Improve error handling for perspectives Lambda errors

[hablutzel1]

Currently, when making a simple request like:

{
    "check_type": "caa",
    "domain_or_ip_target": ""
}

a failure occurs within the perspectives Lambda and the returned error response looks like this:

{
...
    "perspectives": [
        {
            "perspective_code": "eu-west-2",
...
                "errors": [
                    {
                        "error_type": "mpic_error:coordinator:remote_check",
                        "error_message": "The remote check failed to complete. - Check failed for perspective eu-west-2: 'body'"
                    }
                ],

Where the error doesn't provide too much useful information.

With this update, the full error details are provided, including the Lambda execution exception and stack trace:

{
...
    "perspectives": [
        {
            "perspective_code": "eu-west-2",
...
                "errors": [
                    {
                        "error_type": "mpic_error:coordinator:remote_check",
                        "error_message": "The remote check failed to complete. - Check failed for perspective eu-west-2: Lambda execution error: {\"errorMessage\": \"Invalid domain name: Empty domain\", \"errorType\": \"ValueError\", \"requestId\": \"6f89dbce-f259-4a7b-849f-9d1b1e36a29d\", \"stackTrace\": [\"  File \\\"/opt/python/lib/python3.11/site-packages/aws_lambda_powertools/middleware_factory/factory.py\\\", line 137, in wrapper\\n    response = middleware()\\n\", \"  File \\\"/opt/python/lib/python3.11/site-packages/aws_lambda_powertools/utilities/parser/parser.py\\\", line 119, in event_parser\\n    return handler(parsed_event, context, **kwargs)\\n\", \"  File \\\"/var/task/mpic_caa_checker_lambda_function.py\\\", line 84, in lambda_handler\\n    return get_handler().process_invocation(event)\\n\", \"  File \\\"/var/task/mpic_caa_checker_lambda_function.py\\\", line 34, in process_invocation\\n    caa_response = event_loop.run_until_complete(self.caa_checker.check_caa(caa_request))\\n\", \"  File \\\"/var/lang/lib/python3.11/asyncio/base_events.py\\\", line 654, in run_until_complete\\n    return future.result()\\n\", \"  File \\\"/opt/python/lib/python3.11/site-packages/open_mpic_core/mpic_caa_checker/mpic_caa_checker.py\\\", line 88, in check_caa\\n    caa_request.domain_or_ip_target = DomainEncoder.prepare_target_for_lookup(caa_request.domain_or_ip_target)\\n\", \"  File \\\"/opt/python/lib/python3.11/site-packages/open_mpic_core/common_util/domain_encoder.py\\\", line 25, in prepare_target_for_lookup\\n    raise ValueError(f\\\"Invalid domain name: {str(e)}\\\")\\n\"]}"
                    }
                ],

Although returning the complete stack trace may seem excessive, I think it accelerates debugging and since this service is internal to the CAs, exposing detailed errors in responses might be considered safe.

[birgelee]

@hablutzel1 this looks great thanks. I agree about the point of the service being internal to CAs justifying the full error response.