Skip to main content

GCP Compute Network Endpoint Group

A Google Cloud Compute Network Endpoint Group (NEG) is a first-class object that aggregates a set of individual network endpoints—such as VM IP addresses and ports, Cloud Run revisions or Cloud Functions—which can then be referenced by Google Cloud Load Balancers. NEGs give you fine-grained control over which back-end endpoints receive traffic and allow load balancers to perform health-checks directly against those endpoints rather than against an entire instance group.
For a complete description see the official documentation: https://cloud.google.com/load-balancing/docs/negs

Terrafrom Mappings:

  • google_compute_network_endpoint_group.name

Supported Methods​

  • GET: Get a gcp-compute-network-endpoint-group by its "name"
  • LIST: List all gcp-compute-network-endpoint-group
  • SEARCH

gcp-cloud-functions-function​

Serverless Network Endpoint Groups are automatically created for each Cloud Function that is configured to be reachable behind a load balancer. Overmind links the NEG back to the originating gcp-cloud-functions-function, showing you which function will receive traffic through this group.

gcp-compute-network​

For VM-based (GCE) NEGs, the endpoints must reside in a specific VPC network. Overmind relates the NEG to the gcp-compute-network so you can trace load-balanced traffic paths within the wider network topology.

gcp-compute-subnetwork​

Regional VM NEGs are also bound to a particular subnetwork. Linking the NEG to its gcp-compute-subnetwork helps you understand which CIDR ranges and firewall rules apply to the back-end endpoints.

gcp-run-service​

Each Cloud Run revision exposed through a load balancer appears as a serverless NEG. Overmind associates the NEG with the relevant gcp-run-service, making it clear which Cloud Run service instances are part of the group.