Use case 1: Interconnect two or more VPCs in the same MZR

Connect two VPCs in the same region with a local transit gateway.

Use case 2: Interconnect two or more VPCs across multiple MZRs

Connect VPCs in multiple regions by using a global transit gateway.

For higher resiliency, you can deploy multiple global transit gateways across regions and group them by using redundancy groups.

Use case 3: Interconnect one or more VPCs in the same MZR and an IBM classic network

Connect VPCs in the same region with IBM Cloud classic through a local transit gateway.

Use case 4: Interconnect VPCs and an IBM classic network to access all your resources across all MZRs

Connect VPCs from multiple regions with IBM Cloud classic through a global transit gateway.

Use case 5: Interconnect VPCs across accounts

Connect VPCs in the same region owned by different IBM Cloud accounts through a local transit gateway.

Use case 6: Connect networks (VPC and classic) to multiple local gateways

Keep in mind:

• Your local traffic is kept on a local transit gateway, which reduces latency.
• Highly Available (HA) capabilities are provided, as data in VPCs C and D might be replicated in VPCs in E and F.
• Classic infrastructure transit gateway connections are required to be in the same account as the transit gateway owner.

Use case 7: Interconnect networks (VPC and classic) across accounts

Connect cross-account IBM Cloud classic accounts to one or more transit gateways. To do so, the IBM Cloud account that owns the transit gateway requests permission from the IBM Cloud classic account to connect it to the transit gateway. The IBM Cloud classic account must approve the request before the connection is made. You can repeat this process for multiple IBM Cloud classic account connections as shown.

Use case 8: Connect networks by using a High Availability GRE tunnel

Connect IBM Cloud classic infrastructure by using a GRE tunnel to a local transit gateway.

This diagram shows a highly available GRE tunnel configuration. When you set up a GRE tunnel configuration, an availability zone must be specified. To make this use case highly available, you must set up two GRE tunnels with the same endpoints, but by using different availability zones.

Transit gateway GRE connections require the gateway owner to specifically configure HA for their needs. A GRE connection is a point-to-point connection, has no built-in redundancy, and is a single point of failure. When you configure a GRE connection on a transit gateway, you must specify the availability zone. For a robust HA solution, configure multiple GRE connections by using different availability zones.

Use case 9: Connect an on-premises network by using Direct Link

Connect IBM Cloud Direct Link to allow on-premises connectivity to IBM Cloud networks through a transit gateway. This connection allows the on-premises network to access all networks that are connected to the transit gateway. In the following example, the Direct Link gateway connects to a global transit gateway, along with 4 VPCs and IBM Cloud Classic Infrastructure. The inverse is also true, in that all other networks that are connected to the transit gateway are now connected to the on-premises network.

Direct Link can be connected to either local or remote transit gateways.

Use case 10: Using VPN for VPC as a redundant spoke

In this use case, a VPN gateway is configured as a spoke to the transit gateway to provide a redundant network path between the on-premises environment and IBM Cloud. The primary connectivity is established using Direct Link, which offers private, dedicated connectivity with assured bandwidth and low-latency performance. This direct connection terminates in the IBM Cloud Transit VPC, which connects to the transit gateway. From there, traffic is routed to multiple environments including Power Virtual Servers, virtual server instances in spoke VPCs, and classic infrastructure.

To enhance availability without the added cost of a second direct link, a VPN gateway is deployed as a secondary path. This path provides secure connectivity over the internet, using BGP VPN tunnels between the on-prem network and the VPN gateway. Multiple GRE tunnels are established between the VPN gateway appliances and the transit gateway routers, allowing the VPN to act as a spoke within the Transit Gateway topology. While this path doesn't provide the same performance guarantees as Direct Link, it ensures continued connectivity during a primary path failure, making it a cost-effective and resilient solution for hybrid cloud networking.

Use case 11: Build a highly available multi-region network

Deploy two global transit gateways in separate regions and group them into a redundancy group to create a resilient network architecture. In this example, Global Transit GW A is deployed in eu-es and Global Transit GW B is deployed in br-sao. Both gateways are members of the same redundancy group.

Two VPCs under Account A are connected to both transit gateways:

• VPC A (us-south, 10.4.0.0/16) — connects locally to Transit GW A or B, and reaches VPC B (10.5.0.0/16) via Transit GW A or B
• VPC B (us-east, 10.5.0.0/16) — connects locally to Transit GW A or B, and reaches VPC A (10.4.0.0/16) via Transit GW A or B

Because each VPC is connected to both transit gateways, traffic can be routed through either gateway. If one region becomes unavailable, the other gateway continues to forward traffic between the VPCs without interruption.

Power Virtual Server use cases using Transit Gateway

For use cases involving Power Virtual Server workspaces, see Power Edge Router use cases.