Which architecture best fit the requirements, considering its original requirement for AWS?

[gungkrisna]

I have a requirement to create a VPC diagram originally designed for AWS but now need to adapt it for implementation in GCP. Which design would satisfy this requirement, and how can I ensure its correctness? Here is the diagram URL

1. The architecture is high-availability (in AWS, it should have multi AZs).

2. Determine the appropriate IP address range for each subnet.

3. There are public and private subnets.

4. Use some components in the VPC, including:
   Internet gateway (I have no idea about the equivalent service in GCP)
   VPC peering

5. Design the architecture in the private data center.

6. The private data center can connect to resources in AWS.

[lawrencenelson]

Hi @gungkrisna,

Welcome to the Google Cloud Community!

Based on your screenshots, you have three designs from which to choose.

The first and second designs are almost identical; however, in the first design, both vpc-a and vpc-b have their own dedicated load balancers. This setup, in my opinion, is preferable to the second design where vpc-a and vpc-b share a single load balancer. Having separate load balancers, as seen in the first design, is advantageous if you wish to isolate the services running in each VPC. However, the potential drawbacks of the first design include added complexity and increased costs. Nonetheless, the second design might be more appropriate if you're seeking a simpler architecture and are willing to accept some level of interdependence between the VPCs.

Among all three designs, I would recommend the third design. It not only incorporates Cloud DNS, which is beneficial for domain name system management, but it also includes a standby database replica for higher availability.

I hope this helps. Thank you. 😃