Waiting for Execution to Complete
You can monitor the state of execution by issuing a GET to the provided
statusPath.
The result is a JSON object containing the following values:
- id: The id of the execution.
- status: Text representing the response from the executing runtime. Valid values include OK and NOT_COMPLETED.
- exitCode: The value returned by the model or problem's exit statement, or 0 if there was no explicit exit.
- statusPath: The path from which the current status can be refreshed.
- inputPath, resultPath, runLogPath: Paths to the relevant files for the current execution.
- modelParameters: A record of how parameters were set prior to execution - this takes the form of a JSON object mapping names to text values.
Of the above, only start, id and statusPath are guaranteed to be present. Other properties may be present, depending on the state of the execution.
The four path properties: statusPath, inputPath, resultPath and runLogPath are used in GET requests to retrieve the relevant data if it exists. If the data is not present, a 404 Not Found response is returned.
When an execution has not finished, the status field is set to NOT_COMPLETED. In this case, query the statusPath again after a short delay, repeating until it returns something other than NOT_COMPLETED.
In MPS/LP problems, when the solver has finished - and regardless of the actual final search status - OK is returned. In this case, check the
exitCode field of the
executionStatus structure to check if your problem was solved optimally. Valid values are:
- 0: Optimization not started
- 1: Locally optimal
- 2: Optimal
- 3: Locally infeasible
- 4: Infeasible
- 5: Unbounded
- 6: Unfinished
The Optimizer solve log is available in the execution's run log.