// the find
apecloud/ape-dts
ApeCloud's Data Transfer Suite, written in Rust. Provides ultra-fast data replication between MySQL, PostgreSQL, Redis, MongoDB, Kafka and ClickHouse, ideal for disaster recovery (DR) and migration scenarios.
ape-dts is a CDC and snapshot migration tool written in Rust that moves data between MySQL, PostgreSQL, Redis, MongoDB, and analytical targets like ClickHouse, Doris, and StarRocks. It's aimed at teams running database migrations or setting up DR pipelines who find Debezium too heavy. The no-external-dependencies design means you can drop a single binary and go.
The benchmark numbers are striking — 30x Debezium's throughput on MySQL snapshot migration using the same hardware, with a 71MB image vs Debezium's 1.38GB. The data verification story (check/revise/review modes) is something most CDC tools skip entirely and is genuinely useful for migration validation. Lua scripting for in-flight transformations is a practical escape hatch that avoids needing a full stream processor just to rename a column. Resume-from-breakpoint on both snapshot and CDC tasks is table stakes but it's here and documented well.
Target support is asymmetric — MySQL gets connectors to 7+ destinations, PostgreSQL gets fewer, and MongoDB/Redis are mostly homogeneous only. If you need mongo-to-postgres or redis-to-kafka, you're out of luck. The community skews heavily Chinese (WeChat as the primary support channel, bilingual docs where the Chinese version appears more complete) which is fine if you read it, but adds friction for English-only teams hitting edge cases. Configuration is INI files with a lot of surface area — the config docs are long and there's no schema validation, so typos fail silently at runtime rather than startup. No managed cloud offering or Kubernetes operator means you own the operability story entirely.