NodeConfig

Parameters that describe the nodes in a cluster.

GKE Autopilot clusters do not recognize parameters in NodeConfig. Use AutoprovisioningNodePoolDefaults instead.

JSON representation
{
  "machineType": string,
  "diskSizeGb": integer,
  "oauthScopes": [
    string
  ],
  "serviceAccount": string,
  "metadata": {
    string: string,
    ...
  },
  "imageType": string,
  "labels": {
    string: string,
    ...
  },
  "localSsdCount": integer,
  "tags": [
    string
  ],
  "preemptible": boolean,
  "accelerators": [
    {
      object (AcceleratorConfig)
    }
  ],
  "diskType": string,
  "minCpuPlatform": string,
  "workloadMetadataConfig": {
    object (WorkloadMetadataConfig)
  },
  "taints": [
    {
      object (NodeTaint)
    }
  ],
  "sandboxConfig": {
    object (SandboxConfig)
  },
  "nodeGroup": string,
  "reservationAffinity": {
    object (ReservationAffinity)
  },
  "shieldedInstanceConfig": {
    object (ShieldedInstanceConfig)
  },
  "linuxNodeConfig": {
    object (LinuxNodeConfig)
  },
  "kubeletConfig": {
    object (NodeKubeletConfig)
  },
  "bootDiskKmsKey": string,
  "gcfsConfig": {
    object (GcfsConfig)
  },
  "advancedMachineFeatures": {
    object (AdvancedMachineFeatures)
  },
  "gvnic": {
    object (VirtualNIC)
  },
  "spot": boolean,
  "confidentialNodes": {
    object (ConfidentialNodes)
  },
  "resourceLabels": {
    string: string,
    ...
  },
  "loggingConfig": {
    object (NodePoolLoggingConfig)
  },
  "windowsNodeConfig": {
    object (WindowsNodeConfig)
  },
  "localNvmeSsdBlockConfig": {
    object (LocalNvmeSsdBlockConfig)
  },
  "ephemeralStorageLocalSsdConfig": {
    object (EphemeralStorageLocalSsdConfig)
  },
  "soleTenantConfig": {
    object (SoleTenantConfig)
  },
  "resourceManagerTags": {
    object (ResourceManagerTags)
  },
  "enableConfidentialStorage": boolean,
  "secondaryBootDisks": [
    {
      object (SecondaryBootDisk)
    }
  ],
  "fastSocket": {
    object (FastSocket)
  }
}
Fields
machineType

string

The name of a Google Compute Engine machine type

If unspecified, the default machine type is e2-medium.

diskSizeGb

integer

Size of the disk attached to each node, specified in GB. The smallest allowed disk size is 10GB.

If unspecified, the default disk size is 100GB.

oauthScopes[]

string

The set of Google API scopes to be made available on all of the node VMs under the "default" service account.

The following scopes are recommended, but not required, and by default are not included:

  • https://www.googleapis.com/auth/compute is required for mounting persistent storage on your nodes.
  • https://www.googleapis.com/auth/devstorage.read_only is required for communicating with gcr.io (the Google Container Registry).

If unspecified, no scopes are added, unless Cloud Logging or Cloud Monitoring are enabled, in which case their required scopes will be added.

serviceAccount

string

The Google Cloud Platform Service Account to be used by the node VMs. Specify the email address of the Service Account; otherwise, if no Service Account is specified, the "default" service account is used.

metadata

map (key: string, value: string)

The metadata key/value pairs assigned to instances in the cluster.

Keys must conform to the regexp [a-zA-Z0-9-_]+ and be less than 128 bytes in length. These are reflected as part of a URL in the metadata server. Additionally, to avoid ambiguity, keys must not conflict with any other metadata keys for the project or be one of the reserved keys:

  • "cluster-location"
  • "cluster-name"
  • "cluster-uid"
  • "configure-sh"
  • "containerd-configure-sh"
  • "enable-os-login"
  • "gci-ensure-gke-docker"
  • "gci-metrics-enabled"
  • "gci-update-strategy"
  • "instance-template"
  • "kube-env"
  • "startup-script"
  • "user-data"
  • "disable-address-manager"
  • "windows-startup-script-ps1"
  • "common-psm1"
  • "k8s-node-setup-psm1"
  • "install-ssh-psm1"
  • "user-profile-psm1"

Values are free-form strings, and only have meaning as interpreted by the image running in the instance. The only restriction placed on them is that each value's size must be less than or equal to 32 KB.

The total size of all keys and values must be less than 512 KB.

An object containing a list of "key": value pairs. Example: { "name": "wrench", "mass": "1.3kg", "count": "3" }.

imageType

string

The image type to use for this node. Note that for a given image type, the latest version of it will be used. Please see https://cloud.google.com/kubernetes-engine/docs/concepts/node-images for available image types.

labels

map (key: string, value: string)

The map of Kubernetes labels (key/value pairs) to be applied to each node. These will added in addition to any default label(s) that Kubernetes may apply to the node. In case of conflict in label keys, the applied set may differ depending on the Kubernetes version -- it's best to assume the behavior is undefined and conflicts should be avoided. For more information, including usage and the valid values, see: https://kubernetes.io/docs/concepts/overview/working-with-objects/labels/

An object containing a list of "key": value pairs. Example: { "name": "wrench", "mass": "1.3kg", "count": "3" }.

localSsdCount

integer

The number of local SSD disks to be attached to the node.

The limit for this value is dependent upon the maximum number of disks available on a machine per zone. See: https://cloud.google.com/compute/docs/disks/local-ssd for more information.

tags[]

string

The list of instance tags applied to all nodes. Tags are used to identify valid sources or targets for network firewalls and are specified by the client during cluster or node pool creation. Each tag within the list must comply with RFC1035.

preemptible

boolean

Whether the nodes are created as preemptible VM instances. See: https://cloud.google.com/compute/docs/instances/preemptible for more information about preemptible VM instances.

accelerators[]

object (AcceleratorConfig)

A list of hardware accelerators to be attached to each node. See https://cloud.google.com/compute/docs/gpus for more information about support for GPUs.

diskType

string

Type of the disk attached to each node (e.g. 'pd-standard', 'pd-ssd' or 'pd-balanced')

If unspecified, the default disk type is 'pd-standard'

minCpuPlatform

string

Minimum CPU platform to be used by this instance. The instance may be scheduled on the specified or newer CPU platform. Applicable values are the friendly names of CPU platforms, such as minCpuPlatform: "Intel Haswell" or minCpuPlatform: "Intel Sandy Bridge". For more information, read how to specify min CPU platform

workloadMetadataConfig

object (WorkloadMetadataConfig)

The workload metadata configuration for this node.

taints[]

object (NodeTaint)

List of kubernetes taints to be applied to each node.

For more information, including usage and the valid values, see: https://kubernetes.io/docs/concepts/configuration/taint-and-toleration/

sandboxConfig

object (SandboxConfig)

Sandbox configuration for this node.

nodeGroup

string

Setting this field will assign instances of this pool to run on the specified node group. This is useful for running workloads on sole tenant nodes.

reservationAffinity

object (ReservationAffinity)

The optional reservation affinity. Setting this field will apply the specified Zonal Compute Reservation to this node pool.

shieldedInstanceConfig

object (ShieldedInstanceConfig)

Shielded Instance options.

linuxNodeConfig

object (LinuxNodeConfig)

Parameters that can be configured on Linux nodes.

kubeletConfig

object (NodeKubeletConfig)

Node kubelet configs.

bootDiskKmsKey

string

The Customer Managed Encryption Key used to encrypt the boot disk attached to each node in the node pool. This should be of the form projects/[KEY_PROJECT_ID]/locations/[LOCATION]/keyRings/[RING_NAME]/cryptoKeys/[KEY_NAME]. For more information about protecting resources with Cloud KMS Keys please see: https://cloud.google.com/compute/docs/disks/customer-managed-encryption

gcfsConfig

object (GcfsConfig)

Google Container File System (image streaming) configs.

advancedMachineFeatures

object (AdvancedMachineFeatures)

Advanced features for the Compute Engine VM.

gvnic

object (VirtualNIC)

Enable or disable gvnic in the node pool.

spot

boolean

Spot flag for enabling Spot VM, which is a rebrand of the existing preemptible flag.

confidentialNodes

object (ConfidentialNodes)

Confidential nodes config. All the nodes in the node pool will be Confidential VM once enabled.

resourceLabels

map (key: string, value: string)

The resource labels for the node pool to use to annotate any related Google Compute Engine resources.

An object containing a list of "key": value pairs. Example: { "name": "wrench", "mass": "1.3kg", "count": "3" }.

loggingConfig

object (NodePoolLoggingConfig)

Logging configuration.

windowsNodeConfig

object (WindowsNodeConfig)

Parameters that can be configured on Windows nodes.

localNvmeSsdBlockConfig

object (LocalNvmeSsdBlockConfig)

Parameters for using raw-block Local NVMe SSDs.

ephemeralStorageLocalSsdConfig

object (EphemeralStorageLocalSsdConfig)

Parameters for the node ephemeral storage using Local SSDs. If unspecified, ephemeral storage is backed by the boot disk.

soleTenantConfig

object (SoleTenantConfig)

Parameters for node pools to be backed by shared sole tenant node groups.

resourceManagerTags

object (ResourceManagerTags)

A map of resource manager tag keys and values to be attached to the nodes.

enableConfidentialStorage

boolean

Optional. Reserved for future use.

secondaryBootDisks[]

object (SecondaryBootDisk)

List of secondary boot disks attached to the nodes.

fastSocket

object (FastSocket)

Enable or disable NCCL fast socket for the node pool.

AcceleratorConfig

AcceleratorConfig represents a Hardware Accelerator request.

JSON representation
{
  "acceleratorCount": string,
  "acceleratorType": string,
  "gpuPartitionSize": string,
  "gpuSharingConfig": {
    object (GPUSharingConfig)
  },
  "gpuDriverInstallationConfig": {
    object (GPUDriverInstallationConfig)
  }
}
Fields
acceleratorCount

string (int64 format)

The number of the accelerator cards exposed to an instance.

acceleratorType

string

The accelerator type resource name. List of supported accelerators here

gpuPartitionSize

string

Size of partitions to create on the GPU. Valid values are described in the NVIDIA mig user guide.

gpuSharingConfig

object (GPUSharingConfig)

The configuration for GPU sharing options.

gpuDriverInstallationConfig

object (GPUDriverInstallationConfig)

The configuration for auto installation of GPU driver.

GPUSharingConfig

GPUSharingConfig represents the GPU sharing configuration for Hardware Accelerators.

JSON representation
{
  "maxSharedClientsPerGpu": string,
  "gpuSharingStrategy": enum (GPUSharingStrategy)
}
Fields
maxSharedClientsPerGpu

string (int64 format)

The max number of containers that can share a physical GPU.

gpuSharingStrategy

enum (GPUSharingStrategy)

The type of GPU sharing strategy to enable on the GPU node.

GPUSharingStrategy

The type of GPU sharing strategy currently provided.

Enums
GPU_SHARING_STRATEGY_UNSPECIFIED Default value.
TIME_SHARING GPUs are time-shared between containers.

GPUDriverInstallationConfig

GPUDriverInstallationConfig specifies the version of GPU driver to be auto installed.

JSON representation
{
  "gpuDriverVersion": enum (GPUDriverVersion)
}
Fields
gpuDriverVersion

enum (GPUDriverVersion)

Mode for how the GPU driver is installed.

GPUDriverVersion

The GPU driver version to install.

Enums
GPU_DRIVER_VERSION_UNSPECIFIED Default value is to not install any GPU driver.
INSTALLATION_DISABLED Disable GPU driver auto installation and needs manual installation
DEFAULT "Default" GPU driver in COS and Ubuntu.
LATEST "Latest" GPU driver in COS.

WorkloadMetadataConfig

WorkloadMetadataConfig defines the metadata configuration to expose to workloads on the node pool.

JSON representation
{
  "mode": enum (Mode)
}
Fields
mode

enum (Mode)

Mode is the configuration for how to expose metadata to workloads running on the node pool.

Mode

Mode is the configuration for how to expose metadata to workloads running on the node.

Enums
MODE_UNSPECIFIED Not set.
GCE_METADATA Expose all Compute Engine metadata to pods.
GKE_METADATA Run the GKE Metadata Server on this node. The GKE Metadata Server exposes a metadata API to workloads that is compatible with the V1 Compute Metadata APIs exposed by the Compute Engine and App Engine Metadata Servers. This feature can only be enabled if Workload Identity is enabled at the cluster level.

NodeTaint

Kubernetes taint is composed of three fields: key, value, and effect. Effect can only be one of three types: NoSchedule, PreferNoSchedule or NoExecute.

See here for more information, including usage and the valid values.

JSON representation
{
  "key": string,
  "value": string,
  "effect": enum (Effect)
}
Fields
key

string

Key for taint.

value

string

Value for taint.

effect

enum (Effect)

Effect for taint.

Effect

Possible values for Effect in taint.

Enums
EFFECT_UNSPECIFIED Not set
NO_SCHEDULE NoSchedule
PREFER_NO_SCHEDULE PreferNoSchedule
NO_EXECUTE NoExecute

SandboxConfig

SandboxConfig contains configurations of the sandbox to use for the node.

JSON representation
{
  "type": enum (Type)
}
Fields
type

enum (Type)

Type of the sandbox to use for the node.

Type

Possible types of sandboxes.

Enums
UNSPECIFIED Default value. This should not be used.
GVISOR Run sandbox using gvisor.

ReservationAffinity

ReservationAffinity is the configuration of desired reservation which instances could take capacity from.

JSON representation
{
  "consumeReservationType": enum (Type),
  "key": string,
  "values": [
    string
  ]
}
Fields
consumeReservationType

enum (Type)

Corresponds to the type of reservation consumption.

key

string

Corresponds to the label key of a reservation resource. To target a SPECIFIC_RESERVATION by name, specify "compute.googleapis.com/reservation-name" as the key and specify the name of your reservation as its value.

values[]

string

Corresponds to the label value(s) of reservation resource(s).

Type

Indicates whether to consume capacity from a reservation or not.

Enums
UNSPECIFIED Default value. This should not be used.
NO_RESERVATION Do not consume from any reserved capacity.
ANY_RESERVATION Consume any reservation available.
SPECIFIC_RESERVATION Must consume from a specific reservation. Must specify key value fields for specifying the reservations.

LinuxNodeConfig

Parameters that can be configured on Linux nodes.

JSON representation
{
  "sysctls": {
    string: string,
    ...
  },
  "cgroupMode": enum (CgroupMode)
}
Fields
sysctls

map (key: string, value: string)

The Linux kernel parameters to be applied to the nodes and all pods running on the nodes.

The following parameters are supported.

net.core.busy_poll net.core.busy_read net.core.netdev_max_backlog net.core.rmem_max net.core.wmem_default net.core.wmem_max net.core.optmem_max net.core.somaxconn net.ipv4.tcp_rmem net.ipv4.tcp_wmem net.ipv4.tcp_tw_reuse

An object containing a list of "key": value pairs. Example: { "name": "wrench", "mass": "1.3kg", "count": "3" }.

cgroupMode

enum (CgroupMode)

cgroupMode specifies the cgroup mode to be used on the node.

CgroupMode

Possible cgroup modes that can be used.

Enums
CGROUP_MODE_UNSPECIFIED CGROUP_MODE_UNSPECIFIED is when unspecified cgroup configuration is used. The default for the GKE node OS image will be used.
CGROUP_MODE_V1 CGROUP_MODE_V1 specifies to use cgroupv1 for the cgroup configuration on the node image.
CGROUP_MODE_V2 CGROUP_MODE_V2 specifies to use cgroupv2 for the cgroup configuration on the node image.

NodeKubeletConfig

Node kubelet configs.

JSON representation
{
  "cpuManagerPolicy": string,
  "cpuCfsQuota": boolean,
  "cpuCfsQuotaPeriod": string,
  "podPidsLimit": string,
  "insecureKubeletReadonlyPortEnabled": boolean
}
Fields
cpuManagerPolicy

string

Control the CPU management policy on the node. See https://kubernetes.io/docs/tasks/administer-cluster/cpu-management-policies/

The following values are allowed. * "none": the default, which represents the existing scheduling behavior. * "static": allows pods with certain resource characteristics to be granted increased CPU affinity and exclusivity on the node. The default value is 'none' if unspecified.

cpuCfsQuota

boolean

Enable CPU CFS quota enforcement for containers that specify CPU limits.

This option is enabled by default which makes kubelet use CFS quota (https://www.kernel.org/doc/Documentation/scheduler/sched-bwc.txt) to enforce container CPU limits. Otherwise, CPU limits will not be enforced at all.

Disable this option to mitigate CPU throttling problems while still having your pods to be in Guaranteed QoS class by specifying the CPU limits.

The default value is 'true' if unspecified.

cpuCfsQuotaPeriod

string

Set the CPU CFS quota period value 'cpu.cfs_period_us'.

The string must be a sequence of decimal numbers, each with optional fraction and a unit suffix, such as "300ms". Valid time units are "ns", "us" (or "µs"), "ms", "s", "m", "h". The value must be a positive duration.

podPidsLimit

string (int64 format)

Set the Pod PID limits. See https://kubernetes.io/docs/concepts/policy/pid-limiting/#pod-pid-limits

Controls the maximum number of processes allowed to run in a pod. The value must be greater than or equal to 1024 and less than 4194304.

insecureKubeletReadonlyPortEnabled

boolean

Enable or disable Kubelet read only port.

AdvancedMachineFeatures

Specifies options for controlling advanced machine features.

JSON representation
{
  "threadsPerCore": string
}
Fields
threadsPerCore

string (int64 format)

The number of threads per physical core. To disable simultaneous multithreading (SMT) set this to 1. If unset, the maximum number of threads supported per core by the underlying processor is assumed.

VirtualNIC

Configuration of gVNIC feature.

JSON representation
{
  "enabled": boolean
}
Fields
enabled

boolean

Whether gVNIC features are enabled in the node pool.

FastSocket

Configuration of Fast Socket feature.

JSON representation
{
  "enabled": boolean
}
Fields
enabled

boolean

Whether Fast Socket features are enabled in the node pool.

WindowsNodeConfig

Parameters that can be configured on Windows nodes. Windows Node Config that define the parameters that will be used to configure the Windows node pool settings

JSON representation
{
  "osVersion": enum (OSVersion)
}
Fields
osVersion

enum (OSVersion)

OSVersion specifies the Windows node config to be used on the node

OSVersion

Possible OS version that can be used.

Enums
OS_VERSION_UNSPECIFIED When OSVersion is not specified
OS_VERSION_LTSC2019 LTSC2019 specifies to use LTSC2019 as the Windows Servercore Base Image
OS_VERSION_LTSC2022 LTSC2022 specifies to use LTSC2022 as the Windows Servercore Base Image

LocalNvmeSsdBlockConfig

LocalNvmeSsdBlockConfig contains configuration for using raw-block local NVMe SSDs

JSON representation
{
  "localSsdCount": integer
}
Fields
localSsdCount

integer

Number of local NVMe SSDs to use. The limit for this value is dependent upon the maximum number of disk available on a machine per zone. See: https://cloud.google.com/compute/docs/disks/local-ssd for more information.

A zero (or unset) value has different meanings depending on machine type being used: 1. For pre-Gen3 machines, which support flexible numbers of local ssds, zero (or unset) means to disable using local SSDs as ephemeral storage. 2. For Gen3 machines which dictate a specific number of local ssds, zero (or unset) means to use the default number of local ssds that goes with that machine type. For example, for a c3-standard-8-lssd machine, 2 local ssds would be provisioned. For c3-standard-8 (which doesn't support local ssds), 0 will be provisioned. See https://cloud.google.com/compute/docs/disks/local-ssd#choose_number_local_ssds for more info.

EphemeralStorageLocalSsdConfig

EphemeralStorageLocalSsdConfig contains configuration for the node ephemeral storage using Local SSDs.

JSON representation
{
  "localSsdCount": integer
}
Fields
localSsdCount

integer

Number of local SSDs to use to back ephemeral storage. Uses NVMe interfaces.

A zero (or unset) value has different meanings depending on machine type being used: 1. For pre-Gen3 machines, which support flexible numbers of local ssds, zero (or unset) means to disable using local SSDs as ephemeral storage. The limit for this value is dependent upon the maximum number of disk available on a machine per zone. See: https://cloud.google.com/compute/docs/disks/local-ssd for more information. 2. For Gen3 machines which dictate a specific number of local ssds, zero (or unset) means to use the default number of local ssds that goes with that machine type. For example, for a c3-standard-8-lssd machine, 2 local ssds would be provisioned. For c3-standard-8 (which doesn't support local ssds), 0 will be provisioned. See https://cloud.google.com/compute/docs/disks/local-ssd#choose_number_local_ssds for more info.

SoleTenantConfig

SoleTenantConfig contains the NodeAffinities to specify what shared sole tenant node groups should back the node pool.

JSON representation
{
  "nodeAffinities": [
    {
      object (NodeAffinity)
    }
  ]
}
Fields
nodeAffinities[]

object (NodeAffinity)

NodeAffinities used to match to a shared sole tenant node group.

NodeAffinity

Specifies the NodeAffinity key, values, and affinity operator according to shared sole tenant node group affinities.

JSON representation
{
  "key": string,
  "operator": enum (Operator),
  "values": [
    string
  ]
}
Fields
key

string

Key for NodeAffinity.

operator

enum (Operator)

Operator for NodeAffinity.

values[]

string

Values for NodeAffinity.

Operator

Operator allows user to specify affinity or anti-affinity for the given key values.

Enums
OPERATOR_UNSPECIFIED Invalid or unspecified affinity operator.
IN Affinity operator.
NOT_IN Anti-affinity operator.

SecondaryBootDisk

SecondaryBootDisk represents a persistent disk attached to a node with special configurations based on its mode.

JSON representation
{
  "mode": enum (Mode),
  "diskImage": string
}
Fields
mode

enum (Mode)

Disk mode (container image cache, etc.)

diskImage

string

Fully-qualified resource ID for an existing disk image.

Mode

Mode specifies how the secondary boot disk will be used. This triggers mode-specified logic in the control plane.

Enums
MODE_UNSPECIFIED MODE_UNSPECIFIED is when mode is not set.
CONTAINER_IMAGE_CACHE CONTAINER_IMAGE_CACHE is for using the secondary boot disk as a container image cache.