Database Connections

You can connect directly to your own databases—useful for building dashboards on custom applications, data warehouses, or legacy systems.

Supported databases

PostgreSQL
MySQL / MariaDB
Microsoft SQL Server
IBM DB2
Snowflake — see the Snowflake setup guide

Prerequisites

You’ll need this info from your database administrator:

Host/IP: The server’s network address
Port: Connection port (e.g., 5432 for Postgres, 3306 for MySQL)
Database name: The specific schema to access
Read-only credentials: Username and password with SELECT permissions

Setup instructions

Step 1: Whitelist our IPs

If your database is behind a firewall, you’ll need to whitelist our sync servers. Contact support@resplendentdata.com for the current IP list.

Step 2: Connect in Resplendent Data

Go to Settings → Integrations
Find your database type and click Connect on the card
Enter your connection details:
- Host
- Port
- Database name
- Username & password
- Require SSL (off by default). Turn this on for cloud hosts that require encrypted connections, such as Supabase, Neon, or Amazon RDS.
Click Test Connection

Sync performance and refresh rates

Tier	Refresh frequency
Free Forever	Every 60 minutes
Starter	Every 15 minutes
Professional / Business / Starship	Every 60 seconds

Incremental sync

For larger tables, incremental sync pulls only changed rows after the initial load. You choose:

Unique Key — one column, or a Virtual Primary Key built from multiple columns when no single column identifies each row
Sorting Column — a value that advances whenever a row is inserted or edited, such as updated_at

The native SQL connectors let you select these columns even when the database has no primary-key or unique constraint on them. Catalog metadata may suggest defaults, but you are responsible for confirming that the identity stays unique and that the sorting column tracks changes. SQL Server binary rowversion/timestamp columns cannot be used as sorting columns.

Without a suitable identity and sorting column, use snapshot syncing instead.

Incremental history and updates use bounded batches. Timestamp updates replay a five-minute overlap to cover equal timestamps and short commit delays. Changes backdated beyond that window and hard deletes require a resync. Adding columns also performs a full resync of the configured scope: existing rows are replaced, and incremental history loads again in batches.

SQL discovery and Eric inspection

Eric inspects schema, declared keys, row-count evidence, and explicitly requested column statistics before proposing a sync for native SQL sources and Snowflake. Source records become available to Eric only after you approve the Dataset and it syncs. Planner statistics from the source are estimates and may be stale. Snowflake aggregate profiling is bounded; unavailable statistics remain unknown.

Snowflake standard-table key declarations are checked against aggregate counts when requested; you still need to establish that those keys remain stable. A column name alone is not evidence that it tracks changes. Without a suitable identity and watermark, choose a snapshot of up to 100,000 rows.

Explore with SQL Studio

Open SQL Studio from the Dataset picker to run a read-only SELECT or WITH … SELECT against native SQL sources before adding a Dataset. SQL Studio returns at most 1,000 rows and allows up to five minutes per query. Only users with full Dataset access can open it.

On-prem connections run SQL Studio through your sync agent. IBM DB2 requires an on-prem agent with the IBM i Access ODBC driver; hosted DB2 connections are not supported. Snowflake SQL Studio, including SHOW commands and custom SELECT Datasets, is documented in the Snowflake setup guide.

On-premise databases

If your database isn’t on the public internet, use the Resplendent Sync Agent (available on Business tier and higher) for MySQL, SQL Server, IBM DB2, and other native SQL connectors except Snowflake. Snowflake uses the cloud connector.

IBM DB2 connections require an on-prem sync agent with the IBM i Access ODBC driver installed on the agent host. Deploy the updated agent when you enable SQL Studio or incremental sync for on-prem SQL sources.