Console

You must create a sole-tenant node template before you create a node group.By using the Google Cloud console, you must create the node template during thecreation of a sole-tenant node group. The new node template is created inthe same region that you specify in the node group properties.

1. In the Google Cloud console, go to theSole-tenant nodes page.

   Go to Sole-tenant nodes

2. ClickCreate node group.

3. Specify aName for the node group.

4. Specify aRegion to create the node template in. You can use the nodetemplate to create node groups in any zone of this region.

5. Specify theZone and clickContinue.

6. In theNode template list, clickCreate node template to begincreating a sole-tenant node template.

7. Specify aName for the node template.

8. Specify theNode type for each sole-tenant node in the node group tocreate based on this node template.

9. Optional: you can also specify the following properties for the nodetemplate.

   • Add alocal SSD andGPU accelerator.

   • SelectEnable CPU overcommit to controlCPU overcommitlevels foreach VM scheduled on the node.

   • AddNode affinity labels. Affinity labels let you logically groupnodes and node groups, and later, when provisioning VMs, you can specifyaffinity labels on the VMs to schedule VMs on a specific set of nodes ornode groups. For more information, seeNode affinity andanti-affinity.

10. ClickCreate to finish creating your node template.

11. Optional: to add a new sole-tenant node template in a different region,repeat the preceding steps.

To view the node templates, clickNode templates in theSole-tenant nodes page.

gcloud

Use thegcloud compute sole-tenancy node-templatescreate commandto create a node template:

gcloud compute sole-tenancy node-templates createTEMPLATE_NAME \  --node-type=NODE_TYPE \  [--region=REGION \]  [--node-affinity-labels=AFFINITY_LABELS \]  [--accelerator type=GPU_TYPE,count=GPU_COUNT \]  [--disk type=local-ssd,count=DISK_COUNT,size=DISK_SIZE \]  [--cpu-overcommit-type=CPU_OVERCOMMIT_TYPE]

Replace the following:

• TEMPLATE_NAME: the name for the new node template.

• NODE_TYPE: the node type for sole-tenant nodescreated based on this template. Use thegcloud compute sole-tenancynode-types list commandto get a list of the node types available in each zone.

• REGION: the region to create the node template in.You can use this template to create node groups in any zone of thisregion.

• AFFINITY_LABELS: the keys and values,[KEY=VALUE,...], for affinity labels. Affinity labels letyou logically group nodes and node groups and later, when provisioningVMs, you can specify affinity labels on the VMs to schedule VMs on aspecific set of nodes or node groups. For more information, seeNodeaffinity and anti-affinity.

• GPU_TYPE: the type of GPU for each sole-tenantnode created based on this node template. For information on the zonalavailability of the required sole-tenant node type, use thegcloud compute sole-tenancy node-types list command. For example, to see zones fora2-highgpunode types, add the flag--filter="name~'a2-highgpu'". For theGPU and node types available, see the table in theGPU_COUNTdescription.

• GPU_COUNT: the number of GPUs to attach to eachsole-tenant node. The value forGPU_COUNT depends ontheGPU_TYPE and the sole-tenant node type. SetGPU_COUNT to the value shown in the following table:

  Node type	GPU_TYPE	GPU_COUNT
  a2-highgpu	nvidia-a100-40gb	8
  a2-megagpu	nvidia-a100-40gb	16
  a2-ultragpu	nvidia-a100-80gb	8
  a3-highgpu	nvidia-h100-80gb	8
  a3-megagpu	nvidia-h100-mega-80gb	8
  g2	nvidia-l4	8
  n1	nvidia-tesla-p100	4
  n1	nvidia-tesla-p4	4
  n1	nvidia-tesla-t4	4
  n1	nvidia-tesla-v100	8

• DISK_COUNT: number of Local SSD disks. Set to16or24. This parameter is not required for A2 Ultra, A3 High, and A3 Meganode types because they include a fixed number of Local SSD disks.

• DISK_SIZE: optional value for the partition sizeof the local SSD in GB. The only supported partition size is375, and ifyou do not set this value it defaults to375.

• CPU_OVERCOMMIT_TYPE: the overcommit type for CPUson a VM. Set toenabled ornone.

REST

Use thenodeTemplates.insert methodto create a node template:

POST https://compute.googleapis.com/compute/v1/projects/PROJECT_ID/regions/REGION/nodeTemplates{  "name": "TEMPLATE_NAME",  "nodeType": "NODE_TYPE",  "nodeAffinityLabels": {    "KEY": "VALUE",    ...  },  "accelerators": [    {      "acceleratorType": "GPU_TYPE",      "acceleratorCount":GPU_COUNT    }  ],  "disks": [    {      "diskType": "local-ssd",      "diskSizeGb":DISK_SIZE,      "diskCount":DISK_COUNT    }  ],  "cpuOvercommitType":CPU_OVERCOMMIT_TYPE}

Replace the following:

• PROJECT_ID: the project ID.

• REGION: the region to create the node template in.You can use this template to create node groups in any zone of thisregion.

• TEMPLATE_NAME: the name for the new node template.

• NODE_TYPE: the node type for sole-tenant nodescreated based on this template. Use thenodeTypes.list methodto get a list of the node types available in each zone.

• KEY: thenodeAffinityLabels value that specifiesthe key portion of a node affinity label expressed as a key-value pair.Affinity labels let you logically group nodes and node groups, and later,when provisioning VMs, you can specify affinity labels on the VMs toschedule VMs on a specific set of nodes or node groups. For moreinformation, seeNode affinity and anti-affinity.

• VALUE: thenodeAffinityLabels value thatspecifies the value portion of a node affinity label key-value pair.

• GPU_TYPE: the type of GPU for each sole-tenantnode created based on this node template. For information on the zonalavailability of the required sole-tenant node type, use thenodeTypes.listmethod. For example, to seezones fora2-highgpu node types, use the filtername~"a2-highgpu.*". Forthe GPU and node types available, see the table in theGPU_COUNTdescription.

• GPU_COUNT: the number of GPUs for each sole-tenantnode created based on this node template. The value forGPU_COUNT depends on theGPU_TYPE and the sole-tenant node type. SetGPU_COUNT to the value shown in the following table:

  Node type	GPU_TYPE	GPU_COUNT
  a2-highgpu	nvidia-a100-40gb	8
  a2-megagpu	nvidia-a100-40gb	16
  a2-ultragpu	nvidia-a100-80gb	8
  a3-highgpu	nvidia-h100-80gb	8
  a3-megagpu	nvidia-h100-mega-80gb	8
  g2	nvidia-l4	8
  n1	nvidia-tesla-p100	4
  n1	nvidia-tesla-p4	4
  n1	nvidia-tesla-t4	4
  n1	nvidia-tesla-v100	8

• DISK_SIZE: optional value for the partition sizeof the local SSD in GB. The only supported partition size is375, and ifyou don't set this value it defaults to375.

• DISK_COUNT: number of Local SSD disks. Set to16 or24. This parameter is not required for A3 High and A3 Mega node typesbecause they include a fixed number of Local SSD disks.

• CPU_OVERCOMMIT_TYPE: CPU overcommit type. Set toenabled,none, orCPU_OVERCOMMIT_TYPE_UNSPECIFIED.

Console

1. In the Google Cloud console, go to theSole-tenant nodes page.

   Go to Sole-tenant nodes

2. ClickCreate node group to begin creating a node group.

3. Specify aName for the node group.

4. Specify theRegion for the node group to display the available nodetemplates in that region.

5. Specify theZone within the region to create the node group in.

6. Specify theNode template to create the node group or clickCreate node template tocreate a new sole-tenant node template. The selected node template is applied to the node group.

7. Choose one of the following for theAutoscaling mode for thenodegroup autoscaler:

   • Off: Manually manage the size of the nodegroup.

   • On: Have nodes automatically added to or removed from thenode group.

   • Only scale out: Add nodes to the node group when extracapacity is required.

8. Specify theNumber of nodes for the group. If you enable thenodegroup autoscaler, specify arange for the size of the node group. You can manually change the valueslater.

9. Set theMaintenance policy of the sole-tenant node group in theConfigure Maintenance Settings section to one of the following values.The maintenance policy lets you configure the behavior of VMs on the nodegroup during host maintenance events. For more information, seeMaintenance policies.

   • Default
   • Restart in place
   • Migrate within node group

10. You can choose betweenregular maintenance windows andadvanced maintenance control to handle maintenance for your sole-tenantnode group, as follows:

    • Maintenance Window: Select the time period during which you wantplanned maintenance events to happen for the nodes in the sole-tenantnode groups.

    • Opt-in for advanced maintenance control for sole-tenancy:Advanced maintenance control for sole-tenancy lets you controlplanned maintenance events for sole-tenant node groups and minimizemaintenance-related disruptions. To opt-in for advanced maintenancecontrol, click theOpt-in for advanced maintenance control for sole-tenancytoggle to theon position. If you choose to use this option for nodemaintenance, theMaintenance window field is disabled, andmaintenance occurs as configured in advanced maintenance control.

    Note that advanced maintenance control only supports theDefaultmaintenance policy.

11. Configure the share settings by specifying one of the following inConfigure share settings:

    • To share the node group with all projects in your organization, chooseShare this node group with all projects within the organization.
    • To share the node group with specific projects within yourorganization, chooseShare this node group with selected projectswithin the organization.

    If you do not want to share the node group, chooseDo not share thisnode group with other projects. For more information about sharing nodegroups, seeShare sole-tenant node groups.

12. ClickCreate to finish creating the node group.

gcloud

Run thegcloud compute sole-tenancy node-groupscreate commandcommand to create a node group based on a previously created nodetemplate:

gcloud compute sole-tenancy node-groups createGROUP_NAME \  --node-template=TEMPLATE_NAME \  --target-size=TARGET_SIZE \  [--zone=ZONE \]  [--maintenance-policy=MAINTENANCE_POLICY \]  [--maintenance-window-start-time=START_TIME \]  [--autoscaler-mode=AUTOSCALER_MODE: \  --min-nodes=MIN_NODES \  --max-nodes=MAX_NODES]

Replace the following:

• GROUP_NAME: the name for the new node group.

• TEMPLATE_NAME: the name of the node template touse to create this group.

• TARGET_SIZE: the number of nodes to create in thegroup.

• ZONE: the zone to create the node group in. This mustbe the same region as the node template on which you are basing the nodegroup.

• MAINTENANCE_POLICY: the maintenance policy for thenode group. For more information, seeMaintenancepolicies. This must be one of thefollowing values:

  • default
  • restart-in-place
  • migrate-within-node-group

  Alternately, you can opt in for advanced maintenance control for thesole-tenant node group, using the--maintenance-interval flag.For more information, seeEnable advanced maintenance control on a sole-tenant node.

• START_TIME: the start time in GMT for themaintenance window for the VMs in this node group. Set to one of:00:00,04:00,08:00,12:00,16:00, or20:00. If not set, the nodegroup does not have a set maintenance window.

• AUTOSCALER_MODE: the autoscaler policy for the nodegroup. This must be one of:

  • off: manually manage the size of the node group.

  • on: have nodes automatically added to or removed from thenode group.

  • only-scale-out: add nodes to the node group when extracapacity is required.

• MIN_NODES: the minimim size of the node group. Thedefault value is0 and must be an integer value less than or equal toMAX_NODES.

• MAX_NODES: the maximum size of the node group.This must be less than or equal to100 and greater than or equal toMIN_NODES. Required ifAUTOSCALER_MODE is not settooff.

REST

Use thenodeGroups.insert methodto create a node group based on a previously created nodetemplate:

POST https://compute.googleapis.com/compute/v1/projects/PROJECT_ID/zones/ZONE/nodeGroups?initialNodeCount=TARGET_SIZE
{  "nodeTemplate": "regions/REGION/nodeTemplates/TEMPLATE_NAME",  "name": "GROUP_NAME",  "maintenancePolicy":MAINTENANCE_POLICY,  "maintenanceWindow": {      "startTime": "START_TIME"  }  "autoscalingPolicy": {      "mode":AUTOSCALER_MODE,      "minNodes":MIN_NODES,      "maxNodes":MAX_NODES  },}

Replace the following:

• PROJECT_ID: the project ID.

• ZONE: the zone to create the node group in. Thismust be in the same region as the node template on which you are basing thenode group.

• TARGET_SIZE: the number of nodes to create in thegroup.

• REGION: the region to create the node group in.You must have a node template in the selected region.

• TEMPLATE_NAME: the name of the node template touse to create this group.

• GROUP_NAME: the name for the new node group.

• MAINTENANCE_POLICY: themaintenancepolicy for the node group. Thismust be one of the following values:

  • DEFAULT
  • RESTART_IN_PLACE
  • MIGRATE_WITHIN_NODE_GROUP

  Alternately, you can opt-in for advanced maintenance control for thesole-tenant node group, using themaintenanceInterval field.For more information, seeEnable advanced maintenance control on a sole-tenant node.

• START_TIME: the start time in GMT for themaintenance window for the VMs in this node group. Set to one of:00:00,04:00,08:00,12:00,16:00, or20:00. If not set, the nodegroup does not have a set maintenance window.

• AUTOSCALER_MODE: the autoscaler policy for the nodegroup. This must be one of the following values:

  • OFF: manually manage the size of the node group.

  • ON: have nodes automatically added to or removed from thenode group.

  • ONLY_SCALE_OUT: add nodes to the node group when extracapacity is required.

• MIN_NODES: the minimim size of the node group. The defaultis0 and must be an integer value less than or equal toMAX_NODES.

• MAX_NODES: the maximum size of the node group.This must be less than or equal to100 and greater than or equal toMIN_NODES. Required ifAUTOSCALER_MODE is not settoOFF.

Console

1. In the Google Cloud console, go to theSole-tenant nodes page.

   Go to Sole-tenant nodes

2. ClickNode groups.

3. Click theName of the node group to provision a VM instance on, andthen optionally, to provision a VM on a specific sole-tenant node, click thename of the specific sole-tenant node to provision the VM.

4. ClickCreate instance to provision a VM instance on this node group,note the values automatically applied for theName,Region, andZone, and modify those values as necessary.

5. Select aMachine configuration by specifying theMachine family,Series, andMachine type. Choose theSeries that corresponds tothesole-tenant node type.

6. Modify theBoot disk,Firewall, and other settings as necessary.

7. ClickSole Tenancy, note the automatically assignedNodeaffinity labels, and useBrowse to adjust as necessary.

8. ClickManagement, and forOn host maintenance, choose one of thefollowing:

   • Migrate VM instance (recommended): VM migrated to another node inthe node group during maintenance events.

   • Terminate: VM stopped during maintenance events.

9. Choose one of the following for theAutomatic restart:

   • On (recommended): Automatically restarts VMs if they arestopped for maintenance events.

   • Off: Does not automatically restart VMs after a maintenance event.

10. ClickCreate to finish creating your sole-tenant VM.

gcloud

Use thegcloud compute instancescreate command command toprovision a VM on a sole-tenant node group:

gcloud compute instances createVM_NAME \  [--zone=ZONE \]  --image-family=IMAGE_FAMILY \  --image-project=IMAGE_PROJECT \  --node-group=GROUP_NAME \  --machine-type=MACHINE_TYPE \  [--maintenance-policy=MAINTENANCE_POLICY \]  [--accelerator type=GPU_TYPE,count=GPU_COUNT \]  [--local-ssd interface=SSD_INTERFACE \]  [--restart-on-failure]

The--restart-on-failure flag indicates whether sole-tenant VMsrestart after stopping. This flag is enabled by default. Use--no-restart-on-failure to disable.

Replace the following:

• VM_NAME: the name of the new sole-tenant VM.

• ZONE: the zone to provision the sole-tenant VM in.

• IMAGE_FAMILY: theimagefamily of the image to use to create theVM.

• IMAGE_PROJECT: theimageproject of the image family.

• GROUP_NAME: the name of the node group toprovision the VM on.

• MACHINE_TYPE: the machine type of the sole-tenantVM. Use thegcloud compute machine-types listcommand to get a list ofavailable machine types for the project.

• MAINTENANCE_POLICY: specifies restart behavior ofsole-tenant VMs during maintenance events. Set to one of the following:

  • MIGRATE: VM migrated to another node in the node groupduring maintenance events.

  • TERMINATE: VM stopped during maintenance events.

• GPU_TYPE: type of GPU. Set to one of theaccelerator types specified when the node template was created.

• GPU_COUNT: number of GPUs of the total specifiedby the node template to attach to this VM. Default value is1.

• SSD_INTERFACE: type of local SSD interface. Youcan only set this for instances created from a node template with local SSDsupport. If you specify this while creating the instance, and the nodetemplate does not support local SSD, instance creation fails. Set tonvmeif the boot disk image drivers are optimized for NVMe, otherwise set toscsi. Specify this flag and a corresponding value once for each local SSDpartition.

REST

Use theinstances.insert methodto provision a VM on a sole-tenant node group:

POST https://compute.googleapis.com/compute/v1/projects/PROJECT_ID/zones/VM_ZONE/instances
{  "machineType": "zones/MACHINE_TYPE_ZONE/machineTypes/MACHINE_TYPE",  "name": "VM_NAME",  "scheduling": {    "onHostMaintenance":MAINTENANCE_POLICY,    "automaticRestart":RESTART_ON_FAILURE,    "nodeAffinities": [      {        "key": "compute.googleapis.com/node-group-name",        "operator": "IN",        "values": [          "GROUP_NAME"        ]      }    ]  },  "networkInterfaces": [    {      "network": "global/networks/NETWORK",      "subnetwork": "regions/REGION/subnetworks/SUBNETWORK"    }  ],  "guestAccelerators": [    {      "acceleratorType":GPU_TYPE,      "acceleratorCount":GPU_COUNT    }  ],  "disks": [    {      "boot": true,      "initializeParams": {        "sourceImage": "projects/IMAGE_PROJECT/global/images/family/IMAGE_FAMILY"      }    },    {      "type":"SCRATCH",      "initializeParams":{        "diskType":"zones/LOCAL_SSD_ZONE/diskTypes/local-ssd"      },      "autoDelete":true,      "interface":"SSD_INTERFACE"    }  ]}

Replace the following:

• PROJECT_ID: the project ID.

• VM_ZONE: the zone to provision the sole-tenant VMin.

• MACHINE_TYPE_ZONE: the zone of the machine type.

• MACHINE_TYPE: the machine type of the sole-tenantVM. Use themachineTypes.listmethod to get a list ofavailable machine types for the project.

• VM_NAME: the name of the new sole-tenant VM.

• MAINTENANCE_POLICY: specifies restart behavior ofsole-tenant VMs during maintenance events. Set to one of the following:

  • MIGRATE: VM migrated to another node in the node groupduring maintenance events.

  • TERMINATE: VM stopped during maintenance events.

• RESTART_ON_FAILURE: indicates whether sole-tenantVMs restart after stopping. Default istrue.

• GROUP_NAME: the name of the node group toprovision the VM on.

• NETWORK: the URL of the network resource for thisVM.

• REGION: the region containing the subnetwork forthis VM.

• SUBNETWORK: the URL of the subnetwork resource forthis VM.

• GPU_TYPE: the type of GPU. Set to one of theaccelerator types specified when the node template was created.

• GPU_COUNT: the number of GPUs of the totalspecified by the node template to attach to this VM. Default value is1.

• IMAGE_PROJECT:imageproject of the image family.

• IMAGE_FAMILY:imagefamily of the image to use to create theVM.

• LOCAL_SSD_ZONE: the local SSD's zone.

• SSD_INTERFACE: the type of local SSD interface.Set toNVME if the boot disk image drivers are optimized for NVMe,otherwise set toSCSI.

gcloud

1. Use thegcloud compute instance-templates createcommand to createa managed instance group template for a group of VMs to create on asole-tenant node group:

   gcloud compute instance-templates createINSTANCE_TEMPLATE \  --machine-type=MACHINE_TYPE \  --image-project=IMAGE_PROJECT \  --image-family=IMAGE_FAMILY \  --node-group=GROUP_NAME \  [--accelerator type=GPU_TYPE,count=GPU_COUNT \]  [--local-ssd interface=SSD_INTERFACE]

   Replace the following:

   • INSTANCE_TEMPLATE: the name for the newinstance template.

   • MACHINE_TYPE: the machine type of the sole-tenant VM. Use thegcloud compute machine-types listcommand to get a listof available machine types for the project.

   • IMAGE_PROJECT: theimageproject of the image family.

   • IMAGE_FAMILY: theimagefamily of the image to use to createthe VM.

   • GROUP_NAME: the name of the node group toprovision the VM on. Alternatively, if you want to use this instancetemplate to create a regional MIG that exists in more than one zone, usethe--node-affinity-file flagto specify a list of values for the regional MIG's node groups.

   • GPU_TYPE: type of GPU. Set to one of theaccelerator types specified when the node template was created.

   • GPU_COUNT: number of GPUs of the totalspecified by the node template to attach to this VM. Default value is1.

   • SSD_INTERFACE: type of local SSD interface.Set tonvme if the boot disk image drivers are optimized for NVMe,otherwise set toscsi. Specify this flag and a corresponding valueonce for each local SSD partition.

2. Use thegcloud compute instance-groups managedcreate commandcommand to create a managed instance group within your sole-tenant nodegroup:

   gcloud compute instance-groups managed createINSTANCE_GROUP_NAME \  --size=SIZE \  --template=INSTANCE_TEMPLATE \  --zone=ZONE

   Replace the following:

   • INSTANCE_GROUP_NAME: the name for this instancegroup.

   • SIZE: the number of VMs to include in thisinstance group. Your node group must have enough resources toaccommodate the instances in this managed instance group. Use themanaged instance group autoscaler toautomatically manage the size of managed instance groups.

   • INSTANCE_TEMPLATE: the name of the instancetemplate to use to create this MIG. The template must have one or morenode affinity labelspointing to the appropriate node groups.

   • ZONE: the zone to create the managed instancegroup in. For a regional MIG, replace the--zone flag with the--region flag and specify a region; also add the--zones flag tospecify all of the zones where the node groups exist.

REST

1. Use theinstanceTemplates.insert methodto create a managed instance group template within your sole-tenantnode group:

   POST https://compute.googleapis.com/compute/v1/projects/PROJECT_ID/zones/TEMPLATE_ZONE/instance-templates
   {  "name": "INSTANCE_TEMPLATE",  "properties": {    "machineType": "zones/MACHINE_TYPE_ZONE/machineTypes/MACHINE_TYPE",    "scheduling": {      "nodeAffinities": [        {          "key": "compute.googleapis.com/node-group-name",          "operator": "IN",          "values": [            "GROUP_NAME"          ]        }      ]    },    "networkInterfaces": [      {        "network": "global/networks/NETWORK",        "subnetwork": "regions/REGION/subnetworks/SUBNETWORK"      }    ],    "guestAccelerators": [      {        "acceleratorType":GPU_TYPE,        "acceleratorCount":GPU_COUNT      }    ],    "disks": [      {        "boot": true,        "initializeParams": {          "sourceImage": "projects/IMAGE_PROJECT/global/images/family/IMAGE_FAMILY"        }      },      {        "type":"SCRATCH",        "initializeParams":{          "diskType":"zones/LOCAL_SSD_ZONE/diskTypes/local-ssd"          },        "autoDelete":true,        "interface":"SSD_INTERFACE"      }    ]  }}

   Replace the following:

   • PROJECT_ID: the project ID.

   • TEMPLATE_ZONE: the zone to create the instancetemplate in.

   • INSTANCE_TEMPLATE: the name of the new instancetemplate.

   • MACHINE_TYPE_ZONE: the zone of the machine type.

   • MACHINE_TYPE: the machine type of the sole-tenantVM. Use themachineTypes.list methodto get a list of available machine types for the project.

   • GROUP_NAME: name of the node group toprovision the VM on. If you want to use this instance template to createa regional MIG that exists in more than one zone, specify a list of nodegroups that exist in the same zones as the regional MIG's zones.

   • NETWORK: the URL of the network resource for thisinstance template.

   • REGION: the region containing the subnetwork forthis instance template.

   • SUBNETWORK: the URL of the subnetwork resource forthis instance template.

   • GPU_TYPE: the type of GPU. Set to one of theaccelerator types specified when the node template was created.

   • GPU_COUNT: the number of GPUs of the total specifiedby the node template to attach to this VM. Default value is1.

   • IMAGE_PROJECT: theimageproject of the image family.

   • IMAGE_FAMILY: theimagefamily of the image to use to create theVM.

   • LOCAL_SSD_ZONE: the local SSD's zone.

   • SSD_INTERFACE: the type of local SSD interface.Set toNVME if the boot disk image drivers are optimized for NVMe,otherwise set toSCSI.

2. Use theinstanceGroupManagers.insert methodto create a MIG within your sole-tenant node group based on the previouslycreated instance template. Or, if you want to create a regional MIG, use theregionInstanceGroupManagers.insert methodand specify the region and zones of all of the node groups as specified inthe instance template.

   For example, to create a zonal MIG, use the following request:

   POST https://compute.googleapis.com/compute/v1/projects/PROJECT_ID/zones/ZONE/instanceGroupManagers{  "baseInstanceName": "NAME_PREFIX",  "name": "INSTANCE_GROUP_NAME",  "targetSize":SIZE,  "instanceTemplate": "global/instanceTemplates/INSTANCE_TEMPLATE"}

   Replace the following:

   • PROJECT_ID: the project ID.

   • ZONE: the zone to create the managed instancegroup in.

   • NAME_PREFIX: the prefix name for each of theinstances in the managed instance group.

   • INSTANCE_GROUP_NAME: the name for the instancegroup.

   • SIZE: the number of VMs to include in thisinstance group. Your node group must have enough resources toaccommodate the instances in this managed instance group. Use themanaged instance group autoscaler toautomatically manage the size of managed instance groups.

   • INSTANCE_TEMPLATE: the URL of the instancetemplate to use to create this group. The template must have anodeaffinity labelpointing to the appropriate node group.

gcloud

1. Use thegcloud compute sole-tenancy node-templatescreate commandto create a node template with a set of affinity labels for aproduction workload:

   gcloud compute sole-tenancy node-templates create prod-template \  --node-type=n1-node-96-624 \  --node-affinity-labels workload=frontend,environment=prod

2. Use thegcloud compute sole-tenancy node-templatesdescribe commandto view the node affinity labels assigned to the node template.

3. Use thegcloud compute sole-tenancy node-groupscreate commandto create a node group that uses the production template:

   gcloud compute sole-tenancy node-groups create prod-group \  --node-template=prod-template \  --target-size=1

4. For your production VMs, create anode-affinity-prod.json file tospecify the affinity of your production VMs. For example, you might create afile that specifies that VMs run only on nodes with both theworkload=frontend andenvironment=prod affinities. Create the nodeaffinity file by usingCloud Shellor create it in a location of your choice.

   [  {    "key" : "workload",    "operator" : "IN",    "values" : ["frontend"]  },  {    "key" : "environment",    "operator" : "IN",    "values" : ["prod"]  }]

5. Use thenode-affinity-prod.json file with thegcloud computeinstances create commandto schedule a VM on the node group with matching affinity labels.

   gcloud compute instances create prod-vm \  --node-affinity-file node-affinity-prod.json \  --machine-type=n1-standard-2

6. Use thegcloud compute instancesdescribe commandand check thescheduling field to view the node affinitiesassigned to the VM.

gcloud

1. For development VMs, specify the affinity of your development VMs bycreating anode-affinity-dev.json withCloud Shell, or by creating itin a location of your choice. For example, create a file that configures VMsto run on any node group with theworkload=frontend affinity as long as itis notenvironment=prod:

   [  {    "key" : "workload",    "operator" : "IN",    "values" : ["frontend"]  },  {    "key" : "environment",    "operator" : "NOT_IN",    "values" : ["prod"]  }]

2. Use thenode-affinity-dev.json file with thegcloud compute instancescreate command to createthe development VM:

   gcloud compute instances create dev-vm \  --node-affinity-file=node-affinity-dev.json \  --machine-type=n1-standard-2

3. Use thegcloud compute instancesdescribe commandand check thescheduling field to view the node anti-affinitiesassigned to the VM.