Manage GDC Edge infrastructure

This page covers managing GDCE blueprints and deployments, and creating and deleting a GDC Edge cluster.

GDCE Workload cluster blueprint

The GDC Edge Workload cluster blueprint contains custom resources you can combine to specify the intent to deploy a GDC Edge cluster. For an example Workload cluster custom resource definition, see Example WorkloadCluster CRD.

The following list defines the role of each custom resource:

  • WorkloadCluster: creates a local control plane based GDCE cluster with one node pool and VPN Connection between the created GDCE cluster and Telecom Network Automation Orchestration cluster's VPC.

  • ClusterInit: initializes the cluster's lifecycle that Telecom Network Automation manages. This lifecycle includes installing Telecom Network Automation components on the GDCE cluster required for the lifecycle management.

  • SourceRepoRepository and Repository: sets up the GitOps pipeline for the created GDCE cluster using Cloud Source Repository and Porch along with ConfigSync, installed from the previous resource.

  • ComputeFirewall: creates a firewall rule in the Telecom Network Automation Orchestration cluster's VPC to allow for TCP ingress traffic on port 3000 that comes from the same IP address range used as the GDCE Cluster's pod address range.

  • ImagePull: sets up the required credentials on the GDCE cluster for kubelet to pull images from the artifact registry.

When creating a deployment you can customize the blueprint using following parameters in the setters.yaml file:

Parameter Description
project-id The ID of the target Google Cloud project where the
Telecom Network Automation. orchestration cluster is created.
cluster-name A unique name that identifies a given cluster. This name must be unique across all deployments.
username The name of the user account within the target Google Cloud project authorized to modify cluster resources.
cluster-ipv4-cidr-blocks The IPv4 CIDR block for Kubernetes Pods that run on the cluster.
services-ipv4-cidr-blocks The IPv4 CIDR block for Kubernetes Services that run on the cluster
cluster-location The Google Cloud region where you want to create the cluster.
nodepool-name A unique name that identifies the node pool. This name must be unique across all deployments.
node-count The number of nodes you want the node pool to hold.
scale-down-grace-period-seconds Specifies the time period in seconds to wait for the NF workloads to migrate from the nodes being removed during scale down. For example, node-count is reduced during an update. This process is to ensure a non-disruptive scale down of the cluster. Note: This profcessis only applicable when the NF workloads are using the required pod label, infra.tna.cloud.google.com/nf-workload, for Telecom Network Automation to identify the critical NF workloads.
edge-zone-name The name of the target GDC Edge zone. This zone must be in the region specified in cluster-location.
vpn-connection-name A unique name that identifies the VPN connection. This name must be unique across all deployments.
vpc-name The VPC network name of the
Telecom Network Automation Orchestration cluster. If you opted for Enable management config during the Orchestration cluster creation, this must be the VPC network name. Otherwise, the default VPC network that Telecom Network Automation creates is tnavpc-01.
nat-gateway-ip The NAT gateway IP address for the target cluster
image-registry-server The image registry server where GDCE workload images are stored. The images are stored in us-central1-docker.pkg.dev.
control-plane-node-count The number of nodes to run the local control plane workloads. Valid values are 3 for high availability and 1 for standard operation.
control-plane-shared-deployment-policy Specifies whether application workloads can run on the nodes that run the local control plane workloads. Valid values are DISALLOWED and ALLOWED.
external-lb-ipv4-address-pools A comma-delimited list of IPv4 addresses, address ranges, or subnetworks for ingress traffic on services that run behind the GDC Edge load balancer. The range must belong to the load balancer IP subnet pool mentioned during the rack turnup.
targetVersion Specifies the GDC Edge software version that you want this cluster to run in the format 1.X.Y where X is the minor version, and Y is the patch version. For example 1.5.0.

GDC Edge blueprints

Create a blueprint

  1. From the navigation menu, click Blueprints.
  2. In the drop-down menu, select your preferred Orchestration cluster.
  3. Click Import from Library.
  4. Select the blueprint, GDCE Workload cluster blueprint.
  5. Click Add Blueprint.
  6. On the next screen, specify an appropriate name. For example, workloadcluster1 for the GDCE cluster blueprint.
  7. Optionally, add any labels.
  8. Afterwards, click Create.
  9. The blueprint appears in the Draft Blueprints list.

Edit a blueprint

  1. From the navigation menu, click Blueprints > Drafts Blueprint.
  2. Select your preferred blueprint. A blueprint editor appears.
  3. Edit the blueprint. Any changes to the particular blueprint at the current stage are discouraged.

Propose a blueprint

  1. From the navigation menu, click Blueprints > Drafts Blueprint.
  2. Select your preferred blueprint. A blueprint editor appears.
  3. Click Propose. The status of the blueprint updates to Proposed.

Approve a blueprint

  1. From the navigation menu, click Blueprints.
  2. In your preferred blueprint, click Action.
  3. Select Approve. The blueprint appears in Approved Blueprints.

GDC Edge deployment

Create a blueprint deployment

  1. From the navigation menu, click **Blueprints > Approved Blueprints.
  2. Select the name of your preferred blueprint.
  3. Click Create Deployment.
  4. Provide a name for the deployment. For example, workloadcluster1-deploy1.
  5. Click Create Deployment. The deployment appears in Draft Deployment.

Edit a blueprint deployment

  1. From the navigation menu, click Blueprints.
  2. Select the name of your preferred deployment. A deployment editor appears.
  3. Edit the setters.yaml file.
    1. Note: Ensure you carefully edit the fields and make the necessary changes to the values. For a description of these fields, see GDCE Workload cluster blueprints.
  4. Click "Save File".

Apply a blueprint deployment

  1. From the navigation menu, click Blueprints.
  2. Select the name of your preferred deployment. A deployment editor appears.
  3. Click Deploy. A dialog appears.
  4. In the dialog, click Deploy. The deployment appears in Applied Deployment.

Scale up and down GDCE cluster

  1. From the navigation menu, click Blueprints > Applied Deployments.
  2. Select your preferred deployment.
  3. Click View.
  4. Click Edit Deployment.
  5. Increase or decrease the node-count value from the setters.yaml file. When decreasing node-count set the scale-down-grace-period-seconds field to let NF workloads to migrate from the node being removed before scaling down.
  6. Click Save File. The deployment appears in Draft Deployment.
  7. Navigate to Deployments.
  8. In Draft Deployment, select the name of your preferred deployment. A deployment editor appears.
  9. Click Deploy.
  10. Confirm the deployment. The deployment appears in Applied Deployment.

Delete GDC Edge cluster

  1. From the navigation menu, click Blueprints.
  2. From the drop-down menu, Select your preferred Orchestration cluster.
  3. Click Applied Deployment.
  4. Click the Action icon for the deployment.
  5. Click Delete. A confirmation dialog appears.
  6. Confirm the deletion by clicking Delete.

The deletion process might complete in up to 30 minutes.