How to load data from Postgres to Kafka

Learn how to use Airbyte to synchronize your Postgres data into Kafka within minutes.

Trusted by data-driven companies

Building your pipeline or Using Airbyte

Airbyte is the only open source solution empowering data teams  to meet all their growing custom business demands in the new AI era.

Building in-house pipelines
Bespoke pipelines are:
  • Inconsistent and inaccurate data
  • Laborious and expensive
  • Brittle and inflexible
Furthermore, you will need to build and maintain Y x Z pipelines with Y sources and Z destinations to cover all your needs.
After Airbyte
Airbyte connections are:
  • Reliable and accurate
  • Extensible and scalable for all your needs
  • Deployed and governed your way
All your pipelines in minutes, however custom they are, thanks to Airbyte’s connector marketplace and AI Connector Builder.

Start syncing with Airbyte in 3 easy steps within 10 minutes

Set up a Postgres connector in Airbyte

Connect to Postgres or one of 400+ pre-built or 10,000+ custom connectors through simple account authentication.

Set up Kafka for your extracted Postgres data

Select Kafka where you want to import data from your Postgres source to. You can also choose other cloud data warehouses, databases, data lakes, vector databases, or any other supported Airbyte destinations.

Configure the Postgres to Kafka in Airbyte

This includes selecting the data you want to extract - streams and columns -, the sync frequency, where in the destination you want that data to be loaded.

Take a virtual tour

Check out our interactive demo and our how-to videos to learn how you can sync data from any source to any destination.

Demo video of Airbyte Cloud

Demo video of AI Connector Builder

What sets Airbyte Apart

Modern GenAI Workflows

Streamline AI workflows with Airbyte: load unstructured data into vector stores like Pinecone, Weaviate, and Milvus. Supports RAG transformations with LangChain chunking and embeddings from OpenAI, Cohere, etc., all in one operation.

Move Large Volumes, Fast

Quickly get up and running with a 5-minute setup that supports both incremental and full refreshes, for databases of any size.

An Extensible Open-Source Standard

More than 1,000 developers contribute to Airbyte’s connectors, different interfaces (UI, API, Terraform Provider, Python Library), and integrations with the rest of the stack. Airbyte’s AI Connector Builder lets you edit or add new connectors in minutes.

Full Control & Security

Airbyte secures your data with cloud-hosted, self-hosted or hybrid deployment options. Single Sign-On (SSO) and Role-Based Access Control (RBAC) ensure only authorized users have access with the right permissions. Airbyte acts as a HIPAA conduit and supports compliance with CCPA, GDPR, and SOC2.

Fully Featured & Integrated

Airbyte automates schema evolution for seamless data flow, and utilizes efficient Change Data Capture (CDC) for real-time updates. Select only the columns you need, and leverage our dbt integration for powerful data transformations.

Enterprise Support with SLAs

Airbyte Self-Managed Enterprise comes with dedicated support and guaranteed service level agreements (SLAs), ensuring that your data movement infrastructure remains reliable and performant, and expert assistance is available when needed.

What our users say

Jean-Mathieu Saponaro
Data & Analytics Senior Eng Manager

"The intake layer of Datadog’s self-serve analytics platform is largely built on Airbyte.Airbyte’s ease of use and extensibility allowed any team in the company to push their data into the platform - without assistance from the data team!"

Learn more
Chase Zieman headshot
Chase Zieman
Chief Data Officer

“Airbyte helped us accelerate our progress by years, compared to our competitors. We don’t need to worry about connectors and focus on creating value for our users instead of building infrastructure. That’s priceless. The time and energy saved allows us to disrupt and grow faster.”

Learn more
Alexis Weill
Data Lead

“We chose Airbyte for its ease of use, its pricing scalability and its absence of vendor lock-in. Having a lean team makes them our top criteria.
The value of being able to scale and execute at a high level by maximizing resources is immense”

Learn more

How to Sync Postgres to Kafka Manually

1. Download Kafka: Go to the Apache Kafka website and download the latest version of Kafka.

2. Extract Kafka: Unzip the downloaded file to your desired location.

3. Start ZooKeeper: Kafka uses ZooKeeper, so you need to start it before starting Kafka.

```shell

bin/zookeeper-server-start.sh config/zookeeper.properties

```

4. Start Kafka Server: Open another terminal and run the Kafka server.

```shell

bin/kafka-server-start.sh config/server.properties

```

1. Create Topic: Create a Kafka topic where the data from PostgreSQL will be published.

```shell

bin/kafka-topics.sh --create --topic postgres-data --bootstrap-server localhost:9092 --replication-factor 1 --partitions 1

```

1. Install PostgreSQL: If you haven't already, install PostgreSQL on your system.

2. Create a Database: Create a database and tables that you want to move data from.

3. Enable Logical Replication: Modify `postgresql.conf` to enable logical replication by setting `wal_level` to `logical`.

4. Create Publication: For PostgreSQL 10 and above, you can create a publication for the tables you want to watch.

```sql

CREATE PUBLICATION my_publication FOR TABLE my_table;

```

1. Set Up Your Development Environment: Make sure you have a suitable programming environment with necessary dependencies installed, such as a PostgreSQL driver and Kafka client library.

2. Database Connection: Write code to connect to your PostgreSQL database.

3. Polling or Trigger-Based Data Retrieval: Decide on a method for retrieving data from PostgreSQL. You can poll the database at intervals or use triggers to act on data changes.

4. Read Data: Write a function to read data from PostgreSQL. This could be new rows or updated rows depending on your application.

5. Format Data: Format the data into a structure suitable for sending to Kafka.

6. Kafka Producer: Write a Kafka producer in your application that connects to the Kafka cluster.

7. Send Data to Kafka: Send the formatted data to the Kafka topic you created earlier.

8. Error Handling: Implement error handling to deal with any issues during data retrieval or publishing.

9. Logging: Add logging to your application for monitoring and debugging purposes.

1. Compile and Run: Compile your application and run it.

2. Monitor: Monitor the application for any errors and ensure data is being published to Kafka successfully.

1. Consume Messages: Use a Kafka consumer to consume messages from the Kafka topic to validate that data is being moved correctly.

```shell

bin/kafka-console-consumer.sh --bootstrap-server localhost:9092 --topic postgres-data --from-beginning

```

2. Check Data Integrity: Verify that the data in Kafka matches the source data in PostgreSQL.

1. Deploy: Deploy your custom application to a suitable production environment.

2. Scale: If necessary, scale your Kafka cluster and application to handle the load.

1. Set Up Monitoring: Set up monitoring for both your Kafka cluster and custom application to ensure they are running smoothly.

2. Regular Maintenance: Perform regular maintenance checks on both PostgreSQL and Kafka.

Additional Considerations:

Security: Ensure that your Kafka cluster and PostgreSQL database are secured, using SSL/TLS and SASL if necessary.

Transaction Support: If your application requires transaction support, make sure to handle transactions properly in your custom application.

Data Serialization: Choose an appropriate data serialization format for Kafka messages (e.g., JSON, Avro, Protobuf).

Backpressure Handling: Implement backpressure handling in your application in case Kafka cannot handle the rate of incoming messages.

Remember, this approach requires a significant amount of custom development and testing. It's important to consider the trade-offs between building a custom solution and using existing third-party connectors that may offer additional features and robustness.

How to Sync Postgres to Kafka Manually - Method 2: PostgreSQL Kafka connector using Confluent CLI

First, let’s see what connectors are at our disposal:

```

confluent connect plugin list

```

Look for “PostgresSource” in the output. If you don’t see it, double-check your Confluent CLI setup.

Before we configure our connector, let’s examine its properties:

```

confluent connect plugin describe PostgresSource

```

This command provides a wealth of information. Take some time to read through it – understanding these properties will help you troubleshoot issues later.

Now comes the fun part – creating your configuration file. I recommend using a text editor you’re comfortable with. Let’s call our file postgres-to-kafka.json:

```json

{

  "name": "pg-to-kafka-stream",

  "config": {

    "connector.class": "PostgresSource",

    "kafka.auth.mode": "SERVICE_ACCOUNT",

    "kafka.service.account.id": "sa-abc123",

    "topic.prefix": "pg_",

    "connection.host": "your-db-host.com",

    "connection.port": "5432",

    "connection.user": "your_username",

    "connection.password": "your_password",

    "db.name": "your_database",

    "table.whitelist": "users,orders",

    "timestamp.column.name": "updated_at",

    "incrementing.column.name": "id",

    "output.data.format": "JSON",

    "db.timezone": "UTC",

    "tasks.max": "1",

    "mode": "timestamp+incrementing"

  }

}

```

Some notes on this configuration:

  • The name field is how you’ll refer to this connector later. Choose something memorable!
  • kafka.auth.mode: Using a service account is generally more secure than API keys.
  • topic.prefix: This helps organize your Kafka topics. I like using short prefixes like “pg_”.
  • mode: “timestamp+incrementing” is often a good choice for capturing both new and updated rows.

Remember, this is just a starting point. You may need to adjust based on your specific PostgreSQL setup.

With your configuration file ready, it’s time to create the connector:

```

confluent connect cluster create --config-file postgres-to-kafka.json

```

If successful, you’ll get a response with your connector’s ID. Save this ID – you’ll need it for monitoring.

Connecting systems can be tricky. Let’s check on our connector’s status:

```

confluent connect cluster status <your-connector-id>

```

Look for “RUNNING” in the output. If you see “FAILED”, check the error messages and revisit your configuration.

Now for the moment of truth – is data flowing into Kafka? Let’s check:

```

confluent kafka topic consume pg_users --from-beginning

```

Replace “pg_users” with whatever your topic name is based on your prefix and table name.

If you see JSON data streaming by, congratulations! You’ve successfully connected PostgreSQL to Kafka.

FAQs

ETL, an acronym for Extract, Transform, Load, is a vital data integration process. It involves extracting data from diverse sources, transforming it into a usable format, and loading it into a database, data warehouse or data lake. This process enables meaningful data analysis, enhancing business intelligence.

An object-relational database management system, PostgreSQL is able to handle a wide range of workloads, supports multiple standards, and is cross-platform, running on numerous operating systems including Microsoft Windows, Solaris, Linux, and FreeBSD. It is highly extensible, and supports more than 12 procedural languages, Spatial data support, Gin and GIST Indexes, and more. Many webs, mobile, and analytics applications use PostgreSQL as the primary data warehouse or data store.

PostgreSQL gives access to a wide range of data types, including:  

1. Numeric data types: This includes integers, floating-point numbers, and decimal numbers.  

2. Character data types: This includes strings, text, and character arrays.  

3. Date and time data types: This includes dates, times, and timestamps.  

4. Boolean data types: This includes true/false values.  

5. Network address data types: This includes IP addresses and MAC addresses.  

6. Geometric data types: This includes points, lines, and polygons.  

7. Array data types: This includes arrays of any of the above data types.  

8. JSON and JSONB data types: This includes JSON objects and arrays.  

9. XML data types: This includes XML documents.  

10. Composite data types: This includes user-defined data types that can contain multiple fields of different data types.  

Overall, PostgreSQL's API provides access to a wide range of data types, making it a versatile and powerful tool for data management and analysis.

This can be done by building a data pipeline manually, usually a Python script (you can leverage a tool as Apache Airflow for this). This process can take more than a full week of development. Or it can be done in minutes on Airbyte in three easy steps: 
1. Set up PostgreSQL to Kafka as a source connector (using Auth, or usually an API key)
2. Choose a destination (more than 50 available destination databases, data warehouses or lakes) to sync data too and set it up as a destination connector
3. Define which data you want to transfer from PostgreSQL to Kafka and how frequently
You can choose to self-host the pipeline using Airbyte Open Source or have it managed for you with Airbyte Cloud. 

ELT, standing for Extract, Load, Transform, is a modern take on the traditional ETL data integration process. In ELT, data is first extracted from various sources, loaded directly into a data warehouse, and then transformed. This approach enhances data processing speed, analytical flexibility and autonomy.

ETL and ELT are critical data integration strategies with key differences. ETL (Extract, Transform, Load) transforms data before loading, ideal for structured data. In contrast, ELT (Extract, Load, Transform) loads data before transformation, perfect for processing large, diverse data sets in modern data warehouses. ELT is becoming the new standard as it offers a lot more flexibility and autonomy to data analysts.

What should you do next?

Hope you enjoyed the reading. Here are the 3 ways we can help you in your data journey:

flag icon
Easily address your data movement needs with Airbyte Cloud
Take the first step towards extensible data movement infrastructure that will give a ton of time back to your data team. 
Get started with Airbyte for free
high five icon
Talk to a data infrastructure expert
Get a free consultation with an Airbyte expert to significantly improve your data movement infrastructure. 
Talk to sales
stars sparkling
Improve your data infrastructure knowledge
Subscribe to our monthly newsletter and get the community’s new enlightening content along with Airbyte’s progress in their mission to solve data integration once and for all.
Subscribe to newsletter