dbmazz vs OpenLogReplicator
dbmazz
github.comOpenLogReplicator
bersler.comSide by side
| Figure | dbmazz | OpenLogReplicator |
|---|---|---|
| Price from | Free | Free |
| Free plan | Yes | Yes |
| Free trial | — | — |
| Open source | — | — |
| macOS | Yes | — |
| Linux | Yes | Yes |
| Self-hosted | Yes | Yes |
| Plans published | 2 | 1 |
| Spec score · Change Data Capture Software | 8.0 (#9) | 7.6 (#13) |
| Support | The project points users to GitHub Discussions and Issues, and its documentation lists support by email or GitHub issue. | Production assistance and proof-of-concept help are available by contacting the author through the contact form. |
Plans
In their own words
dbmazz
dbmazz is a Rust daemon for moving database changes into a sink. It reads PostgreSQL changes through logical replication and streams inserts, updates, and deletes. The project lists PostgreSQL and MySQL as sources, and StarRocks, PostgreSQL, and Snowflake as stable sinks; S3/Iceberg is planned. Each instance handles one replication job. The project reports sub-second replication lag and about 11 MB of resident memory per daemon. Teams can try its ez-cdc CLI quickstart or run the daemon from the official Docker image. The CLI supports Linux and macOS on amd64 and arm64. Listed observability tools include Prometheus metrics, gRPC reflection, an HTTP control plane, and a live terminal dashboard. Automatic setup can create publications, replication slots, sink tables, and audit columns. Self-hosting is free; BYOC is also listed at free in a customer's AWS or GCP account, while support and optional PrivateLink cost extra. The Elastic License 2.0 permits commercial and non-commercial use but restricts offering the software to third parties as a managed service.
OpenLogReplicator
OpenLogReplicator is a free, open-source Oracle change data capture solution written in C++ for self-hosted Linux deployments. It reads transactions directly from Oracle binary redo logs and streams changes as JSON or Protobuf. Destinations include Kafka, flat files, plain TCP/IP or ZeroMQ network streams, and a discard output for testing. The log-based design is intended to minimize impact on the source database. Archived redo logs can be processed, including offline operation while the database is shut down, and actual DDL commands can be tracked. Basic DML filters can restrict processing, such as selecting INSERT operations for a table. Direct Prometheus metrics include DML counters that can carry schema-name tags. Oracle 23c is supported; binary XMLType support for Oracle 11.2 through 21c is experimental. Debezium use is supported with version 1.3.0 as the minimum and 1.9.0 recommended for Debezium 2.x and 3.x. The GPL v3 plan costs nothing.