Connectors
Connectors are configured data sources that feed data into the DataPool. Each connector defines where to get data, how often to fetch it, and how to map it into the DataPool’s stream/item/field hierarchy.
Connector Types
| Type | Data Source | Fetch Mode |
|---|---|---|
| SQL | SQL Server, MySQL, PostgreSQL, SQLite | Polling (configurable interval) |
| File | CSV, Excel, JSON, XML | Upload, URL polling, or watch folder |
| Webhook | HTTP POST | Push (real-time) |
| Premium | Custom API integration | Push via ingest API |
Creating a Connector
Navigate to Data Sources
From the sidebar, click Data Sources.
Create new connector
Click New Connector and select the connector type.
Configure the connection
Each connector type has its own configuration. See the individual connector pages for details.
Map the data
Define how source data maps to DataPool streams, items, and fields:
- Stream name — The top-level collection name
- Item key — Which source field uniquely identifies each item
- Field mapping — Which source columns/properties become DataPool fields
Test the connection
Click Test to verify the connection and see sample data.
Save and activate
Save the connector. Data begins flowing immediately.
Ingest Keys
Each current-generation connector is assigned a Data Ingest API key in the format ik_<key_id>_<secret>. Think of it like a Stripe-style API credential:
key_idis the public selector used by Inspire to resolve the connector quicklysecretis the private proof that authenticates the request- The full key is presented as one Bearer token in API calls
- The full key can be rotated without recreating the connector
Treat the full ingest key like a password. You can store the key_id in logs or dashboards, but never commit or expose the full ik_<key_id>_<secret> credential.
How connectors use the key
For webhooks and custom API integrations, send the full key in the Authorization header:
Authorization: Bearer ik_<key_id>_<secret>The connector detail page only shows the complete key when it is first created or rotated, so copy it into your connector configuration or secret store immediately.
Webhook Signing Key
Webhook connectors can optionally require HMAC-SHA256 signature verification by setting a Webhook Signing Key on the connector. When configured, every incoming webhook request must include valid X-Webhook-Signature and X-Webhook-Timestamp headers, preventing forged payloads from being written to the DataPool.
See Webhook Signature Verification for header format, computation, and code examples.
Connector State
Each connector displays its current state:
| State | Description |
|---|---|
| Active | Running normally, data flowing |
| Idle | Connected but no recent data |
| Error | Connection or data error — check logs |
| Disabled | Manually paused by a user |
Connector Logs
Each connector maintains a log of recent activity:
- Successful data fetches with item count
- Errors with details and timestamps
- Connection attempts and results
View logs from the connector detail page. Logs are retained for 30 days.
Monitoring
The Data Sources page shows a summary of all connectors with:
- Status indicators (green/yellow/red)
- Last successful fetch time
- Item count
- Error count in the last 24 hours
Best Practices
| Practice | Why |
|---|---|
| Set appropriate polling intervals | Too frequent wastes resources; too slow means stale data |
| Use unique item keys | Prevents duplicate items in the DataPool |
| Monitor error counts | Catch data source issues before they affect displays |
| Name streams descriptively | sales-by-region is better than data1 |
| Test before deploying | Verify data mapping before going live |