This guide provides step-by-step instructions to connect your GCP Cloud SQL database to Sequin.

​
Provision a Postgres user for Sequin

When in development, it’s probably fine to supply Sequin with an existing user.

However, when in production, you should create a dedicated user for Sequin. The user needs the following permissions:

  1. connect permission on the database.
  2. select permission on all the tables you want to connect to Sequin.
  3. replication permission to read from replication slots.

Here’s how to create a dedicated user with the minimum required permissions:

-- Create user with a secure password
create user sequin_user with password 'REPLACE_WITH_SECURE_PASSWORD';

-- Grant connect permission
grant connect on database your_database to sequin_user;

-- Grant select permission on tables you want to replicate
grant select on table table1, table2, table3 to sequin_user;
-- OR grant select on all tables in a schema
grant select on all tables in schema public to sequin_user;

-- Grant replication permission
alter user sequin_user with replication;

To generate a secure password, if you have OpenSSL installed, you can use the following command:

openssl rand -base64 32

​
Enable logical replication on GCP Cloud SQL

By default, logical replication is not enabled on GCP Cloud SQL. You can double check if logical replication is enabled by connecting to your database and running the following command:

SHOW rds.logical_replication;

If the output is off, then logical replication is not enabled.

To enable it, follow these steps:

Enabling replication requires a restart of your database instance.
1

Enable pglogical flag

In the Google Cloud Console, navigate to your Cloud SQL instance and follow these steps:

  1. Go to the “Configuration” tab for your database instance
  2. Click “Edit configuration”
  3. Under “Flags” section, add a new flag: cloudsql.enable_pglogical=on
  4. Click “Save” to apply the changes
2

Restart your database instance

After setting the flag, you need to restart your Cloud SQL instance for the changes to take effect. In the Google Cloud Console:

  1. Go to the “Overview” tab of your Cloud SQL instance
  2. Click the “Restart” button at the top of the page
  3. Confirm the restart when prompted

​
Connect Sequin to your GCP Cloud SQL database

After enabling logical replication on GCP Cloud SQL, you’ll now connect to your database in Sequin.

1

Enter connection details in Sequin

In the Sequin Console, click on the Connect Database button and enter the credentials for your GCP Cloud SQL database:

You can find these connection details in your GCP Cloud SQL instance overview.

  • Host: Your Cloud SQL instance IP address
  • Port: 5432 (default Postgres port)
  • Database: Your database name
  • Username: The sequin database user you created earlier
  • Password: The password for your sequin database user

Make sure to enable the “SSL” option for secure connection.

2

Create a publication

Connect to your database using the SQL client of your choice and execute the following SQL query to create a publication:

CREATE PUBLICATION sequin_pub FOR TABLE table1, table2, table3 WITH (publish_via_partition_root = true);

If you want to publish changes from all tables, you can use:

CREATE PUBLICATION sequin_pub FOR ALL TABLES WITH (publish_via_partition_root = true);
3

Create a replication slot

Next, create a replication slot to capture changes from the publication:

SELECT pg_create_logical_replication_slot('sequin_slot', 'pgoutput');
4

Enter the replication slot details in Sequin

Back in the Sequin Console, enter the name of the replication slot (e.g. sequin_slot) and publication (e.g. sequin_pub) you just created. Then, name your database and click “Create Database”.

​
Create a sink

With your GCP Cloud SQL database connected to Sequin, you are ready to create a sink. Follow one of our guides below to get started:

Stream to GCP Pub/Sub

Send changes to GCP Pub/Sub topics to trigger Cloud Functions and power event-driven architectures

Stream to Webhooks

Send database changes to your HTTP endpoints to trigger workflows and keep services in sync

Stream to Redis

Stream changes to Redis streams for real-time data processing and caching

Stream to Kafka

Publish database changes to Kafka topics for event streaming and processing

Was this page helpful?