pgstream

pgstream is a free, open-source PostgreSQL change-data-capture tool, available as a command-line program or Go library. It suits teams operating their own PostgreSQL infrastructure who need to replicate data and schema changes. Its snapshots and masking options are useful, but initial-release limits make it a poor fit for pipelines that need row-level filtering or broader Kafka topic support.

Overview

pgstream uses log-based capture to replicate PostgreSQL data and DDL changes to PostgreSQL, Elasticsearch or OpenSearch, Kafka, and webhooks. Tracking schema changes alongside row changes helps downstream targets keep up when tables evolve, rather than receiving data updates alone.

Teams can run it as a standalone CLI or embed it as a Go library. PostgreSQL is its only stated dependency; pgstream does not require ZooKeeper or Kafka to run. That is a focused deployment footprint, but self-hosting still means your team is responsible for operating the tool and its database environment. The project is licensed under Apache License 2.0, and its README points users to GitHub issues for community assistance.

Key features

Snapshots and replication

Initial snapshots can bootstrap a target with a bulk load, while on-demand snapshots support point-in-time copies. Ongoing log-based replication then keeps changes moving. This combination is well suited to initializing a target or keeping a staging database in step with production without treating the initial load as a separate problem.

Schema tracking and destinations

Because pgstream carries DDL changes as well as data changes, it can help downstream systems follow table changes. Built-in targets include PostgreSQL, Elasticsearch and OpenSearch, Kafka, and webhooks. Kafka support includes schema-based partitioning, but the initial release supports a single Kafka topic, a meaningful constraint for teams dividing events across topics.

Column transformation

Column values can be transformed during replication with greenmask, neosync, go-masker, or custom transformers. This gives teams a way to mask production data as it moves into development or staging environments, reducing exposure in those copies. It does not remove the need to decide which transformations suit the data and destination.

Initial-release constraints

The initial release supports only the wal2json PostgreSQL plugin and has no row-level filtering. It also requires a primary key or a unique, non-null column. These limits narrow its usefulness for sources that cannot meet that key requirement, or pipelines that need selective row capture or multiple Kafka topics.

Pricing

pgstream costs nothing. The free plan is the open-source CDC CLI and Go library under Apache License 2.0, with no paid tier described. That makes it accessible to teams willing to run it themselves; the trade-off is community assistance through GitHub issues rather than a stated paid support plan.

Platforms

pgstream supports Linux and macOS and is self-hosted. Installation options include a Go binary, building from source, or Homebrew on macOS or Linux. Its CLI and Go library forms give teams a choice between running it directly and embedding it in Go software, but it is not a managed service.

Who it's for

pgstream is a strong fit for teams that already operate PostgreSQL and want to keep staging databases or other supported targets current, especially when initial snapshots, DDL replication, or masking production values matter. It is less suitable when a pipeline needs row-level filtering, a PostgreSQL plugin beyond wal2json, or more than one Kafka topic. Teams unable to provide a primary key or unique non-null column should also look elsewhere.

Pros and cons

• Pros: Initial and on-demand snapshots support both bulk bootstrap and point-in-time copies.
• Pros: Replicating DDL alongside data helps downstream targets follow schema changes.
• Pros: Column transformation options support masking data as it moves to development or staging.
• Pros: The Apache-licensed CLI and Go library can run without ZooKeeper or Kafka as dependencies.
• Cons: The initial release supports only one Kafka topic and the wal2json PostgreSQL plugin.
• Cons: No row-level filtering limits pipelines that need selective replication.
• Cons: A primary key or unique non-null column is required, which can rule out some source tables.
• Cons: Self-hosting and community assistance through GitHub issues place operations and support expectations on the adopting team.

Alternatives

Canal is another free, open-source option for teams seeking a self-hosted CDC tool across Linux and macOS, with an API platform also listed.

Sequin offers a free, self-hosted open-source option for teams that want to run it on their own infrastructure, with API and web platforms also listed.

Apache Flink CDC is a free Apache-licensed choice for teams needing a CDC option across Linux, macOS, Windows, and self-hosted environments.

Debezium is a free, Apache-licensed distributed change-data-capture platform for teams seeking that broader platform approach, with API, Linux, and Windows platforms listed.

dbmazz offers a free self-hosted open-source engine with a CLI, terminal dashboard, local web UI, and community support.

Estuary is a freemium option with a free plan and trial; its Pay-as-you-go plan is $100 a month, billed monthly. Choose it when a freemium offering is preferable to a wholly self-hosted open-source tool.

Striim is a freemium alternative with a free plan and trial, including a developer plan with 25 million events per month; it may suit readers comparing event-based usage limits.

TiCDC is another free, open-source CDC option for Linux, macOS, and self-hosted environments, with API also listed.

Browse Change Data Capture Software for more options.

Verdict

Choose pgstream if your team can self-host and wants PostgreSQL replication that carries schema changes, supports snapshots, and can mask values on the way to staging or other targets. Its free Apache-licensed distribution and focused dependency footprint are compelling; look elsewhere if you need row filtering, a different PostgreSQL plugin, or multiple Kafka topics.