Job response

Format

{
  "trace_id": "tr-123",
  "status": "success",
  "stdout": "42\n",
  "stderr": null,
  "exception": null,
  "executionTime": 34,
  "exit_code": 0
}

The executionTime value is illustrative.

Fields

Field	Type	Description
trace_id	string	the same as the job's
status	string	"success" when the guest ran the job; "failed" when it rejected it (unknown language, stdin too large) or could not prepare the directory
stdout	string	standard output
stderr	string or null	null when empty
exception	string or null	"Timeout", "Compilation failed" or a preparation error
executionTime	integer	ms measured inside the guest, from the start of compilation/execution to the end
exit_code	integer	the program's exit code; omitted when 0
error	string	rejection reason; omitted when empty

How to read the result

"status": "success" says the execution happened, not that the program succeeded.

Situation	status	exception	stderr	exit_code
program finished fine	success	null	what the program wrote, or null	0
program exited with an error	success	null	the error	≠ 0
compilation failure	success	"Compilation failed"	compiler output	—
timeout exceeded	success	"Timeout"	whatever there is	—

Go is the exception on the compilation row: it runs through go run, which compiles and executes in one step, so a compilation error arrives only in stderr, with exception null.

What the API does with the result

In /api/v1/run and /api/v1/execute, the API passes through status, stdout, stderr, exception, executionTime and error; exit_code does not appear. It appears as exitCode in the result event of /api/v1/execute/stream and in the machine exec response. The CLI shows it on the summary line of rc run and exits with rc 1 when it is non-zero; rc machine exec exits with the command's own code.

If the guest does not respond within the job's timeout plus a host grace period, the VM is killed and the API returns "status": "failed" with "error": "E001: operation timed out".