What do the import session phases mean (analyzing, testing, forwarding)?

When you create a new data import, your setup goes through several phases. You can see the current phase in the progress indicator on the Data Import page:

Pending

Your import request has been created and is queued for processing. This phase is usually very brief (a few seconds).

Analyzing

The AI is reading your data source — parsing the API specification, CSV columns, or webhook payload. It identifies meters, maps fields, and calculates confidence scores. This phase typically takes 10–30 seconds depending on the complexity of your data source.

Testing

For API connections, the system runs a live test request against your vendor's endpoint. It verifies that authentication works, the URL is reachable, and the response matches the expected format. If the test fails, you will see an error with a suggestion (e.g., "Check your API key" or "Use HTTPS instead of HTTP").

Forwarding

The validated configuration is being sent to the data service for scheduling. This includes all field mappings, meter definitions, authentication credentials, and fetch intervals. This phase is typically instant.

Completed

Your import source is fully configured and active. For API connections, the scheduler will begin fetching data on the configured interval. For webhook sources, the endpoint is registered and ready to receive data.

Failed

Something went wrong during one of the phases above. Click on the import to see the specific error message and suggestion. Common causes: invalid API credentials, unreachable URL, unsupported file format, or the AI could not identify any meter data in the source.

What to do if stuck on "analyzing": Wait up to 60 seconds. If it stays in this phase, the AI may be processing a very large specification. If it times out, try simplifying your input (e.g., upload only the relevant portion of a large Swagger spec).