ProjectsInstances

Creates an instance and begins preparing it to begin serving. The returned long-running operation can be used to track the progress of preparing the new instance. The instance name is assigned by the caller. If the named instance already exists, CreateInstance returns ALREADY_EXISTS. Immediately upon completion of this request: * The instance is readable via the API, with all requested attributes but no allocated resources. Its state is CREATING.

Until completion of the returned operation: Cancelling the operation renders the instance immediately unreadable via the API. The instance can be deleted. All other attempts to modify the instance are rejected. Upon completion of the returned operation: Billing for all successfully- allocated resources begins (some types may have lower than the requested levels). Databases can be created in the instance. The instance's allocated resource levels are readable via the API. * The instance's state becomes READY. The returned long-running operation will have a name of the format /operations/ and can be used to track creation of the instance. The metadata field type is CreateInstanceMetadata. The response field type is Instance, if successful. (instances.create)

Deletes an instance. Immediately upon completion of the request: Billing ceases for all of the instance's reserved resources. Soon afterward: The instance and all of its databases immediately and irrevocably disappear from the API. All data in the databases is permanently deleted.

(instances.delete)

Gets information about a particular instance. (instances.get)

Gets the access control policy for an instance resource. Returns an empty policy if an instance exists but does not have a policy set. Authorization requires spanner.instances.getIamPolicy on resource.

(instances.getIamPolicy)

Lists all instances in the given project. (instances.listProjectsInstances)

Moves the instance to the target instance config. The returned long-running operation can be used to track the progress of moving the instance.

MoveInstance returns FAILED_PRECONDITION if the instance meets any of the following criteria: Has an ongoing move to a different instance config Has backups Has an ongoing update Is under free trial Contains any CMEK-enabled databases While the operation is pending: All other attempts to modify the instance, including changes to its compute capacity, are rejected. * The following database and backup admin operations are rejected:

• DatabaseAdmin.CreateDatabase, DatabaseAdmin.UpdateDatabaseDdl (Disabled if default_leader is specified in the request.) DatabaseAdmin.RestoreDatabase DatabaseAdmin.CreateBackup DatabaseAdmin.CopyBackup Both the source and target instance configurations are subject to hourly compute and storage charges. The instance may experience higher read-write latencies and a higher transaction abort rate. However, moving an instance does not cause any downtime. The returned long- running operation will have a name of the format /operations/ and can be used to track the move instance operation. The metadata field type is MoveInstanceMetadata. The response field type is Instance, if successful. Cancelling the operation sets its metadata's cancel_time. Cancellation is not immediate since it involves moving any data previously moved to target instance configuration back to the original instance config. The same operation can be used to track the progress of the cancellation. Upon successful completion of the cancellation, the operation terminates with CANCELLED status. Upon completion(if not cancelled) of the returned operation: Instance would be successfully moved to the target instance config. You are billed for compute and storage in target instance config. Authorization requires spanner.instances.update permission on the resource instance. For more details, please see documentation. (instances.move)

Updates an instance, and begins allocating or releasing resources as requested. The returned long-running operation can be used to track the progress of updating the instance. If the named instance does not exist, returns NOT_FOUND. Immediately upon completion of this request: For resource types for which a decrease in the instance's allocation has been requested, billing is based on the newly-requested level. Until completion of the returned operation: Cancelling the operation sets its metadata's cancel_time, and begins restoring resources to their pre-request values. The operation is guaranteed to succeed at undoing all resource changes, after which point it terminates with a CANCELLED status. All other attempts to modify the instance are rejected. Reading the instance via the API continues to give the pre-request resource levels. Upon completion of the returned operation: Billing begins for all successfully-allocated resources (some types may have lower than the requested levels). All newly-reserved resources are available for serving the instance's tables. * The instance's new resource levels are readable via the API. The returned long-running operation will have a name of the format /operations/ and can be used to track the instance modification. The metadata field type is UpdateInstanceMetadata. The response field type is Instance, if successful.

Authorization requires spanner.instances.update permission on the resource name. (instances.patch)

Sets the access control policy on an instance resource. Replaces any existing policy. Authorization requires spanner.instances.setIamPolicy on resource.

(instances.setIamPolicy)

Returns permissions that the caller has on the specified instance resource.

Attempting this RPC on a non-existent Cloud Spanner instance resource will result in a NOT_FOUND error if the user has spanner.instances.list permission on the containing Google Cloud Project. Otherwise returns an empty set of permissions. (instances.testIamPermissions)