Introduction
When Kleene runs an extract, it sends requests to the source API. If the API returns an error, Kleene surfaces the HTTP status code in the logs. This article explains the most common connector error codes, what they usually mean, and how to resolve them.
Common connector error codes
| HTTP status code | Meaning | How to resolve |
|---|---|---|
| 400 | Bad request | Check the permissions on the API resource being accessed, and review the configuration of your source or extract. |
| 401 | Incorrect authentication | Confirm that your API credentials are correct and still valid. |
| 403 | Insufficient permissions | Check that the authenticated user or API credential has enough permissions and scope to call the API. |
| 404 | Not found | Check for a temporary service outage and confirm that the API resource is still active and available. |
| 414 | URI too long | Use a custom version of the report if applicable, and reduce the number of parameters until the request is accepted by the API. |
| 429 | Too many requests | Reduce the number of concurrent extract runs, and spread out the schedules of extracts in the same source. |
| 500 | Internal server error | The source server is down. Try again later. |
| 502 | Bad gateway | The server returned an invalid response while acting as a gateway or proxy. Try again later. |
| 503 | Service unavailable | The server is not ready to handle the request. Try again later. |
| 504 | Gateway timeout | The gateway did not get a response in time. Reduce the volume of data being queried, and reduce backfill settings if applicable. |
How to troubleshoot connector errors
- Check the HTTP status code in the Kleene logs.
- Match the code to the descriptions in this article.
- Apply the recommended fix for the relevant error.
- Run the extract again to confirm the issue is resolved.
When to try again later
For server-side errors such as 500, 502, 503, and 504, the issue is usually temporary and outside your configuration. If you have already reduced the request size or volume where applicable, the best next step is often to wait and retry later.
Additional information
If you continue to see the same error after checking the relevant settings, review your source configuration, authentication details, and API documentation for any service-specific limits or requirements.