Creating instances

This page describes how to create a Cloud SQL for a PostgreSQL instance.

For detailed information about all instance settings, see Instance Settings.

During instance creation, a single database is added to the instance. You can add additional databases by creating or importing them.

Before you begin

  1. Sign in to your Google Cloud account. If you're new to Google Cloud, create an account to evaluate how our products perform in real-world scenarios. New customers also get $300 in free credits to run, test, and deploy workloads.

  2. In the Google Cloud Console, on the project selector page, select or create a Google Cloud project.

    Go to project selector

  3. Make sure that billing is enabled for your Cloud project. Learn how to confirm that billing is enabled for your project.

  4. Install and initialize the Cloud SDK.
  5. Make sure you have the Cloud SQL Admin and Compute Viewer roles on your user account.

    Go to the IAM page

    Learn more about roles and permissions.

Creating a PostgreSQL instance

To create a PostgreSQL instance:

Console

  1. Go to the Cloud SQL Instances page in the Google Cloud Console.

    Go to the Cloud SQL Instances page

  2. Click Create instance.
  3. Select PostgreSQL and click Next.
  4. Enter a name.

    Do not include sensitive or personally identifiable information in your instance name; it is externally visible.

    You do not need to include the project ID in the instance name. This is done automatically where appropriate (for example, in the log files).

  5. Enter a password for the postgres user.
  6. Select the database version for your instance: PostgreSQL 13 (default), PostgreSQL 12, PostgreSQL 11, PostgreSQL 10, or PostgreSQL 9.6.

    The database version cannot be edited after the instance has been created.

  7. Under Choose region and zonal availability, select the region and zone for your instance.

    Place your instance in the same region as the resources that access it. The region you select can't be modified in the future. In most cases, you don't need to specify a zone.

    If you are configuring your instance for high availability, you can select both a primary and secondary zone.

    The following conditions apply when the secondary zone is used during instance creation:

    • The zones default to Any for the primary zone and Any (different from primary) for the secondary zone.
    • If both the primary and secondary zones are specified, they must be distinct zones.
  8. Under Customize your instance, update settings for your instance. Begin by clicking SHOW CONFIGURATION OPTIONS to display the groups of settings. Then, expand desired groups to review and customize settings. A Summary of all the options you select is shown on the right.

    The following table is a quick reference to instance settings. For more details about each setting, see the instance settings page.

    Setting Notes
    Machine type
    Machine typeSelect from Shared core, Lightweight, Standard (Most common), or High memory.
    Custom Select this button to create an instance with a flexible configuration. When you select this option, you need to select the number of cores and amount of memory for your instance. Learn more.
    Connectivity
    Cores The number of vCPUs for your instance. Learn more.
    Memory The amount of memory for your instance, in GBs. Learn more.
    Storage
    Storage type Determines whether your instance uses SSD or HDD storage. Learn more.
    Storage capacity The amount of storage provisioned for the instance. Learn more.
    Enable automatic storage increases Determines whether Cloud SQL automatically provides more storage for your instance when free space runs low. Learn more.
    Encryption
    Google-managed encryptionThe default option.
    Customer key-managed encryption key (CMEK)Select to use your key with Google Cloud Key Management Service. Learn more.
    Connections
    Private IP Adds a private IP address for your instance. To enable connecting to the instance, additional configuration is required. Learn more.
    Public IP Adds a public IP address for your instance. You can then add authorized networks to connect to the instance. Learn more.
    Authorized networksAdd the name for the new network and the Network address. Learn more.
    Backups
    Automate backups The window of time when you would like backups to start. Learn more.
    Choose where to store your backupsSelect Multi-region for most use cases. If you need to store backups in a specific region, for example, if there are regulatory reasons to do so, select Region and select your region from the Location drop-down menu.
    Choose how many automated backups to storeThe number of automated backups you would like to retain (from 1 to 365 days). Learn more.
    Enable point-in-time recovery Enables point-in-time recovery and write-ahead logging. Learn more.
    Choose how many days of logs to retain Configure write-ahead log retention from 1 to 7 days. The default setting is 7 days. Learn more.
    Maintenance
    Preferred window Determines a one-hour window when Cloud SQL can perform disruptive maintenance on your instance. If you do not set the window, then disruptive maintenance can be done at any time. Learn more.
    Order of updates Your preferred timing for instance updates, relative to other instances in the same project. Learn more.
    Flags
    ADD FLAG You can use database flags to control settings and parameters for your instance. Learn more.
    Labels
    ADD LABELAdd a key and value for each label that you add. You use labels to help organize your instances.
  9. Click Create.

gcloud

For information about installing and getting started with the gcloud command-line tool, see Installing Cloud SDK. For information about starting Cloud Shell, see the Cloud Shell documentation.

  1. Create the instance: 
    gcloud sql instances create INSTANCE_NAME --database-version=POSTGRES_12
 --cpu=NUMBER_CPUS --memory=MEMORY_SIZE
 --region=REGION --gce-zone=GCE_ZONE --zone=ZONE
    Or, alternatively, you can use the `--tier` flag if you choose db-f1-micro or db-g1-small as the machine type: 
    gcloud sql instances create INSTANCE_NAME --tier=API_TIER_STRING --region=REGION

    There are restrictions on the values for vCPUs and memory size:

    • vCPUs must be either 1 or an even number between 2 and 96.
    • Memory must be:
      • 0.9 to 6.5 GB per vCPU
      • A multiple of 256 MB
      • At least 3.75 GB (3840 MB)

    For example, the following string creates an instance with two vCPUs and 7,680 MB of memory: 

     gcloud sql instances create myinstance --database-version=POSTGRES_12 --cpu=2
 --memory=7680MB --region=us-central
    See Flexible instance configuration for more information about how to size --cpu and --memory.

    The default value for REGION is us-central.

    Do not include sensitive or personally identifiable information in your instance name; it is externally visible.
    You do not need to include the project ID in the instance name. This is done automatically where appropriate (for example, in the log files).

    If you are creating an instance for high availability, you can specify both the primary and secondary zones, using the --zone and --secondary-zone parameters. The following conditions apply when the secondary zone is used during instance creation or edit:

    • The zones must be valid zones.
    • If the secondary zone is specified, the primary must also be specified.
    • If the primary and secondary zones are specified, they must be distinct zones.
    • If the primary and secondary zones are specified, they must belong to the same region.

    You can add more parameters to determine other instance settings:

    Setting Parameter Notes
    Required parameters
    Database version --database-version POSTGRES_13 (default), POSTGRES_12, POSTGRES_11, POSTGRES_10, or POSTGRES_9_6
    Region --region See valid values.
    Connectivity
    Private IP --network Specifies the name of the VPC network you want to use for this instance. Private services access must already be configured for the network. Available only for the beta command (gcloud beta sql instances create). Learn more.
    Public IP --authorized-networks For public IP connections, only connections from authorized networks can connect to your instance. Learn more.
    Machine type and storage
    Machine type --tier Used to specify a shared-core instance (db-f1-micro or db-g1-small). For a flexible instance configuration, use the --cpu or --memory parameters instead. See Flexible instance configuration.
    Storage type --storage-type Determines whether your instance uses SSD or HDD storage. Learn more.
    Storage capacity --storage-size The amount of storage provisioned for the instance, in GB. Learn more.
    Automatic storage increase --storage-auto-increase Determines whether Cloud SQL automatically provides more storage for your instance when free space runs low. Learn more.
    Automatic storage increase limit --storage-auto-increase-limit Determines how large Cloud SQL can automatically grow storage. Available only for the beta command (gcloud beta sql instances create). Learn more.
    Automatic backups and high availability
    High availability --availability-type For a highly-available instance, set to REGIONAL. Learn more.
    Secondary zone --secondary-zone If you are creating an instance for high availability, you can specify both the primary and secondary zones using the >--zone and --secondary-zone parameters. The following restrictions apply when the secondary zone is used during instance creation or edit:
    • The zones must be valid zones.
    • If the secondary zone is specified, the primary must also be specified.

      • If the primary and secondary zones are specified, they must be distinct zones.

      If the primary and secondary zones are specified, they must belong to the same region.
    Automatic backups --backup-start-time The window of time when you would like backups to start. Learn more.
    Retention settings for automated backups --retained-backups-count The number of automated backups to retain. Learn more.
    Point-in-time recovery --enable-point-in-time recovery Enables point-in-time recovery and write-ahead logging. Learn more.
    Retention settings for binary logging --retained-transaction-log-days The number of days to retain write-ahead logs for point-in-time recovery.Learn more.
    Add database flags
    Database flags --database-flags You can use database flags to control settings and parameters for your instance. Learn more about database flags.
    Maintenance schedule
    Maintenance window --maintenance-window-day,
    --maintenance-window-hour    		 Determines a one-hour window when Cloud SQL can perform disruptive maintenance on your instance. If you do not set the window, then disruptive maintenance can be done at any time. Learn more.
    Maintenance timing --maintenance-release-channel Your preferred timing for instance updates, relative to other instances in the same project. Use preview for earlier updates, and production for later updates. Learn more.
  2. Note the automatically assigned IP address.

    If you are not using the Cloud SQL Auth proxy, you will use this address as the host address that your applications or tools use to connect to the instance.

  3. Set the password for the postgres user: 
    gcloud sql users set-password postgres --instance=INSTANCE_NAME --password=PASSWORD

REST API v1beta4

Create the instance

This example creates an instance with backups and binary logging enabled. These settings are optional. For a complete list of parameters for this call, see the Instances:insert page. For information about instance settings, including valid values for region, see Instance Settings.

Do not include sensitive or personally identifiable information in your instance ID; it is externally visible.
You do not need to include the project ID in the instance name. This is done automatically where appropriate (for example, in the log files).

Before using any of the request data below, make the following replacements:

  • project-id: Your project ID
  • instance-id: The desired instance ID
  • region: The desired region
  • database-version: Enum string of the database version. For example: POSTGRES_12
  • password: The password for the root user
  • machine-type: Enum string of the machine (tier) type, as: db-custom-[CPUS]-[MEMORY_MBS]

HTTP method and URL: 

POST https://sqladmin.googleapis.com/sql/v1beta4/projects/project-id/instances

Request JSON body: 

{
  "name": "instance-id",
  "region": "region",
  "databaseVersion": "database-version",
  "rootPassword": "password",
  "settings": {
    "tier": "machine-type",
    "backupConfiguration": {
      "enabled": true
    }
  }
}

To send your request, expand one of these options:

You should receive a JSON response similar to the following:

{
  "kind": "sql#operation",
  "targetLink": "https://sqladmin.googleapis.com/sql/v1beta4/projects/project-id/instances/instance-id",
  "status": "PENDING",
  "user": "user@example.com",
  "insertTime": "2020-01-01T19:13:21.834Z",
  "operationType": "CREATE",
  "name": "operation-id",
  "targetId": "instance-id",
  "selfLink": "https://sqladmin.googleapis.com/sql/v1beta4/projects/project-id/operations/operation-id",
  "targetProject": "project-id"
}

The response is a long-running operation, which may take a few minutes to complete.

Retrieve the IPv4 address

Retrieve the automatically assigned IPv4 address for the new instance:

Before using any of the request data below, make the following replacements:

  • project-id: your project ID
  • instance-id: instance ID created in prior step

HTTP method and URL: 

GET https://sqladmin.googleapis.com/sql/v1beta4/projects/project-id/instances/instance-id

To send your request, expand one of these options:

You should receive a JSON response similar to the following:

{
  "kind": "sql#instance",
  "state": "RUNNABLE",
  "databaseVersion": "MYSQL_5_7",
  "settings": {
    "authorizedGaeApplications": [],
    "tier": "db-f1-micro",
    "kind": "sql#settings",
    "pricingPlan": "PER_USE",
    "replicationType": "SYNCHRONOUS",
    "activationPolicy": "ALWAYS",
    "ipConfiguration": {
      "authorizedNetworks": [],
      "ipv4Enabled": true
    },
    "locationPreference": {
      "zone": "us-west1-a",
      "kind": "sql#locationPreference"
    },
    "dataDiskType": "PD_SSD",
    "backupConfiguration": {
      "startTime": "18:00",
      "kind": "sql#backupConfiguration",
      "enabled": true,
      "binaryLogEnabled": true
    },
    "settingsVersion": "1",
    "storageAutoResizeLimit": "0",
    "storageAutoResize": true,
    "dataDiskSizeGb": "10"
  },
  "etag": "--redacted--",
  "ipAddresses": [
    {
      "type": "PRIMARY",
      "ipAddress": "10.0.0.1"
    }
  ],
  "serverCaCert": {
    ...
  },
  "instanceType": "CLOUD_SQL_INSTANCE",
  "project": "project-id",
  "serviceAccountEmailAddress": "redacted@gcp-sa-cloud-sql.iam.gserviceaccount.com",
  "backendType": "SECOND_GEN",
  "selfLink": "https://sqladmin.googleapis.com/sql/v1beta4/projects/project-id/instances/instance-id",
  "connectionName": "project-id:region:instance-id",
  "name": "instance-id",
  "region": "us-west1",
  "gceZone": "us-west1-a"
}

Look for the ipAddress field in the response.

To see how the underlying REST API request is constructed for this task, see the APIs Explorer on the instances:insert page.

Flexible instance configurations

Flexible instance configurations let you select the amount of memory and CPUs that your instance needs. This flexibility lets you choose the appropriate VM shape for your workload. Machine type names use the format db-custom-CPU-RAM, where CPU is the number of CPUs in the machine, and RAM is the amount of memory in the machine.

When selecting the number of CPUs and amount of memory, there are some restrictions on the configuration you choose:

  • vCPUs must be either 1 or an even number between 2 and 96.
  • Memory must be:
    • 0.9 to 6.5 GB per vCPU
    • A multiple of 256 MB
    • At least 3.75 GB (3840 MB)

Troubleshooting

Click the links in the table for details:

For this problem... The issue might be... Try this...
Internal error. Missing service networking service account. Disable and re-enable the Service Networking API.
Terraform instance creation fails. Terraform configuration error. Inspect and repair the Terraform configuration file.
HTTP Error 409 in Terraform script. Another operation is already in progress. Fix the Terraform script to wait for each operation to finish.
Unknown error

The Service Networking API might not be enabled.

You may be trying to create an instance with the same name as one recently deleted.

You may be trying to create multiple instances simultaneously.

Your subnet creation may have failed if there were no more available addresses in the IP range.

Enable the Service Networking API.

Use a different name for the instance, or wait until it's been a week since the instance was deleted.

Create instances consecutively.

See other Unknown error messages if this doesn't match your case.

Allocate new ranges.
Failed to create subnetwork. No more available addresses in the IP range. Allocate new ranges.

Internal error

You see the error message {"ResourceType":"sqladmin.v1beta4.instance", "ResourceErrorCode":"INTERNAL_ERROR","ResourceErrorMessage":null}.

The issue might be

The service project is likely missing the service networking service account required for this feature.

Things to try

To repair service permissions, disable the Service Networking API , wait five minutes and then re-enable it.

Terraform instance creation fails

Terraform instance creation fails.

The issue might be

This is usually an issue within the Terraform script itself.

Things to try

Inspect and repair the Terraform configuration file.

409 error in Terraform script

You see the error message HTTP Error 409 in Terraform scripts.

The issue might be

Operation failed because another operation was already in progress

Things to try

Revise the script to halt execution until each instance operation is completed. Have the script poll and wait until a 200 is returned for the previous operation ID before continuing to the next step.

Unknown error

When trying to create an instance, you see an error message like Cloud SQL creation failed, error UNKNOWN.

The issue might be

  • The Service Networking API enabled might not be enabled.
  • You may be trying to re-use the name of an instance you recently deleted. Instance names cannot be re-used for one week after deletion.
  • You may be trying to create multiple instances simultaneously. In this case, only the first instance is created and the ret fail with Unknown error. You can only run one create operation at a time.
  • Your subnet creation may have failed if there were no more available addresses in the IP range.

Things to try

  • Enable the Service Networking API.
  • Use a different name for the instance, or wait one week to create a new one using that name.
  • Create multiple instances consecutively instead of simultaneously.
  • See the Failed to create subnetwork section below.

Failed to create subnetwork

You get the error message: Failed to create subnetwork. Couldn't find free blocks in allocated IP ranges. Please allocate new ranges for this service provider.

The issue might be

There are no more available addresses in the allocated IP range.

If you encounter this error when attempting to create a Cloud SQL instance with private IP against a Shared VPC network using private service connections. there are five possible scenarios:

  • The size of the allocated IP range for the private service connection is smaller than /24.
  • The size of the allocated IP range for the private service connection is too small for the number of Cloud SQL instances.
  • You're attempting to create both MySQL or SQL Server and PostgreSQL instances on the same private service connection in the VPC host project. MySQL and SQL Server can share the same service connection. PostgreSQL requires its own service connection.
  • You're attempting to create instances on the same private service connection in different regions, which is not supported.

Things to try

For each of the above scenarios you can elect to either expand the existing or allocate an additional IP range to the private service connection.

If you're allocating a new range, take care to not create an allocation that overlaps with any existing allocations.

After creating a new IP range, update the vpc peering with the following command:

gcloud services vpc-peerings update --service=servicenetworking.googleapis.com
--ranges=[OLD_RESERVED_RANGE_NAME],[NEW_RESERVED_RANGE_NAME] --network=[VPC_NETWORK]
--project=[PROJECT_ID] --force

If you're expanding an existing allocation, take care to only increase the allocation range and not decrease it. For example, if the original allocation was 10.0.10.0/24 then make the new allocation at least 10.0.10.0/23.

In general, if starting from a /24 allocation, decrementing the /mask by 1 for each condition (additional instance type group, additional region) is a good rule of thumb. For example, if trying to create both instance type groups on the same allocation, going from /24 to /23 is enough.

After expanding an existing IP range, update the vpc peering with following command:

gcloud services vpc-peerings update --service=servicenetworking.googleapis.com
--ranges=[RESERVED_RANGE_NAME] --network=[VPC_NETWORK] --project=[PROJECT_ID]

What's next