diff --git a/controller/README.md b/controller/README.md index 696a25e93ed91b635962ff82444dc46f8df26c8b..c4f1596821de72d7cd4df1b76179af1a23763053 100644 --- a/controller/README.md +++ b/controller/README.md @@ -1,11 +1,40 @@ # Controller The controller application controls the AS Solver workflow and allows communication with it via REST. The requests should have a JSON content with appropriate fields filled. -The required fields are described in /src/models directory +The required fields are also described in /src/models directory # Rest end-points to communicate with our Solver - [POST] /train - Train the network with models extracted from CDO + - application_id: unique id of the application + - cp_cdo_path: path to the constraint problem in CDO + - camel_model_cdo_path: path to the camel model in CDO + - proc_finished_url: URL at which a notification will be sent after training has finished + - max_utility_error: maximal MAE error of the predicted configurations to deem a trained network successful + - force_retrain: whether to force retraining of the model. Forcing retraining when another training is in progress may + lead to undefined behaviour. + +- [POST] /constraintProblemSolution - Solve the constraint problem with models extracted from CDO + - application_id: unique id of the application + - problem_datetime: time for which the prediction will be calculated + - cp_CDO_path: path to the constraint problem in CDO + - notification_uri: uri of the ESB at which a notification will be sent + - request_uuid: request uuid which will be passed in the notification to ESB + - [POST] /trainFromFile - Train the network with models extracted from files -- [POST] /solveCP - Solve the constraint problem with models extracted from CDO -- [POST] /solveCPFromFile - Solve the constraint problem with models extracted from files \ No newline at end of file + - application_id: unique id of the application + - cp_file_path: path to the file containing constraint problem + - camel_model_file_path: path to the file containing camel model + - node_candidates_file_path: path to the file containing node candidates + - proc_finished_url: URL at which a notification will be sent after training has finished + - max_utility_error: maximal MAE error of the predicted configurations to deem a trained network successful + - force_retrain: whether to force retraining of the model. Forcing retraining when another training is in progress may + lead to undefined behaviour. + +- [POST] /constraintProblemSolutionFromFile - Solve the constraint problem with network input extracted from file + - application_id: unique id of the application + - input_file_path: path to the file containing the input for supervised gym + - cp_CDO_path: path to the constraint problem in CDO + - notification_uri: uri of the ESB at which a notification will be sent + - request_uuid: request uuid which will be passed in the notification to ESB + \ No newline at end of file