CLI reference

ase fleet

Dedicated servers: their performance, and the local provider.

ase fleet local

Run the local provider: machines are brain-agent processes on this computer.

Requires: a sign-in (ase login) or, in CI, a release credential. An ase-brain checkout, and brain-agent beside ase or given with --agent.

Changes: Runs a local provider on 127.0.0.1 and starts brain-agent processes on this computer, which keep running after it exits; its temporary brain-local-* folder stays. Nothing in the cloud changes.

ase fleet local [OPTIONS]

Argument	Description
`--port <PORT>`	The port the local provider listens on. Default `8799`.
`--agent <AGENT>`	The agent binary; `brain-agent` next to `ase` by default.

Examples

ase fleet local --port 8799

ase fleet loadtest

A load test on Fleet with synthetic players, or its stored reports (list, show, compare). A run prints its cost estimate and spends only with --yes.

Requires: a sign-in (ase login) with the studio permission run_load_test (Owner and Admin). Without --yes it only prints the cost estimate, for any role.

Potentially destructive. With --yes it spends the studio's Fleet money: it mints a synthetic Play player per bot, each for at most 2 hours, runs a real lobby and its bots on Fleet, writes the report into --out and stores it with the game, where the studio sees it. Without --yes nothing runs. The money spent can't be returned.

ase fleet loadtest [OPTIONS] [COMMAND]

Argument	Description
`--game <GAME>`	The game's id.
`--version <VERSION>`	The release whose Linux server and client run.
`--region <REGION>`	The game's region or hosting pool the lobby runs in.
`--players <PLAYERS>`	Synthetic players at the top of the ramp. Default `32`.
`--ramp <RAMP>`	Bots added per step. Default `4`.
`--step <STEP>`	Time between ramp steps, e.g. `10s`. Default `10s`.
`--warmup <WARMUP>`	Time at full load before measuring. Default `10s`.
`--measure <MEASURE>`	How long the run is measured. Default `60s`.
`--mode <MODE>`	The lobby's mode, as the game asks for it.
`--client-arg <CLIENT_ARGS>...`	The game's own flags for its bots, after the client's `bot` launch profile (e.g. `--client-arg=--autostart=32`); repeat for several.
`--yes`	Spend: without it nothing runs.
`--out <OUT>`	Where the report, samples and server log go; `<run id>/` by default.
`--json`	Print JSON.

Examples

ase fleet loadtest --game neon-drift --version 1.4.0 --region eu --players 64
ase fleet loadtest --game neon-drift --version 1.4.0 --region eu --players 64 --yes

ase fleet loadtest list

The game's stored load tests, newest first.

Requires: a sign-in (ase login), in any studio role.

Changes: nothing; it only reads.

ase fleet loadtest list [OPTIONS]

Argument	Description
`--game <GAME>`	The game's id.
`--json`	Print JSON.

Examples

ase fleet loadtest list --game neon-drift

ase fleet loadtest show

One run's report.

Requires: a sign-in (ase login), in any studio role.

Changes: nothing; it only reads.

ase fleet loadtest show [OPTIONS] <RUN>

Argument	Description
`<RUN>`	The run's id, as `list` prints it.
`--game <GAME>`	The game's id.
`--json`	Print JSON.

Examples

ase fleet loadtest show run-2 --game neon-drift

ase fleet loadtest compare

Whether the second run regressed from the first: fails when it did.

Requires: a sign-in (ase login), in any studio role.

Changes: nothing; it only reads.

ase fleet loadtest compare [OPTIONS] <BEFORE> <AFTER>

Argument	Description
`<BEFORE>`	The run to compare with.
`<AFTER>`	The run that may have regressed.
`--game <GAME>`	The game's id.
`--json`	Print JSON.

Examples

ase fleet loadtest compare run-1 run-2 --game neon-drift

ase fleet perf

A game's server performance on Fleet: live with --follow (one JSON line per agent report, every second while followed), or its minute history.

Requires: a sign-in (ase login), in any studio role. History from more than 14 days back scans BigQuery, which counts against the studio's monthly scan allowance and is billed to it.

Changes: nothing; it only reads.

ase fleet perf [OPTIONS] --game <GAME>

Argument	Description
`--game <GAME>`	Required. The game's id.
`--machine <MACHINE>`	One machine.
`--lobby <LOBBY>`	One match: its lobby id.
`--follow`	Stream every report as it arrives.
`--for <DURATION>`	Stop following after this long, e.g. `10m`.
`--from <FROM>`	History from: a span back from now (`1h`, `14d`) or epoch milliseconds. Past 14 days the game's samples in `BigQuery` answer, grouped. Default `1h`.
`--to <TO>`	History to, as `--from` takes it; now by default.
`--by <BY>`	Group the matches' minutes: `version`, `plan` or `players`.
`--json`	Print Fleet's answer as JSON.

Examples

ase fleet perf --game neon-drift --follow --for 10m
ase fleet perf --game neon-drift --from 7d --by version