Configuring the Network Load Balancer

The next step is to give a name to the load balancer, and to assign it to the Region and Network according to your Google Cloud worksheet.

Backend Configuration

Each node pool group requires a backend configuration created for it. GKE creates an instance group for each zone where the node pool was deployed, therefore if the deployment was done in a single zone, only one backend needs to be configured. If the deployment was done in a multi-zone, it would need a backend for each configured zone.

A new backend must have the following selections made for it:

• IP Stack type: must be IPv4 (single-stack)

• Instance group: if configured as a single zone deployment, the Instance group box will contain a single option. If configured as a multi-zone deployment, it will contain all the instance groups created by GKE. Use the following command to identify the correct instance group according to region:

  gcloud container clusters describe <CLUSTER_NAME> --region <REGION>  --format json | jq -r .nodePools[].instanceGroupUrls[] | awk -F'/' '{print $NF}'

  Where:

  <CLUSTER_NAME> is the cluster created in Google Kubernetes Engine Cluster

  <REGION> is the cluster compute region (check the Google Cloud worksheet)

  Example commands and outputs:

  For single zone deployments:

  gcloud container clusters describe gcp-arcsight-test-gks2  --zone us-central1-a --format json | jq -r .nodePools[].instanceGroupUrls[] | awk -F'/' '{print $NF}'

  gke-gcp-arcsight-tes-gcp-arcsight-tes-86d1d1b9-grp

  For multi-zone deployments:

  gcloud container clusters describe th-test-gke --region us-central1 --format json | jq -r .nodePools[].instanceGroupUrls[] | awk -F'/' '{print $NF}'

  gke-th-test-gke-default-pool-a9301022-grp
  gke-th-test-gke-default-pool-5792242b-grp
  gke-th-test-gke-default-pool-76aca1cd-grp

• Health check: the Health Check can be created anew, or an already available one can be used as long as it has the following settings and health criteria:

  • Path: /healthz

  • Protocol: HTTP

  • Port: 10256

  • Interval: 8 seconds

  • Timeout: 1 second

  • Healthy threshold: 1 success

  • Unhealthy threshold: 3 consecutive failures

Frontend Configuration

Access must be granted to the following cluster external communication ports:

Port (s)	Description
443	Used by Transformation Hub, ArcMC, Fusion, etc., for UI access
32080	Used for Transformation Hub and ArcMC communication
32081	Used by the Schema Registry as the external communication port to serve HTTP requests (to provide Schemas information for external Avro consumers)
9092	Optional - only needs to be opened if Transformation Hub is configured to accept connections over a clear text channel
9093	Used by Transformation Hub as a requirement for secure communications with clients
32101 – 32150	Optional - these ports are needed only if Connectors are be deployed in Transformation Hub

Go to Frontend Configuration and select New Frontend IP and port.

For each necessary port, the New Frontend IP and port options must be set as follows:

• Name and Description: optional but recommended to identify each port

• IP version: IPv4

• Subnetwork: private-subnet

• Internal IP purpose: shared. The IP address entered here must be the same for all the ports configured

• Ports: single

• Port number: the port number being configured (from the table above)

• Global access: disabled

Click Done before proceeding with the next port.