PorticolPerf
Usage
porticol-cli perf <collection-file|folder> [options]
porticol-cli perf --server <url> --collection <id|name> [options]
Perf options
| Option | What it does |
|---|---|
-u, --users <n> |
virtual users, 1 to 200 (default 10) |
--duration <time> |
how long the load lasts, as 1500ms, 90s, 5m or 1h, from 1s to 1h (default 60s) |
--ramp-up <time> |
the users start spread across this (default 0s) |
--threshold <measure=n> |
the run fails unless this holds: p50, p90, p95, p99, max and mean in ms, and error_rate in percent, at most n; requests_per_second at least n. Repeatable |
--timeout <ms> |
as for run |
-d, --data <file> |
a CSV or JSON dataset, read round-robin by the users |
--max-time <time> |
stop the load after this, as 30s or 10m |
--retry <n> |
try a request the transport could not send n more times |
Sources
| Option | What it does |
|---|---|
--server <url> |
a Porticol Cloud or Team Server, or PORTICOL_SERVER |
--token <token> |
an API token for it, or PORTICOL_TOKEN, or the file named by PORTICOL_TOKEN_FILE, or the one login kept |
--collection <id|name> |
the server's collection to run |
--folder <name|id> |
run this folder's requests alone |
-e, --environment <file|name> |
an environment file, or the name or id of one of the server's environments; file: and env: say which |
--remote |
run on the server rather than here (run only) |
Secrets
| Option | What it does |
|---|---|
--secret <name=value> |
a secret variable, repeatable; above the environment file and the server's values, masked in every report |
--secret <name> |
the same, with the value read from the variable name |
--env-var <name> |
the same: the value is the process variable of that name |
--env-file <file> |
names to read from a .env file, one NAME=VALUE a line |
Project
porticol.toml beside the collection names environment, data, timeout and reports; a flag always wins over it
Reports
| Option | What it does |
|---|---|
-r, --reporter <name[:path]> |
cli, junit, json or html, to standard output or to the file at path; repeat for several. perf reports as cli, json or html |
--junit <path> |
the same as -r junit:<path> |
--json <path> |
the same as -r json:<path> |
--html <path> |
the same as -r html:<path> |
Output
| Option | What it does |
|---|---|
--color <auto|always|never> |
colour in the report; auto is a terminal, or CLICOLOR_FORCE in the environment |
--no-color |
the same as --color never |
-q, --quiet |
the totals alone: no header, no result lines, and no notices |
--silent |
nothing on standard output; file reports still written |
--no-progress |
the results at the end rather than as they finish |
--verbose |
every assertion and every console line |
Exit codes
| Code | Meaning |
|---|---|
| 0 | every request, assertion and threshold passed |
| 1 | an assertion, a schema check or a threshold failed, a flow stopped, or the run was stopped with Ctrl-C |
| 2 | a request could not be made |
| 3 | the command line, a file, the configuration, the token or a variable is wrong |
| 4 | the licence refused the run |
| 5 | the head document breaks the base document, at the level --fail-on names |
| 6 | the document has a lint finding of severity error |