Composite storage — the right database for every job
DataLake
One logical data lake, three storage engines. Time-series data in InfluxDB 2.x or TimescaleDB, relational data in PostgreSQL or MS SQL, unstructured files (images, PDFs, videos, reports) in S3-compatible object storage like MinIO. Each workload lands where it performs best — you query them all through the same DataBridge REST API.
Key Features
What DataLake Does
// RIGHT DB FOR THE RIGHT JOB
One lake, three engines. Each workload lands where it performs best.
One REST API via DataBridge · edge-to-cloud replication · offline-first
Time-series engines
InfluxDB 2.x and TimescaleDB for millions of points per second — compression, downsampling, continuous aggregates, hot/warm/cold retention policies.
Relational engines
PostgreSQL or Microsoft SQL Server for business data — assets, batches, orders, maintenance history, quality results. Joins, ACID, referential integrity.
Object storage (S3 / MinIO)
Images, PDFs, videos, lab reports, raw audit logs. S3-compatible API — keep them on-premise in MinIO or push to any cloud bucket. Versioned, lifecycle-managed.
One API, every backend
DataBridge sits on top and exposes a single REST API. You never learn three query languages — you learn one. Edge-to-cloud replication works offline-first.
How It Works
Getting Started with DataLake
Four steps from zero to production.
Route by workload
FlowMaker and DataBridge pick the right backend automatically — time-series goes to InfluxDB/Timescale, relational to Postgres/MS SQL, files to MinIO/S3.
Index & compress
Columnar compression for time-series, normalized tables for relational, content-hash keys for object storage. 10× smaller than raw.
Tier automatically
Retention policies per tag group, hot/warm/cold tiers, lifecycle rules on object storage. Data migrates between tiers without downtime.
Query anywhere
REST API, WebSocket, Grafana plugin, native connectors for Power BI and Python. One grammar, any backend.
Integrations