What is Peering?

Peering is a service that connects two virtual private clouds (VPCs) to synchronize data between them. The two VPCs that are connected through a peering connection act like one VPC. You can create a peering connection between two VPCs that are created in the same or different accounts in the same region or across different regions. You can configure routing policies in the two VPCs that are connected through a peering connection to transfer traffic between them. Peering does not rely on independent hardware, and is therefore free from single points of failure (SPOFs) and bandwidth bottlenecks.

Limits

1. The two VPCs that are connected through a peering connection cannot overlap in CIDR blocks.
2. The peer CIDR blocks of the tunnel gateway in one VPC cannot overlap the CIDR block of the peer VPC.

Assume that you create a peering connection between VPCs A and B.

The peer CIDR blocks of the tunnel gateway in VPC A are A1 and A2, and the peer CIDR blocks of the tunnel gateway in VPC B are B1 and B2.

• VPCs A and B cannot overlap in CIDR blocks.
• The CIDR block of VPC A cannot overlap CIDR blocks B1 and B2.
• The CIDR block of VPC B cannot overlap CIDR blocks A1 and A2.

Notes:

• Peering connectivity is nontransferable. Assume that you create a peering connection between VPCs A and B and another one between VPCs A and C. VPCs B and C cannot transfer traffic to each other.
• To enable the two VPCs of a peering connection to communicate with each other, you must configure routing rules directing traffic to the peer end in the route tables of both VPCs.

Local peering and cross-region peering

You can create a peering connection between two VPCs that are created in the same region or across different regions. The two peering modes vary in features and billing rules because of the different physical distances between two VPCs and different underlying implementation architectures.

• Local peering is used to pipeline applications in different VPCs in the same region.

• Cross-region peering is typically applied to the following scenarios: cross-region disaster recovery (DR), interconnection of VPCs in different regions, rapid deployment of the geo-redundancy DR solution (three data centers in two geographic locations), large-bandwidth high reliability, and financial-level network DR.