Transformations > Machine Learning transformation > Error handling
  

Error handling

The Machine Learning transformation handles errors based on the type of issue that occurs and the response code that it receives from the REST API. You can review the response details in the Spark executor log.
The mapping fails when the following types of issues occur:
The following table describes the action that occurs for each response code:
HTTP Status Code
Action
100 Continue
Row is skipped.
Partial requests are not permitted.
101,102,103 Informational, interim
Mapping fails.
3XX Redirection
Row is skipped.
Mapping fails if the redirected URL cannot be reached or the redirected URL returns a code that fails the mapping.
400 Bad request
Row is skipped.
401 Unauthorized
Mapping fails.
403 Forbidden
Mapping fails.
404 Resource not found
Mapping fails.
405 Method not allowed
Mapping fails.
407 Proxy authentication required
Request is tried again.
408 Request timeout
Mapping fails.
409 Conflict
Row is skipped.
415 Unsupported media type
Mapping fails.
418 I'm a teapot
Mapping fails.
421 Misdirected request
Mapping fails.
423 Resource is locked
Row is skipped.
429 Too many requests
Mapping fails.
500 Internal server error
Mapping fails.
501 Not implemented
Mapping fails.
502 Bad gateway
Mapping fails.
503 Service unavailable
Mapping fails.
504 Gateway timeout
Mapping fails.
505 HTTP version not supported
Mapping fails.
506 Server internal configuration error
Mapping fails.
511 Network authentication required
Mapping fails.