Customize your development environment

This page describes the workstation configuration parameters that allow you to customize the Cloud Workstations environment, including the VM type, home disk size, base container image, and more. In the API, these workstation configuration parameters are collectively referred to as the WorkstationConfig.

Before you use these workstation configuration parameters, familiarize yourself with the following:


REST Resource: projects.locations.workstationClusters.workstationConfigs

Resource: WorkstationConfig

A workstation configuration resource in the Cloud Workstations API.

Workstation configurations act as templates for workstations. The workstation configuration defines details such as the workstation virtual machine (VM) instance type, persistent storage, container image defining environment, which IDE or Code Editor to use, and more. Administrators and platform teams can also use Identity and Access Management (IAM) rules to grant access to teams or to individual developers.

JSON representation
{
  "name": string,
  "displayName": string,
  "uid": string,
  "reconciling": boolean,
  "annotations": {
    string: string,
    ...
  },
  "labels": {
    string: string,
    ...
  },
  "createTime": string,
  "updateTime": string,
  "deleteTime": string,
  "etag": string,
  "idleTimeout": string,
  "runningTimeout": string,
  "host": {
    object (Host)
  },
  "persistentDirectories": [
    {
      object (PersistentDirectory)
    }
  ],
  "ephemeralDirectories": [
    {
      object (EphemeralDirectory)
    }
  ],
  "container": {
    object (Container)
  },
  "encryptionKey": {
    object (CustomerEncryptionKey)
  },
  "readinessChecks": [
    {
      object (ReadinessCheck)
    }
  ],
  "replicaZones": [
    string
  ],
  "degraded": boolean,
  "conditions": [
    {
      object (Status)
    }
  ],
  "enableAuditAgent": boolean,
  "disableTcpConnections": boolean
}
Fields
name

string

Identifier. Full name of this workstation configuration.

displayName

string

Optional. Human-readable name for this workstation configuration.

uid

string

Output only. A system-assigned unique identifier for this workstation configuration.

reconciling

boolean

Output only. Indicates whether this workstation configuration is currently being updated to match its intended state.

annotations

map (key: string, value: string)

Optional. Client-specified annotations.

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

labels

map (key: string, value: string)

Optional. Labels that are applied to the workstation configuration and that are also propagated to the underlying Compute Engine resources.

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

createTime

string (Timestamp format)

Output only. Time when this workstation configuration was created.

A timestamp in RFC3339 UTC "Zulu" format, with nanosecond resolution and up to nine fractional digits. Examples: "2014-10-02T15:01:23Z" and "2014-10-02T15:01:23.045123456Z".

updateTime

string (Timestamp format)

Output only. Time when this workstation configuration was most recently updated.

A timestamp in RFC3339 UTC "Zulu" format, with nanosecond resolution and up to nine fractional digits. Examples: "2014-10-02T15:01:23Z" and "2014-10-02T15:01:23.045123456Z".

deleteTime

string (Timestamp format)

Output only. Time when this workstation configuration was soft-deleted.

A timestamp in RFC3339 UTC "Zulu" format, with nanosecond resolution and up to nine fractional digits. Examples: "2014-10-02T15:01:23Z" and "2014-10-02T15:01:23.045123456Z".

etag

string

Optional. Checksum computed by the server. May be sent on update and delete requests to make sure that the client has an up-to-date value before proceeding.

idleTimeout

string (Duration format)

Optional. Number of seconds to wait before automatically stopping a workstation after it last received user traffic.

A value of "0s" indicates that Cloud Workstations VMs created with this configuration should never time out due to idleness. Provide duration terminated by s for seconds—for example, "7200s" (2 hours). The default is "1200s" (20 minutes).

A duration in seconds with up to nine fractional digits, ending with 's'. Example: "3.5s".

runningTimeout

string (Duration format)

Optional. Number of seconds that a workstation can run until it is automatically shut down. We recommend that workstations be shut down daily to reduce costs and so that security updates can be applied upon restart. The idleTimeout and runningTimeout fields are independent of each other. Note that the runningTimeout field shuts down VMs after the specified time, regardless of whether or not the VMs are idle.

Provide duration terminated by s for seconds—for example, "54000s" (15 hours). Defaults to "43200s" (12 hours). A value of "0s" indicates that workstations using this configuration should never time out. If encryptionKey is set, it must be greater than "0s" and less than "86400s" (24 hours).

Warning: A value of "0s" indicates that Cloud Workstations VMs created with this configuration have no maximum running time. This is strongly discouraged because you incur costs and will not pick up security updates.

A duration in seconds with up to nine fractional digits, ending with 's'. Example: "3.5s".

host

object (Host)

Optional. Runtime host for the workstation.

persistentDirectories[]

object (PersistentDirectory)

Optional. Directories to persist across workstation sessions.

ephemeralDirectories[]

object (EphemeralDirectory)

Optional. Ephemeral directories which won't persist across workstation sessions.

container

object (Container)

Optional. Container that runs upon startup for each workstation using this workstation configuration.

encryptionKey

object (CustomerEncryptionKey)

Immutable. Encrypts resources of this workstation configuration using a customer-managed encryption key (CMEK).

If specified, the boot disk of the Compute Engine instance and the persistent disk are encrypted using this encryption key. If this field is not set, the disks are encrypted using a generated key. Customer-managed encryption keys do not protect disk metadata.

If the customer-managed encryption key is rotated, when the workstation instance is stopped, the system attempts to recreate the persistent disk with the new version of the key. Be sure to keep older versions of the key until the persistent disk is recreated. Otherwise, data on the persistent disk might be lost.

If the encryption key is revoked, the workstation session automatically stops within 7 hours.

Immutable after the workstation configuration is created.

readinessChecks[]

object (ReadinessCheck)

Optional. Readiness checks to perform when starting a workstation using this workstation configuration. Mark a workstation as running only after all specified readiness checks return 200 status codes.

replicaZones[]

string

Optional. Immutable. Specifies the zones used to replicate the VM and disk resources within the region. If set, exactly two zones within the workstation cluster's region must be specified—for example, ['us-central1-a', 'us-central1-f']. If this field is empty, two default zones within the region are used.

Immutable after the workstation configuration is created.

degraded

boolean

Output only. Whether this resource is degraded, in which case it may require user action to restore full functionality. See also the conditions field.

conditions[]

object (Status)

Output only. Status conditions describing the current resource state.

enableAuditAgent

boolean

Optional. Whether to enable Linux auditd logging on the workstation. When enabled, a service account must also be specified that has logging.buckets.write permission on the project. Operating system audit logging is distinct from Cloud Audit Logs.

disableTcpConnections

boolean

Optional. Disables support for plain TCP connections in the workstation. By default the service supports TCP connections through a websocket relay. Setting this option to true disables that relay, which prevents the usage of services that require plain TCP connections, such as SSH. When enabled, all communication must occur over HTTPS or WSS.

Host

Runtime host for a workstation.

JSON representation
{

  // Union field config can be only one of the following:
  "gceInstance": {
    object (GceInstance)
  }
  // End of list of possible types for union field config.
}
Fields
Union field config. Type of host to be used for the workstation's runtime. config can be only one of the following:
gceInstance

object (GceInstance)

Specifies a Compute Engine instance as the host.

GceInstance

A runtime using a Compute Engine instance.

JSON representation
{
  "machineType": string,
  "serviceAccount": string,
  "serviceAccountScopes": [
    string
  ],
  "tags": [
    string
  ],
  "poolSize": integer,
  "pooledInstances": integer,
  "disablePublicIpAddresses": boolean,
  "enableNestedVirtualization": boolean,
  "shieldedInstanceConfig": {
    object (GceShieldedInstanceConfig)
  },
  "confidentialInstanceConfig": {
    object (GceConfidentialInstanceConfig)
  },
  "bootDiskSizeGb": integer,
  "accelerators": [
    {
      object (Accelerator)
    }
  ],
  "disableSsh": boolean
}
Fields
machineType

string

Optional. The type of machine to use for VM instances—for example, "e2-standard-4". For more information about machine types that Cloud Workstations supports, see the list of available machine types.

serviceAccount

string

Optional. The email address of the service account for Cloud Workstations VMs created with this configuration. When specified, be sure that the service account has logginglogEntries.create permission on the project so it can write logs out to Cloud Logging. If using a custom container image, the service account must have Artifact Registry Reader permission to pull the specified image.

If you as the administrator want to be able to ssh into the underlying VM, you need to set this value to a service account for which you have the iam.serviceAccounts.actAs permission. Conversely, if you don't want anyone to be able to ssh into the underlying VM, use a service account where no one has that permission.

If not set, VMs run with a service account provided by the Cloud Workstations service, and the image must be publicly accessible.

serviceAccountScopes[]

string

Optional. Scopes to grant to the serviceAccount. Various scopes are automatically added based on feature usage. When specified, users of workstations under this configuration must have iam.serviceAccounts.actAs on the service account.

tags[]

string

Optional. Network tags to add to the Compute Engine VMs backing the workstations. This option applies network tags to VMs created with this configuration. These network tags enable the creation of firewall rules.

poolSize

integer

Optional. The number of VMs that the system should keep idle so that new workstations can be started quickly for new users. Defaults to 0 in the API.

pooledInstances

integer

Output only. Number of instances currently available in the pool for faster workstation startup.

disablePublicIpAddresses

boolean

Optional. When set to true, disables public IP addresses for VMs. If you disable public IP addresses, you must set up Private Google Access or Cloud NAT on your network. If you use Private Google Access and you use private.googleapis.com or restricted.googleapis.com for Container Registry and Artifact Registry, make sure that you set up DNS records for domains *.gcr.io and *.pkg.dev. Defaults to false (VMs have public IP addresses).

enableNestedVirtualization

boolean

Optional. Whether to enable nested virtualization on Cloud Workstations VMs created using this workstation configuration.

Nested virtualization lets you run virtual machine (VM) instances inside your workstation. Before enabling nested virtualization, consider the following important considerations. Cloud Workstations instances are subject to the same restrictions as Compute Engine instances:

  • Organization policy: projects, folders, or organizations may be restricted from creating nested VMs if the Disable VM nested virtualization constraint is enforced in the organization policy. For more information, see the Compute Engine section, Checking whether nested virtualization is allowed.
  • Performance: nested VMs might experience a 10% or greater decrease in performance for workloads that are CPU-bound and possibly greater than a 10% decrease for workloads that are input/output bound.
  • Machine Type: nested virtualization can only be enabled on workstation configurations that specify a machineType in the N1 or N2 machine series.
  • GPUs: nested virtualization may not be enabled on workstation configurations with accelerators.
  • Operating System: Because Container-Optimized OS does not support nested virtualization, when nested virtualization is enabled, the underlying Compute Engine VM instances boot from an Ubuntu LTS image.
shieldedInstanceConfig

object (GceShieldedInstanceConfig)

Optional. A set of Compute Engine Shielded instance options.

confidentialInstanceConfig

object (GceConfidentialInstanceConfig)

Optional. A set of Compute Engine Confidential VM instance options.

bootDiskSizeGb

integer

Optional. The size of the boot disk for the VM in gigabytes (GB). The minimum boot disk size is 30 GB. Defaults to 50 GB.

accelerators[]

object (Accelerator)

Optional. A list of the type and count of accelerator cards attached to the instance.

disableSsh

boolean

Optional. Whether to disable SSH access to the VM.

GceShieldedInstanceConfig

A set of Compute Engine Shielded instance options.

JSON representation
{
  "enableSecureBoot": boolean,
  "enableVtpm": boolean,
  "enableIntegrityMonitoring": boolean
}
Fields
enableSecureBoot

boolean

Optional. Whether the instance has Secure Boot enabled.

enableVtpm

boolean

Optional. Whether the instance has the vTPM enabled.

enableIntegrityMonitoring

boolean

Optional. Whether the instance has integrity monitoring enabled.

GceConfidentialInstanceConfig

A set of Compute Engine Confidential VM instance options.

JSON representation
{
  "enableConfidentialCompute": boolean
}
Fields
enableConfidentialCompute

boolean

Optional. Whether the instance has confidential compute enabled.

Accelerator

An accelerator card attached to the instance.

JSON representation
{
  "type": string,
  "count": integer
}
Fields
type

string

Optional. Type of accelerator resource to attach to the instance, for example, "nvidia-tesla-p100".

count

integer

Optional. Number of accelerator cards exposed to the instance.

PersistentDirectory

A directory to persist across workstation sessions.

JSON representation
{
  "mountPath": string,

  // Union field directory_type can be only one of the following:
  "gcePd": {
    object (GceRegionalPersistentDisk)
  }
  // End of list of possible types for union field directory_type.
}
Fields
mountPath

string

Optional. Location of this directory in the running workstation.

Union field directory_type. How a persistent directory should be implemented. directory_type can be only one of the following:
gcePd

object (GceRegionalPersistentDisk)

A PersistentDirectory backed by a Compute Engine persistent disk.

GceRegionalPersistentDisk

A Persistent Directory backed by a Compute Engine regional persistent disk. The persistentDirectories field is repeated, but it may contain only one entry. It creates a persistent disk that mounts to the workstation VM at /home when the session starts and detaches when the session ends. If this field is empty, workstations created with this configuration do not have a persistent home directory.

JSON representation
{
  "sizeGb": integer,
  "fsType": string,
  "diskType": string,
  "sourceSnapshot": string,
  "reclaimPolicy": enum (ReclaimPolicy)
}
Fields
sizeGb

integer

Optional. The GB capacity of a persistent home directory for each workstation created with this configuration. Must be empty if sourceSnapshot is set.

Valid values are 10, 50, 100, 200, 500, or 1000. Defaults to 200. If less than 200 GB, the diskType must be "pd-balanced" or "pd-ssd".

fsType

string

Optional. Type of file system that the disk should be formatted with. The workstation image must support this file system type. Must be empty if sourceSnapshot is set. Defaults to "ext4".

diskType

string

Optional. The type of the persistent disk for the home directory. Defaults to "pd-standard".

sourceSnapshot

string

Optional. Name of the snapshot to use as the source for the disk. If set, sizeGb and fsType must be empty.

reclaimPolicy

enum (ReclaimPolicy)

Optional. Whether the persistent disk should be deleted when the workstation is deleted. Valid values are DELETE and RETAIN. Defaults to DELETE.

ReclaimPolicy

Value representing what should happen to the disk after the workstation is deleted.

Enums
RECLAIM_POLICY_UNSPECIFIED Do not use.
DELETE Delete the persistent disk when deleting the workstation.
RETAIN Keep the persistent disk when deleting the workstation. An administrator must manually delete the disk.

EphemeralDirectory

An ephemeral directory which won't persist across workstation sessions. It is freshly created on every workstation start operation.

JSON representation
{
  "mountPath": string,

  // Union field directory_type can be only one of the following:
  "gcePd": {
    object (GcePersistentDisk)
  }
  // End of list of possible types for union field directory_type.
}
Fields
mountPath

string

Required. Location of this directory in the running workstation.

Union field directory_type. How an ephemeral directory should be implemented. directory_type can be only one of the following:
gcePd

object (GcePersistentDisk)

An EphemeralDirectory backed by a Compute Engine persistent disk.

GcePersistentDisk

An EphemeralDirectory is backed by a Compute Engine persistent disk.

JSON representation
{
  "diskType": string,
  "sourceSnapshot": string,
  "sourceImage": string,
  "readOnly": boolean
}
Fields
diskType

string

Optional. Type of the disk to use. Defaults to "pd-standard".

sourceSnapshot

string

Optional. Name of the snapshot to use as the source for the disk. Must be empty if sourceImage is set. Must be empty if readOnly is false. Updating sourceSnapshot will update content in the ephemeral directory after the workstation is restarted. This field is mutable.

sourceImage

string

Optional. Name of the disk image to use as the source for the disk. Must be empty if sourceSnapshot is set. Updating sourceImage will update content in the ephemeral directory after the workstation is restarted. This field is mutable.

readOnly

boolean

Optional. Whether the disk is read only. If true, the disk may be shared by multiple VMs and sourceSnapshot must be set.

Container

A Docker container.

JSON representation
{
  "image": string,
  "command": [
    string
  ],
  "args": [
    string
  ],
  "env": {
    string: string,
    ...
  },
  "workingDir": string,
  "runAsUser": integer
}
Fields
image

string

Optional. A Docker container image that defines a custom environment.

Cloud Workstations provides a number of preconfigured images, but you can create your own custom container images. If using a private image, the host.gceInstance.serviceAccount field must be specified in the workstation configuration. If using a custom container image, the service account must have Artifact Registry Reader permission to pull the specified image. Otherwise, the image must be publicly accessible.

command[]

string

Optional. If set, overrides the default ENTRYPOINT specified by the image.

args[]

string

Optional. Arguments passed to the entrypoint.

env

map (key: string, value: string)

Optional. Environment variables passed to the container's entrypoint.

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

workingDir

string

Optional. If set, overrides the default DIR specified by the image.

runAsUser

integer

Optional. If set, overrides the USER specified in the image with the given uid.

CustomerEncryptionKey

A customer-managed encryption key (CMEK) for the Compute Engine resources of the associated workstation configuration. Specify the name of your Cloud KMS encryption key and the default service account. We recommend that you use a separate service account and follow Cloud KMS best practices.

JSON representation
{
  "kmsKey": string,
  "kmsKeyServiceAccount": string
}
Fields
kmsKey

string

Immutable. The name of the Google Cloud KMS encryption key. For example, "projects/PROJECT_ID/locations/REGION/keyRings/KEY_RING/cryptoKeys/KEY_NAME". The key must be in the same region as the workstation configuration.

kmsKeyServiceAccount

string

Immutable. The service account to use with the specified KMS key. We recommend that you use a separate service account and follow KMS best practices. For more information, see Separation of duties and gcloud kms keys add-iam-policy-binding --member.

ReadinessCheck

A readiness check to be performed on a workstation.

JSON representation
{
  "path": string,
  "port": integer
}
Fields
path

string

Optional. Path to which the request should be sent.

port

integer

Optional. Port to which the request should be sent.

Methods

create

Creates a new workstation configuration.

delete

Deletes the specified workstation configuration.

get

Returns the requested workstation configuration.

getIamPolicy

Gets the access control policy for a resource.

list

Returns all workstation configurations in the specified cluster.

listUsable

Returns all workstation configurations in the specified cluster on which the caller has the "workstations.workstation.create" permission.

patch

Updates an existing workstation configuration.

setIamPolicy

Sets the access control policy on the specified resource.

testIamPermissions

Returns permissions that a caller has on the specified resource.