that need direct connectivity, peering can be simpler and less expensive than introducing a managed transit service.

Hybrid architectures extend the pattern by connecting on-premises networks to the hub through VPN or private connectivity. For example, several application VPCs can share access to a corporate data center through a transit gateway connected to a Direct Connect gateway. Address ranges should be nonoverlapping for straightforward routed connectivity. Overlapping ranges require additional mechanisms, such as translation or service-specific access, rather than an assumption that the hub will resolve the conflict.

Centralization is not automatically cheaper or faster. Compare attachment charges, data-processing charges, inter-Region and applicable cross-Availability-Zone transfer charges, appliance capacity, and operational effort. Centralized internet egress can reduce the number of NAT gateways, but traffic processing and transfer costs can offset those savings. Sending traffic through an inspection VPC also adds forwarding stages and requires symmetric routing for stateful firewalls. A transit hub does not inspect traffic unless inspection services are explicitly placed in the path.

For global deployments, use regional hubs connected through an inter-Region backbone or hub peering rather than forcing all traffic through one distant region. Validate bandwidth, latency, route limits, and failover behavior, and distribute customer-managed routing or inspection appliances across Availability Zones. Managed hub availability does not make every attached dependency resilient. Multi-tier applications can keep web, application, and database tiers within one virtual network; the hub usually connects larger network boundaries rather than replacing tier-level security. Choose the topology according to traffic patterns, isolation requirements, growth, and total cost, not merely the number of networks.

## Learn Hub-and-Spoke Topology in lessons

5 short lessons on Hub-and-Spoke Topology in 2 certification courses.

- SAA-C03: Hub-and-Spoke Topology: spoke reachability (Premium), https://app.blowstack.com/lessons/hub-and-spoke-topology-spoke-reachability-saa-c03
- SAP-C03: Hub-and-Spoke Topology: segmented transit routing (Premium), https://app.blowstack.com/lessons/hub-and-spoke-topology-segmented-transit-routing-sap-c03
- SAA-C03: Hub-and-Spoke Topology: transit-path trade-offs (Premium), https://app.blowstack.com/lessons/hub-and-spoke-topology-transit-path-trade-offs-saa-c03
- SAP-C03: Hub-and-Spoke Topology vs Network Peering (Premium), https://app.blowstack.com/lessons/hub-and-spoke-topology-vs-network-peering-sap-c03
- SAA-C03: Hub-and-Spoke Topology vs Network Peering (Premium), https://app.blowstack.com/lessons/hub-and-spoke-topology-vs-network-peering-saa-c03
