Table Of Contents

• Prerequisites
  • PosgreSQL
    • Configuration
    • Using a Minimum Priviledged User
  • Redis
• Installation & Usage
  • Create a Pipeline
  • View Existing Pipelines
  • Teardown a Pipeline

Prerequisites

In order for Willow to successfully replicate from PostgreSQL to Redis, the following prerequisites must be met:

• PostgreSQL server is set up for logical replication
• Docker is installed on the server running Willow
• Redis server must have RedisJSON and RediSearch modules installed

PostgreSQL

Instructions are written based on Debezium's documentation for configuring PostgreSQL. The following instructions assume that the PostgreSQL user used for replication is named willow.

Configuration

Ensure your PostgreSQL server is configured to support logical replication with the pgoutput plugin. This includes:

1. Using PostgreSQL version 10+.
2. Setting the WAL (write-ahead log) level to logical.

ALTER SYSTEM SET wal_level = logical;

3. Adding entries to your PostgreSQL's pg_hba.conf file to allow the server running Willow to replicate from the PostgreSQL host.

local   replication   willow                      trust
host    replication   willow     127.0.0.1/32     trust
host    replication   willow     ::1/128          trust

4. Restart the PostgreSQL server for the changes to take effect.

Using a Minimum Privileged User

In order to replicate data from PostgreSQL's WAL, you must provide Willow with a PostgreSQL user. This user can either be a SUPERUSER or have the minimum required privileges.

The following SQL command can be run in the PostgreSQL terminal to create a SUPERUSER. Be sure to provide a unique password.

CREATE ROLE willow WITH LOGIN SUPERUSER PASSWORD <password>

Minimum privileged users must have the following privileges:

• REPLICATION
• LOGIN
• database level CREATE
• table level SELECT

The following SQL commands can be run in the PostgreSQL terminal to create a minimum privileged user:

/* Create a willow user with the REPLICATION and LOGIN permissions */
CREATE ROLE willow REPLICATION LOGIN PASSWORD <password>;


/* Create a willow_replication group for sharing table level privileges with original table owners */
CREATE ROLE willow_replication;


/* Repeat the following command for all databases containing tables you wish to replicate */

/* Provide willow_replication with database level CREATE privileges */
GRANT CREATE ON DATABASE <database_name> TO willow_replication;


/* Repeat the remaining commands for all tables you wish to replicate. */

/* Place the willow user and the original table owner in the willow_replication group */
GRANT willow_replication TO <original_table_owner>;
GRANT willow_replication TO willow;

/* Make the willow_replication group the table owner and grant SELECT permissions */
ALTER TABLE <table_name> OWNER TO willow_replication;
GRANT SELECT ON <table_name> TO willow_replication;

Redis

Ensure your Redis cache has both the RedisJSON and RediSearch modules installed. Redis Stack by default includes both of these modules and is a simple way to ensure your Redis cache is compatible with Willow.

Installation & Usage

To install and start up Willow, clone this repo and run:

docker compose up

Navigate to Willow's home page at https://localhost:3000.

Create a Pipeline

To initialize pipeline setup, click the "CREATE A CDC PIPELINE" button.

Enter PostgreSQL database details, including:

• The host and port to access the PostgreSQL server
• The database name containing the tables to replicate
• The username and password for Willow to use when replicating from PostgreSQL

Select which tables and columns to replicate and provide a unique source connection name. Primary keys (identified with PK) cannot be deselected. Tables without primary keys are not able to be replicated and are not shown.

Enter Redis cache details, including:

• The URL for the Redis cache. Must start with redis://.
• The username and password for Willow to use when replicating to Redis.

Click the "VERIFY" button to verify that the Redis cache is accessible.

Provide a unique sink connection name and click "SUBMIT" to finish CDC pipeline creation.

After pipeline creation, an initial snapshot of the selected tables is taken to prepopulate the cache. Subsequent INSERT, UPDATE, DELETE, and TRUNCATE commands that affect the selected tables will be reflected in the cache.

View Existing Pipelines

All existing pipelines can be viewed at https://localhost:3000/pipelines.

A single pipeline can be clicked on to view details on the associated source and sink connectors.

Teardown a Pipeline

Pipelines can be torn down by clicking the trash can icon.