# Configuring admin and user clusters

This section describes the configuration required for deploying admin and user clusters. This includes:

Common variables across all clusters
Default values that can be over-ridden on a cluster-by-cluster basis
General cluster configuration
Configuring static IPs for each cluster

Background reading material is available at https://cloud.google.com/anthos/gke/docs/on-prem/how-to/admin-user-cluster-basic.

# Common variables

The following variables are common across all clusters:

Variable	File	Description
`gke_bundle_path`	all.yml	The GKE on-prem bundle file contains all of the components in a particular release of GKE on-prem. Set the value of bundlepath to the absolute path of the admin workstation's bundle file. Do not change. `'var/lib/gke/bundles/gke-onprem-vsphere-<<anthos_version>>-full.tgz'`
`gke_data_disk_name`	all.yml	GKE on-prem creates a virtual machine disk (VMDK) to hold the Kubernetes object data for the admin cluster. The installer creates the VMDK for you, but you must provide a name for the VMDK (the `.vmdk` extension will be added). Any directory in the supplied path must be created before deployment. Not required when adding additional user clusters. For example, `'test-deploy'`.
`gke_network_internal`	all.yml	vSphere network for VMs

# Default configuration

These variables will be used for all clusters unless explicitly overridden in the gke_cluster_config:

Variable	File	Description
`gke_cluster_resource_pool`	all.yml	If you are using a non-default resource pool, provide the name of your vSphere resource pool. Provides a default value for the per-cluster field `resource_pool`. For example, `'Anthos_1.4'`
`gke_cluster_gcp_region`	all.yml	Provides a default value for the per-cluster fields `logging_metric_gcp_region` and `audit_logging_gcp_region`. For example, `'us-east1'`
`gke_cluster_antiAffinity_enable`	all.yml	Sets the `antiAffinityGroups.enabled` flag in the user cluster config. Spreads nodes across at least three physical hosts (requires at least three hosts) Defaults to `true`
`gke_loadBalancerKind`	all.yml	One of `F5BigIP`, `Seesaw` or `ManualLB`" Only F5BigIP is currently supported
`enable_vpc`	all.yml	Provides a default value for the per-cluster field `enable_vpc`. Set to `'true'` if you have your cluster's network controlled by a VPC. This ensures that all telemetry flows through Google's restricted IP addresses. One of `'true'` or `'false'`
`enable_cloudrun`	all.yml	Provides a default value for the per-cluster field `enable_cloudrun`. One of `'true'` or `'false'`

# Cluster configuration

General cluster config is supplied in the fields of the gke_cluster_config structure. The embedded structure gke_cluster_config.gkenode_net is used for configuring networking for each cluster.

# General cluster config `gke_cluster_config`

Field	Description	Value
`name`	Unique name for the cluster	For example, `usercluster-1`
`type`	Type of cluster to be configured	One of `admin` or `user`
`loadBalancerKind`	Type of load balancer	One of `F5BigIP`, `Seesaw` or `ManualLB` Only F5BigIP is currently supported
`f5_partition`	The playbooks for configuring F5 create partitions for the admin and user clusters based on this value.
`state`	Whether the F5 partition should exist or not.	One of `present` or `absent`
`cluster_ctrl_vip`	Set to the IP address that you have chosen to configure on the load balancer for the Kubernetes API server of the corresponding cluster	For example, `'172.17.0.21'`
`cluster_ingress_vip`	Set to the IP address you have chosen to configure on the load balancer for the cluster's ingress service
`datastore`	The datastore to use for the cluster. Defaults to the global `vcenter.datastore`
`resource_pool`	If you want to use a cluster-specific resource pool, provide the name of your vSphere resource pool	Defaults to value of `gke_cluster_resource_pool`.
`cluster_network`	If specified it overwrites the network field in global vCenter configuration
`logging_metric_gcp_region`	A GCP region where you would like to store logs and metrics for this cluster.	Defaults to value of `gke_cluster_gcp_region`
`audit_logging_gcp_region`	A GCP region where you would like to store audit logs for this cluster.	Defaults to value of `gke_cluster_gcp_region`
`enable_vpc`	Set to true if you have your cluster's network controlled by a VPC. This ensures that all telemetry flows through Google's restricted IP addresses.	One of `true` or `false`
`enable_cloudrun`	Specify Cloud Run configuration	One of `true` or `false`

# Configuring IPs for each cluster

The structure gke_cluster_config.gkenode_net is used for configuring networking for each cluster.

Field	Description	Value
`gke_cluster_config.gkenode_net.mode`	Config type being specified	One of `dhcp` or `static`

If you choose static, you can configure the static IPs using the other fields in the gke_cluster_config.gkenode_net structure.

You can specify the addresses of the DNS servers (dns), time servers (tod), and default gateway (gateway) and netmask (netmask) that the cluster nodes will use.

The search_domain field is a string of DNS search domains to use in the cluster. These domains are used as part of a domain search list.

The ips field is an array of IP addresses (ip) and hostnames (hostname). These are the IP addresses and hostnames that Anthos GKE on-prem will assign to the cluster nodes.

A sample structure is shown below:

      gkenode_net:
        mode: 'static'
        dns: '192.168.1.1'
        gateway: '192.168.1.1'
        netmask: '255.255.255.0'
        tod: 'hou-ntp1.hcilabs.hpecorp.net'
        search_domain: 'my.local.com'
        ips:
         - ip: 192.168.1.10
           hostname: admin-host1
         - ip: 192.168.1.11
           hostname: admin-host2
         - ip: 192.168.1.12
           hostname: admin-host3

# User cluster configuration - master and worker node

Configuration for the master and worker nodes in a user cluster are supplied in the fields of the gke_cluster_config structure.

Field	Description	Value
`masternode_cpu`	The number of virtual CPUs for the master nodes in the user cluster.	For example, `'6'`
`masternode_memory`	The number of megabytes of memory for the master nodes in the user cluster.	For example, `'8192'`
`masternode_replicas`	The number of master nodes in the user cluster. User cluster master nodes must have either 1 or 3 replicas.	Default is `'1'`
`workernode_cpu`	The number of virtual CPUs for the worker nodes in the user cluster.	Default is `'4'`
`workernode_memory`	The number of megabytes of memory for the worker nodes in the user cluster.	Default is `'8192'`
`workernode_replicas`	The number of worker nodes in the user cluster. User cluster master nodes must have at least 3 replicas.	Default is `'3'`

A sample configuration for the worker and master nodes in a user cluster is shown below:

      masternode_cpu: '6'
      masternode_memory: '16000'
      masternode_replicas: '1'
      workernode_cpu: '6'
      workernode_memory: '16000'
      workernode_replicas: '4'

# User cluster configuration - CSI and Service Mesh

Configuration of the CSI driver for user clusters is documented in the section CSI configuration.

Configuration of the Service Mesh for user clusters is documented in the section Service Mesh configuration.

← Configuring the admin workstation HPE SimpliVity configuration →