CloudSQL - PostgreSQL

Create an integration to manage access to PostgreSQL instances on Google Cloud SQL

Google Cloud SQL PostgreSQL is a fully managed relational database service built for the cloud. It provides a high-performance, scalable, and highly available PostgreSQL database instance without the overhead of managing infrastructure. With Google Cloud SQL, users benefit from Google Cloud's robust infrastructure, which ensures high availability, security, and scalability for their databases.

Through this integration, Apono helps you securely manage access to your Google Cloud SQL PostgreSQL database instances.

To enable Apono to manage Google Cloud SQL PostgreSQL user access, you must create a user and then configure the integration within the Apono UI.

---

Prerequisites

Item	Description
Apono Connector	On-prem connection serving as a bridge between your Google Cloud PostgreSQL databases and Apono Minimum Required Version: 1.4.1 Use the following steps to update an existing connector.
Cloud SQL Admin API	API for managing database instances with resources, such as BackupRuns, Databases, and Instances
Cloud SQL Admin Role	(Cloud IAM authentication only) Google Cloud role that the Apono connector's service user must have at the instance's project or organization level
PostgreSQL Info	Information for the database instance to be integrated: Project ID Dataset Name

---

Create a PostgreSQL user

You must create a user in your PostgreSQL instance for the Apono connector and grant that user permissions to your databases.

You must use the admin account and password to connect to your database.

Following these steps to create a user and grant it permissions:

1. In the Google Cloud console, create a new user with either Built-in authentication or Cloud IAM authentication.

Built-In Authentication

Use apono_connector for the username.

This authentication method grants the user the cloudsqlsuperuser role. Be sure to set a strong password for the user.

As an alternative, you can run the following command from your Postgre client:

CREATE USER 'apono_connector'@'%' IDENTIFIED BY 'password'

Cloud IAM

Use apono-connector-iam-sa@[PROJECT_ID].iam.gserviceaccount.com for the Principal.

This authentication method does not grant the user account database privileges.

Be sure that the Apono connector GCP service account (apono-connector-iam-sa@[PROJECT_ID].iam.gserviceaccount.com) has the Cloud SQL Admin role.

1. (Cloud IAM only) In your preferred client tool, grant cloudsqlsuperuser access to the user account.

ALTER ROLE "<CONNECTOR_USERNAME>" WITH CREATEROLE;
GRANT cloudsqlsuperuser TO "<CONNECTOR_USERNAME>";

1. In your preferred client tool, grant the cloudsqlsuperuser role privileges on all databases except template0 and cloudsqladmin. This allows Apono to perform tasks that are not restricted to a single schema or object within the database, such as creating, altering, and dropping database objects.

DO $$
DECLARE
  database_name text;
BEGIN
  FOR database_name IN (SELECT datname FROM pg_database WHERE datname != 'template0' AND datname != 'cloudsqladmin') LOOP
    EXECUTE 'GRANT ALL PRIVILEGES ON DATABASE ' || quote_ident(database_name) || ' TO cloudsqlsuperuser WITH GRANT OPTION';
  END LOOP;
END; $$

1. For each database to be managed through Apono, connect to the database and grant cloudsqlsuperuser privileges on all objects in the schemas. This allows Apono to perform tasks that are restricted to schemas within the database, such as modifying table structures, creating new sequences, or altering functions.

DO $$
DECLARE
  schema text;
BEGIN
  FOR schema IN (SELECT schema_name FROM information_schema.schemata WHERE schema_name NOT LIKE 'pg_%' AND schema_name != 'information_schema' AND schema_name != 'cron') LOOP
    EXECUTE 'GRANT ALL PRIVILEGES ON SCHEMA ' || quote_ident(schema) || ' TO cloudsqlsuperuser WITH GRANT OPTION';
    EXECUTE 'GRANT ALL PRIVILEGES ON ALL TABLES IN SCHEMA ' || quote_ident(schema) || ' TO cloudsqlsuperuser WITH GRANT OPTION';
    EXECUTE 'GRANT ALL PRIVILEGES ON ALL SEQUENCES IN SCHEMA ' || quote_ident(schema) || ' TO cloudsqlsuperuser WITH GRANT OPTION';
    EXECUTE 'GRANT ALL PRIVILEGES ON ALL FUNCTIONS IN SCHEMA ' || quote_ident(schema) || ' TO cloudsqlsuperuser WITH GRANT OPTION';
  END LOOP;
  EXECUTE 'ALTER DEFAULT PRIVILEGES GRANT ALL PRIVILEGES ON TABLES TO cloudsqlsuperuser WITH GRANT OPTION';
  EXECUTE 'ALTER DEFAULT PRIVILEGES GRANT ALL PRIVILEGES ON SEQUENCES TO cloudsqlsuperuser WITH GRANT OPTION';
  EXECUTE 'ALTER DEFAULT PRIVILEGES GRANT ALL PRIVILEGES ON FUNCTIONS TO cloudsqlsuperuser WITH GRANT OPTION';
  EXECUTE 'ALTER DEFAULT PRIVILEGES GRANT ALL PRIVILEGES ON SCHEMAS TO cloudsqlsuperuser WITH GRANT OPTION';
END; $$

1. Connect to the template1 database and grant cloudsqlsuperuser privileges on all objects in the schemas. For any new databases created in the future, this allows Apono to perform tasks that are restricted to schemas within the database, such as modifying table structures, creating new sequences, or altering functions.

DO $$
DECLARE
  schema text;
BEGIN
  FOR schema IN (SELECT schema_name FROM information_schema.schemata WHERE schema_name NOT LIKE 'pg_%' AND schema_name != 'information_schema' AND schema_name != 'cron') LOOP
    EXECUTE 'GRANT ALL PRIVILEGES ON SCHEMA ' || quote_ident(schema) || ' TO cloudsqlsuperuser WITH GRANT OPTION';
    EXECUTE 'GRANT ALL PRIVILEGES ON ALL TABLES IN SCHEMA ' || quote_ident(schema) || ' TO cloudsqlsuperuser WITH GRANT OPTION';
    EXECUTE 'GRANT ALL PRIVILEGES ON ALL SEQUENCES IN SCHEMA ' || quote_ident(schema) || ' TO cloudsqlsuperuser WITH GRANT OPTION';
    EXECUTE 'GRANT ALL PRIVILEGES ON ALL FUNCTIONS IN SCHEMA ' || quote_ident(schema) || ' TO cloudsqlsuperuser WITH GRANT OPTION';
  END LOOP;
  EXECUTE 'ALTER DEFAULT PRIVILEGES GRANT ALL PRIVILEGES ON TABLES TO cloudsqlsuperuser WITH GRANT OPTION';
  EXECUTE 'ALTER DEFAULT PRIVILEGES GRANT ALL PRIVILEGES ON SEQUENCES TO cloudsqlsuperuser WITH GRANT OPTION';
  EXECUTE 'ALTER DEFAULT PRIVILEGES GRANT ALL PRIVILEGES ON FUNCTIONS TO cloudsqlsuperuser WITH GRANT OPTION';
  EXECUTE 'ALTER DEFAULT PRIVILEGES GRANT ALL PRIVILEGES ON SCHEMAS TO cloudsqlsuperuser WITH GRANT OPTION';
END; $$

1. (Built-in authentication only) Create a secret with the credentials from step 1.

When using Cloud IAM authentication, the service account and its permissions are managed through Google Cloud IAM roles and policies. The service account is used to authenticate to the Cloud SQL instance.

A secret does not need to be created.

---

Integrate Google Cloud SQL - PostgreSQL

Google Cloud SQL - PostgreSQL

You can also use the steps below to integrate with Apono using Terraform.

In step 11, instead of clicking Confirm, follow the Are you integrating with Apono using Terraform? guidance.

Follow these steps to complete the integration:

1. On the Catalog tab, click Google Cloud SQL - PostgreSQL. The Connect Integration page appears.
2. Under Discovery, click one or more resource types and cloud services to sync with Apono.

Apono automatically discovers and syncs all the instances in the environment. After syncing, you can manage access flows to these resources.

1. Click Next. The Apono connector section expands.

2. From the dropdown menu, select a connector.

If the desired connector is not listed, click + Add new connector and follow the instructions for creating a GCP connector.

1. Click Next. The Integration Config section expands.

2. Define the Integration Config settings.\

   Setting	Description
   Integration Name	Unique, alphanumeric, user-friendly name used to identify this integration when constructing an access flow
   Auth Type	Authorization type for the MySQL service account user: IAM Auth: Cloud IAM authentication User / Password: Built-in authentication
   Project ID	ID of the project where the PostgreSQL instance is deployed
   Region	Location where the PostgreSQL instance is deployed
   Instance ID	ID of the PostgreSQL instance
   Instance ID User Override	(Optional) Allows overriding the instance ID for the user
   Database Name	Name of the database to integrate By default, Apono sets this value to postgre.
   SSL Mode	(Optionl) Mode of Secure Sockets Layer (SSL) encryption used to secure the connection with the SQL database server: require: An SSL-encrypted connection must be used. allow: An SSL-encrypted or unencrypted connection is used. If an SSL-encrypted connection is unavailable, the unencrypted connection is used. disable: An unencrypted connection is used. prefer: An SSL-encrypted connection is attempted. If the encrypted connection is unavailable, the unencrypted connection is used. verify-ca: An SSL-encrypted connection must be used and a server certification verification against the provided CA certificates must pass. verify-full: An SSL-encrypted connection must be used and a server certification verification against the provided CA certificates must pass. Additionally, the server hostname is checked against the certificate's names.

3. Click Next. The Secret Store section expands.

4. (User/Password only) Associate the secret or credentials.

A secret is not needed or Cloud IAM authentication.

1. Click Next. The Get more with Apono section expands.

2. Define the Get more with Apono settings.\

   Setting	Description
   Credential Rotation	(Optional) Number of days after which the database credentials must be rotated Learn more about the Credentials Rotation Policy.
   User cleanup after access is revoked (in days)	(Optional) Defines the number of days after access has been revoked that the user should be deleted Learn more about Periodic User Cleanup & Deletion.
   Custom Access Details	(Optional) Instructions explaining how to access this integration's resources Upon accessing an integration, a message with these instructions will be displayed to end users in the User Portal. The message may include up to 400 characters. To view the message as it appears to end users, click Preview.
   Integration Owner	(Optional) Fallback approver if no resource owner is found Follow these steps to define one or several integration owners: From the Attribute dropdown menu, select User or Group under the relevant identity provider (IdP) platform. From the Value dropdown menu, select one or multiple users or groups. NOTE: When Resource Owner is defined, an Integration Owner must be defined.
   Resource Owner	(Optional) Group or role responsible for managing access approvals or rejections for the resource Follow these steps to define one or several resource owners: Enter a Key name. This value is the name of the tag created in your cloud environment. From the Attribute dropdown menu, select an attribute under the IdP platform to which the key name is associated. Apono will use the value associated with the key (tag) to identify the resource owner. When you update the membership of the group or role in your IdP platform, this change is also reflected in Apono. NOTE: When this setting is defined, an Integration Owner must also be defined.

3. Click Confirm.

💡Are you integrating with Apono using Terraform?

If you want to integrate with Apono using Terraform, follow these steps instead of clicking Confirm:

1. At the top of the screen, click View as Code. A modal appears with the completed Terraform configuration code.

2. Click to copy the code.

3. Make any additional edits.

4. Deploy the code in your Terraform.

Refer to Integration Config Metadata for more details about the schema definition.

Now that you have completed this integration, you can create access flows that grant permission to your Google Cloud SQL PostgreSQL instance.