Struct CommonArgs 

pub struct CommonArgs {

    pub mooncake_trace_path: Option<String>,
    pub test: bool,
    pub num_gpu_blocks: usize,
    pub block_size: u32,
    pub trace_simulation_duration_ms: Option<u64>,
    pub benchmark_duration_ms: u64,
    pub num_unique_inference_workers: usize,
    pub inference_worker_duplication_factor: usize,
    pub trace_length_factor: usize,
    pub trace_duplication_factor: usize,
    pub seed: u64,
    pub sweep: bool,
    pub sweep_min_ms: u64,
    pub sweep_max_ms: u64,
    pub sweep_steps: usize,
    pub bench: bool,
    pub sequence_logs: bool,
}

Shared CLI arguments for trace-based benchmarks.

Fields

mooncake_trace_path: Option<String>

Path to a JSONL mooncake trace file.

test: bool

Deprecated compatibility flag. Use cargo test --package dynamo-bench --test ... for the fixture-backed integration tests instead.

num_gpu_blocks: usize

Number of GPU blocks available in the mock engine’s KV cache.

block_size: u32

Number of tokens per KV cache block.

trace_simulation_duration_ms: Option<u64>

Optional wall-clock duration (ms) used to rescale the trace during event generation. Omit to preserve the original Mooncake timestamps.

benchmark_duration_ms: u64

Wall-clock duration (ms) over which the benchmark replays operations.

num_unique_inference_workers: usize

Number of unique simulated inference workers.

inference_worker_duplication_factor: usize

How many times to duplicate unique workers during the benchmark phase.

trace_length_factor: usize

Factor by which to stretch each request’s hash sequence length.

trace_duplication_factor: usize

How many times to duplicate the raw trace data with offset hash_ids.

seed: u64

RNG seed for reproducible worker-to-trace assignment.

sweep: bool

Enable throughput vs p99 latency sweep mode.

sweep_min_ms: u64

Minimum benchmark duration (ms) for sweep mode.

sweep_max_ms: u64

Maximum benchmark duration (ms) for sweep mode.

sweep_steps: usize

Number of logarithmically spaced sweep steps between min and max.

bench: bool

Ignored - passed by cargo bench harness.

sequence_logs: bool

Opt in to runtime warn/error logs from the mocker and sequence tracker.

Trait Implementations

impl Args for CommonArgs

fn group_id() -> Option<Id>

Report the ArgGroup::id for this set of arguments

fn augment_args<'b>(__clap_app: Command) -> Command

Append to Command so it can instantiate Self via FromArgMatches::from_arg_matches_mut

fn augment_args_for_update<'b>(__clap_app: Command) -> Command

Append to Command so it can instantiate self via FromArgMatches::update_from_arg_matches_mut

impl Debug for CommonArgs

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl FromArgMatches for CommonArgs

fn from_arg_matches(__clap_arg_matches: &ArgMatches) -> Result<Self, Error>

Instantiate Self from ArgMatches, parsing the arguments as needed.

fn from_arg_matches_mut( __clap_arg_matches: &mut ArgMatches, ) -> Result<Self, Error>

Instantiate Self from ArgMatches, parsing the arguments as needed.

fn update_from_arg_matches( &mut self, __clap_arg_matches: &ArgMatches, ) -> Result<(), Error>

Assign values from ArgMatches to self.

fn update_from_arg_matches_mut( &mut self, __clap_arg_matches: &mut ArgMatches, ) -> Result<(), Error>

Assign values from ArgMatches to self.