Command-line options¶
To run a test, use the molotov runner and point it to the scenario module or path:
Load test.
usage: molotov [-h] [--single-run] [-s SINGLE_MODE] [--config CONFIG]
[--version] [--debug] [-v] [-w WORKERS] [--ramp-up RAMP_UP]
[--sizing] [--sizing-tolerance SIZING_TOLERANCE]
[--delay DELAY] [--console-update CONSOLE_UPDATE]
[-p PROCESSES] [-d DURATION] [-r MAX_RUNS] [-q] [-x] [-f FAIL]
[-c] [--statsd] [--statsd-address STATSD_ADDRESS] [--uvloop]
[--use-extension USE_EXTENSION [USE_EXTENSION ...]]
[--force-shutdown]
[scenario]
Positional Arguments¶
scenario | path or module name that contains scenarii Default: “loadtest.py” |
Named Arguments¶
--single-run | Run once every existing scenario Default: False |
-s, --single-mode | |
Name of a single scenario to run once. | |
--config | Point to a JSON config file. |
--version | Displays version and exits. Default: False |
--debug | Run the event loop in debug mode. Default: False |
-v, --verbose | Verbosity level. -v will display tracebacks. -vv requests and responses. Default: 0 |
-w, --workers | Number of workers Default: 1 |
--ramp-up | Ramp-up time in seconds Default: 0.0 |
--sizing | Autosizing Default: False |
--sizing-tolerance | |
Sizing tolerance Default: 5.0 | |
--delay | Delay between each worker run Default: 0.0 |
--console-update | |
Delay between each console update Default: 0.2 | |
-p, --processes | |
Number of processes Default: 1 | |
-d, --duration | Duration in seconds Default: 86400 |
-r, --max-runs | Maximum runs per worker |
-q, --quiet | Quiet Default: False |
-x, --exception | |
Stop on first failure. Default: False | |
-f, --fail | Number of failures required to fail |
-c, --console | Use simple console for feedback Default: False |
--statsd | Activates statsd Default: False |
--statsd-address | |
Statsd Address Default: “udp://localhost:8125” | |
--uvloop | Use uvloop Default: False |
--use-extension | |
Imports a module containing Molotov extensions | |
--force-shutdown | |
Cancel all pending workers on shutdown Default: False |