Google Cloud VPC Network Peering allows internal IP address connectivity across two Virtual Private Cloud (VPC) networks regardless of whether they belong to the same project or the same organization.
VPC Network Peering enables you to connect VPC networks so that workloads in different VPC networks can communicate internally. Traffic stays within Google's network and doesn't traverse the public internet.
A VPC Peer must be configured and established on both sides. The VPC Peering module requires the network self link for the local peer as well as the remote peer. To get the respected network_self_link from both sides of the VPC, we can use different data resource blocks to find our desired project which in turn helps find the desired network_self_link in the VPC.
This example uses the data resource blocks to find the local peer project using the filter API to sift projects based on the labels assigned to the project and another data resource block to get the network self link from the project.
data "google_projects" "local_peer_project" {
filter = "labels.application_name=qserv labels.environment=dev"
}
data "google_compute_network" "local_network" {
project = data.google_projects.local_peer_project.projects[0].project_id
name = "qserv-dev-vpc"You may use the same example block to get the network_self_link for the remote peer by changing the appropriate labels for the filter and providing the correct network name.
There may be times where a hub and spoke design is desired where the hub project is peered with multiple spoke projects. The VPC Peering Module only peers between two projects. To have multiple peers, we need to create multiple tfvars files with a dedicated pipeline.
An example project is the butler-dev project where it's peerd with science-platform-dev and panda-dev. In the butler ./env directory there are two different tfvars files with two different pipelines.
A new pipeline needs to be created and deployed using the existing vpc peering pipeline as an example. Be sure to change the appropriate directory names and tfvars file names in the pipeline.