Deploying Datanyx on your own infrastructure
This guide covers the end-to-end on-prem install: system requirements, deployment, licensing, and post-install verification. Plan on a half-day for a standard install.
~15 min read | For: IT and infrastructure teams running Datanyx self-hosted
System requirements
Before you begin, confirm your target environment meets these minimums.
| Component | Minimum |
|---|---|
| OS | Ubuntu 20.04+, RHEL 8+, or equivalent Linux |
| CPU | 8 cores (16 recommended for >20 users) |
| Memory | 32 GB RAM (64 GB recommended) |
| Storage | 250 GB SSD, additional storage for data volumes |
| Database | PostgreSQL 13+ (for the Datanyx metadata store) |
| Container runtime | Docker 20.10+ or Kubernetes 1.24+ |
| Network | Outbound HTTPS for license validation and updates |
Installation steps
Step 1 — Prepare your environment
Provision the host, install the OS, verify the CPU/memory/storage meet the minimums, and ensure your network allowlist permits outbound HTTPS.
Step 2 — Download the Datanyx installer
Use the download link provided by your account manager. Installers are signed — verify the checksum before running.
Step 3 — Run the installer
The installer walks you through configuring the database connection, storage paths, and admin credentials.
Step 4 — Apply your license
Your license key determines which products are active. Apply it during the installer or from the Admin › License screen after first login.
Step 5 — Verify the install
Open the Datanyx URL in a browser. You should see the first-login screen. Run the built-in health check from Admin › System to confirm all services are up.
After install, the flow is the same as SaaS. Name your org, invite your team, connect a datasource, assign roles. See the SaaS onboarding guide from step 2 onward — the only difference is where Datanyx is hosted.
Ongoing responsibilities
Unlike the SaaS version, on-prem customers own the operational side:
Backups. The metadata database and data volumes need regular backups. Datanyx recommends daily incremental, weekly full.
Updates. New Datanyx releases ship roughly quarterly. Update windows should be scheduled during low-usage periods.
Monitoring. CPU, memory, disk, and database health should be monitored continuously. Datanyx exports Prometheus metrics.
Scaling. Horizontal scaling is supported for query workers and workflow runners.