Skip to main content
Setting up Database Access with GCP Cloud SQL
Setting up Database Access with GCP Cloud SQL

Length: 12:32

Version: 14.x

Database Access with PostgreSQL on GCP Cloud SQL

Teleport can provide secure access to PostgreSQL on Google Cloud SQL via the Teleport Database Service. This allows for fine-grained access control through Teleport's RBAC.

In this guide, you will:

  1. Configure an PostgreSQL on Google Cloud SQL with a service account.
  2. Join the PostgreSQL on Google Cloud SQL database to your Teleport cluster.
  3. Connect to the PostgreSQL on Google Cloud SQL database via the Teleport Database Service.

Self-Hosted Teleport Architecture for Cloud SQL Access

Prerequisites​

  • A running Teleport cluster version 14.3.33 or above. If you want to get started with Teleport, sign up for a free trial or set up a demo environment.

  • The tctl admin tool and tsh client tool.

    Visit Installation for instructions on downloading tctl and tsh.

  • Google Cloud account
  • Command-line client psql installed and added to your system's PATH environment variable.
  • A host, e.g., a Compute Engine instance, where you will run the Teleport Database Service
  • To check that you can connect to your Teleport cluster, sign in with tsh login, then verify that you can run tctl commands using your current credentials. tctl is supported on macOS and Linux machines. For example: 
    $ tsh login --proxy=teleport.example.com [email protected]
    $ tctl status
    # Cluster  teleport.example.com
    # Version  14.3.33
    # CA pin   sha256:abdc1245efgh5678abdc1245efgh5678abdc1245efgh5678abdc1245efgh5678
    If you can connect to the cluster and run the tctl status command, you can use your current credentials to run subsequent tctl commands from your workstation. If you host your own Teleport cluster, you can also run tctl commands on the computer that hosts the Teleport Auth Service for full permissions.

Step 1/9. Create a service account for the Teleport Database Service​

A GCP service account will be used by the Teleport Database Service to create ephemeral access tokens for other GCP service accounts when it's acting on the behalf of authorized Teleport users.

Create a service account​

Go to the Service Accounts page and create a service account:

Create Service Account

(Optional) Grant permissions​

The Teleport Database Service needs permissions to automatically download your Cloud SQL instance's root CA certificate and to general an ephemeral client certificate.

If you intend to download the CA certificate manually and your Cloud SQL instance SSL mode is not "require trusted client certificates", then you can skip this step.

Otherwise, in the second step of the service account creation dialogue, assign the service account the pre-defined GCP IAM role "Cloud SQL Client", then click "Done".

Cloud SQL Client permissions

The "Cloud SQL Client" role has the following permissions:

# Used to auto-download the instance's root CA certificate and check SSL mode.
cloudsql.instances.get
# Used to generate an ephemeral client certificate
cloudsql.instances.connect

If you only need one of these permissions, you can define and assign a custom IAM role to the service account instead.

Step 2/9. Create a service account for a database user​

Teleport uses service accounts to connect to Cloud SQL databases.

Create a service account​

Go to the IAM & Admin Service Accounts page and create a new service account named "cloudsql-user":

Create Service Account

Click "Create and continue".

Grant permissions​

On the second step grant this service account the "Cloud SQL Instance User" role which will allow it to connect to Cloud SQL instances using an IAM token for authentication:

Grant Cloud SQL Instance User to Service Account

Click "Done".

Grant access to the service account​

The Teleport Database Service must be able to impersonate this service account. Navigate to the "cloudsql-user" service account overview page and select the "permissions" tab:

Select Service Account Permissions Tab

Click "Grant Access" and add the "teleport-db-service" principal ID. Select the "Service Account Token Creator" role and save the change:

Grant Service Account Token Creator to Database Service

Service account permissions

The "Service Account Token Creator" IAM role includes more permissions than the Teleport Database Service needs. To further restrict the service account, you can create a role that includes only the following permission:

# Used to generate IAM auth tokens when connecting to a database instance.
iam.serviceAccounts.getAccessToken

Step 3/9. Configure your Cloud SQL database​

Teleport uses IAM database authentication with Cloud SQL PostgreSQL instances.

If you're creating a new PostgreSQL instance, make sure to add the cloudsql.iam_authentication database flag under "Customize your instance / Flags" section:

Enable IAM Authentication

To check whether IAM authentication is enabled for an existing Cloud SQL instance, look for the flag on the Configuration panel on the instance's Overview page:

Check IAM Authentication

If it isn't enabled, you can add this flag using the "Edit configuration" dialog at the bottom of the Configuration panel. Changing this setting may require a database instance reboot.

Create a database user​

Now go back to the Users page of your Cloud SQL instance and add a new user account. In the sidebar, choose "Cloud IAM" authentication type and add the "cloudsql-user" service account that you created in the second step:

Add Cloud SQL User Account

Press "Add" and your Users table should look similar to this:

Cloud SQL User Accounts Table

See Creating and managing IAM users in Google Cloud documentation for more info.

Step 4/9. Install Teleport​

Select an edition, then follow the instructions for that edition to install Teleport.

The following command updates the repository for the package manager on the local operating system and installs the provided Teleport version:

$ curl https://cdn.teleport.dev/install-v14.3.33.sh | bash -s 14.3.33

Step 5/9. Configure the Teleport Database Service​

Create a join token​

The Database Service requires a valid join token to join your Teleport cluster. Run the following tctl command and save the token output in /tmp/token on the server that will run the Database Service:

$ tctl tokens add --type=db --format=text
abcd123-insecure-do-not-use-this

(Optional) Download the Cloud SQL CA certificate​

The Cloud SQL instance's root CA certificate is required so that the Teleport Database Service can validate the certificate presented by the database instance.

The Teleport Database Service can automatically download the instance's root CA certificate if it is granted the "cloudsql.instances.get" permission.

Alternatively, you can download the instance's CA certificate file from the "Connections" tab under the "Security" section:

Instance Root Certificate

Generate Teleport config​

Provide the following information and then generate a configuration file for the Teleport Database Service:

  • example.teleport.sh:443 The host and port of your Teleport Proxy Service or Enterprise Cloud site
  • public-ip The Cloud SQL instance public IP address. The address can be found on the "Connect to this instance" panel on the "Overview" page in the Cloud SQL instance's dashboard.
  • project-id The GCP project ID. You can normally see it in the organization view at the top of the GCP dashboard.
  • instance-id The name of your Cloud SQL instance.
$ sudo teleport db configure create \
   -o file \
   --name=cloudsql \
   --protocol=postgres \
   --labels=env=dev \
   --token=/tmp/token \
   --proxy=example.teleport.sh:443  \
   --uri=public-ip:5432 \
   --gcp-project-id=project-id \
   --gcp-instance-id=instance-id

This command will generate a Teleport Database Service configuration file and save it to /etc/teleport.yaml.

Step 6/9. Configure GCP credentials​

The Teleport Database Service must have credentials for the teleport-db-service GCP service account.

If the Teleport Database Service is hosted on a GCE instance, you can change the attached service account. For non-GCE deployments of Teleport, we recommend using workload identity.

Alternatively, go to that service account's Keys tab and create a new key:

Service Account Keys

Make sure to choose JSON format:

Service Account New Key

Save the file. Set the GOOGLE_APPLICATION_CREDENTIALS environment variable to point to the JSON credentials file you downloaded earlier. For example, if you use systemd to start teleport, then you should edit the service's EnvironmentFile to include the env var:

$ echo 'GOOGLE_APPLICATION_CREDENTIALS=/path/to/credentials.json' | sudo tee -a /etc/default/teleport
warning

A service account key can be a security risk - we only describe using a key in this guide for simplicity. We do not recommend using service account keys in production. See authentication in the Google Cloud documentation for more information about service account authentication methods.

Step 7/9. Start the Teleport Database Service​

Configure the Teleport Database Service to start automatically when the host boots up by creating a systemd service for it. The instructions depend on how you installed the Teleport Database Service.

On the host where you will run the Teleport Database Service, enable and start Teleport:

$ sudo systemctl enable teleport
$ sudo systemctl start teleport

You can check the status of the Teleport Database Service with systemctl status teleport and view its logs with journalctl -fu teleport.

Step 8/9. Create a Teleport user​

tip

To modify an existing user to provide access to the Database Service, see Database Access Access Controls

Create a local Teleport user with the built-in access role:

$ tctl users add \
  --roles=access \
  --db-users="*" \
  --db-names="*" \
  alice
FlagDescription
--rolesList of roles to assign to the user. The builtin access role allows them to connect to any database server registered with Teleport.
--db-usersList of database usernames the user will be allowed to use when connecting to the databases. A wildcard allows any user.
--db-namesList of logical databases (aka schemas) the user will be allowed to connect to within a database server. A wildcard allows any database.
warning

Database names are only enforced for PostgreSQL and MongoDB databases.

For more detailed information about database access controls and how to restrict access see RBAC documentation.

Step 9/9. Connect​

Once the Database Service has joined the cluster, log in to see the available databases:

$ tsh login --proxy=teleport.example.com --user=alice
$ tsh db ls
# Name     Description              Labels
# -------- ------------------------ --------
# cloudsql GCP Cloud SQL PostgreSQL env=dev
note

You will only be able to see databases that your Teleport role has access to. See our RBAC guide for more details.

When connecting to the database, use the name of the database's service account that you added as an IAM database user above, minus the ".gserviceaccount.com" suffix. The database user name is shown on the Users page of your Cloud SQL instance. Retrieve credentials for the "cloudsql" example database and connect to it:

$ tsh db connect [email protected] --db-name=postgres cloudsql

To log out of the database and remove credentials:

# Remove credentials for a particular database instance:
$ tsh db logout cloudsql
# Or remove credentials for all databases:
$ tsh db logout

Troubleshooting​

Could not find default credentials​

This error can come from either your client application or Teleport.

For a client application, ensure that you disable GCP credential loading. Your client should not attempt to load credentials because GCP credentials will be provided by the Teleport Database Service.

If you see the credentials error message in the Teleport Database Service logs (at DEBUG log level), then the Teleport Database Service does not have GCP credentials configured correctly.

If you are using a service account key, then ensure that the environment variable GOOGLE_APPLICATION_CREDENTIALS=/path/to/credentials.json is set and restart your Teleport Database Service to ensure that the env var is available to teleport. For example, if your Teleport Database Service runs as a systemd service:

$ echo 'GOOGLE_APPLICATION_CREDENTIALS=/path/to/credentials.json' | sudo tee -a /etc/default/teleport
$ sudo systemctl restart teleport

See authentication in the Google Cloud documentation for more information about service account authentication methods.

Unable to cancel a query​

If you use a PostgreSQL cli client like psql, and you try to cancel a query with ctrl+c, but it doesn't cancel the query, then you need to connect using a tsh local proxy instead. When psql cancels a query, it establishes a new connection without TLS certificates, however Teleport requires TLS certificates not only for authentication, but also to route database connections.

If you enable TLS Routing in Teleport then tsh db connect will automatically start a local proxy for every connection. Alternatively, you can connect via Teleport Connect which also uses a local proxy. Otherwise, you need to start a tsh local proxy manually using tsh proxy db and connect via the local proxy.

If you have already started a long-running query in a psql session that you cannot cancel with ctrl+c, you can start a new client session to cancel that query manually:

First, find the query's process identifier (PID):

SELECT pid,usename,backend_start,query FROM pg_stat_activity WHERE state = 'active';

Next, gracefully cancel the query using its PID. This will send a SIGINT signal to the postgres backend process for that query:

SELECT pg_cancel_backend(<PID>);

You should always try to gracefully terminate a query first, but if graceful cancellation is taking too long, then you can forcefully terminate the query instead. This will send a SIGTERM signal to the postgres backend process for that query:

SELECT pg_terminate_backend(<PID>);

See the PostgreSQL documentation on admin functions for more information about the pg_cancel_backend and pg_terminate_backend functions.

Next steps​

  • See the YAML configuration reference for updating dynamic resource matchers or static database definitions.