finds.dev

datavane/datavines

Know your data better!Datavines is Next-gen Data Observability Platform, support metadata manage and data quality.

View on GitHubHomepage ↗

★ 767JavaApache-2.0updated Jul 2026

// the take

Datavines is a self-hosted data observability platform for data quality checks, data profiling, and metadata/catalog management across warehouses like MySQL, Doris, StarRocks, ClickHouse, and others. It's aimed at data engineering or platform teams who want an open-source alternative to tools like Great Expectations or Monte Carlo, with jobs that run either on Spark or a lightweight local JDBC engine.

The plugin architecture is real, not just marketing: data sources, check rules, execution engines, alert channels, and the registry are all separately pluggable, so swapping in a new warehouse connector or check type doesn't mean forking core code. Supporting both a Spark engine and a no-dependency local JDBC engine means you can run functional checks without standing up a cluster first, then move to Spark once data volume demands it. The 27 built-in check rules cover single-table, cross-table, and custom SQL checks, which is a decent baseline without writing your own rule engine. Server nodes are designed to scale horizontally with job-level fault tolerance, so a single node crash shouldn't lose or duplicate a running check.

The Spark engine is locked to Spark 2.4, which has been end-of-life for years — if your data platform runs on anything current this is a hard compatibility wall, not a config tweak. Alerting is email-only out of the box; there's no Slack, webhook, or PagerDuty integration, which is a real gap for a tool whose whole point is flagging data problems fast. Failed-row error data storage only works with MySQL or local files, and only when running on the Local engine, so Spark-based checks give you a pass/fail without the sample rows you'd need to actually debug a failure. At under 800 stars with a fairly small contributor base for something this operationally heavy (ZooKeeper, multiple DB backends, Spark integration), expect to lean on reading the source and filing issues rather than finding answers already written up.

We read the README, the directory tree and a key source file in November 2026. Nobody pays to be listed. How we judge →

// at a glance

Stars
767
Forks
212
Language
Java
License
Apache-2.0
Last commit
27 Jul 2026

// nearby

If datavines isn't quite it.

All repos →