Transitgateway

AWS Transit Gateway

Overview

Domain expertise for configuring AWS Transit Gateway, the Regional network hub that connects many VPCs and on-premises networks through a single router instead of a mesh of point-to-point connections. Covers building the hub and attaching VPCs, segmenting traffic with route tables, centralizing egress and inspection, east-west inspection with AWS Network Firewall, hybrid connectivity over Site-to-Site VPN and Direct Connect, inter-Region peering, migrating off a VPC peering mesh, and IP multicast.

This skill is a router. Each customer task maps to a procedure file under references/. Read the matching reference in full before acting, then follow its constraints and steps. The reference files are self-contained: each carries its own decision tables, constraints, procedure, and troubleshooting.

Execute commands using the AWS MCP server when connected (sandboxed execution, audit logging, observability). Fall back to the AWS CLI otherwise. All CLI operations require least-privilege, ephemeral credentials (an assumed IAM role through AWS STS or AWS IAM Identity Center / SSO), never long-lived IAM user access keys. A transit gateway is a Regional resource: run each aws ec2 transit gateway command in the Region that holds the hub.

Which Transit Gateway task do you need?

Routing notes

• Decide segmentation before you build. “Default route table association” and “Default route table propagation” are on by default, which wires every attachment into one open mesh. If the customer plans isolated environments, the creating reference disables the defaults up front and hands off to the segmenting reference. Retrofitting isolation onto an open hub is a re-architect.
• North-south egress vs east-west inspection. Centralizing egress sends spoke traffic out to the internet through a central VPC. East-west inspection keeps traffic between spokes internal and forces it through a firewall on the way. They look similar but use different route table recipes. Match the reference to the direction of traffic the customer actually has.
• Appliance vs Gateway Load Balancer for inspection. Raw third-party appliances and a Gateway Load Balancer (GWLB) endpoint are two paths to the same goal. GWLB is the recommended approach for new designs. Both live in the centralizing-egress reference; appliance mode and the GWLB endpoint route table entries differ and the reference covers each.
• Appliance mode is required for stateful cross-Availability-Zone inspection, with a tradeoff. Appliance mode keeps each flow on one Availability Zone’s appliance so request and response do not split. It also disables cross-Availability-Zone failover for that attachment, so the inspection design must pair it with health-check-based failover. Both the egress and east-west references carry this.
• Transit gateway side vs Direct Connect side. The connecting-on-premises reference covers the transit gateway side: Site-to-Site VPN attachment options, route propagation, and equal-cost multi-path (ECMP). The Direct Connect gateway and virtual interface setup belongs to the separate directconnect skill. Do not restate the Direct Connect side here.

Security considerations

A transit gateway is the central routing point for many VPCs and on-premises networks, so a misconfiguration here has blast radius across every attached network. Apply these controls regardless of the specific task; each per-task reference carries the detail.

• You MUST enable Transit Gateway Flow Logs for traffic visibility, audit, and incident response across the hub, and MUST enable encryption at rest on the destination (a KMS key on the CloudWatch log group, or SSE-KMS on the S3 bucket).
• You MUST, when a KMS key encrypts a flow log destination (CloudWatch log group or S3 bucket) or a CloudTrail destination, scope the KMS key policy with condition keys (aws:SourceArn, aws:SourceAccount, and kms:ViaService) so only the specific log group, bucket, or trail in the expected account and service can use the key, preventing cross-account or cross-service misuse.
• You SHOULD apply least-privilege IAM for transit gateway administration, avoiding service wildcards and FullAccess policies, restricting who can create attachments, modify route tables, and change associations or propagations.
• You SHOULD ensure Site-to-Site VPN tunnels use strong encryption (for example AES-256-GCM with IKEv2) and enable tunnel logging to CloudWatch Logs with encryption enabled (a KMS key) to protect sensitive connection state and IKE negotiation detail from unauthorized access (see the connecting on-premises networks reference).
• You MUST treat a misconfigured transit gateway route table as a security risk, since wrong associations or propagations can expose workloads across environments meant to stay isolated (see the segmenting traffic reference).
• You MUST enable AWS CloudTrail to detect unauthorized changes to transit gateway route tables, associations, and propagations, MUST enable encryption at rest on the CloudTrail destination (a KMS key), and use AWS Config rules to detect drift from the intended design.

Additional Resources

• AWS Transit Gateway Guide (opens in a new tab)
• AWS Transit Gateway product page (opens in a new tab)
• AWS Transit Gateway pricing (opens in a new tab)