Struct ServerConfig 

pub struct ServerConfig {

    pub host: String,
    pub port: u16,
    pub max_concurrent: usize,
    pub timeout_secs: u64,
    pub cors_enabled: bool,
    pub api_keys: Vec<String>,
    pub rate_limit_capacity: f64,
    pub rate_limit_rate: f64,
    pub body_limit_bytes: usize,
    pub metrics_enabled: bool,
    pub structured_tracing: bool,
    pub router_capacity: usize,
    pub router_mem_budget_mb: usize,
    pub router_preload: Vec<String>,
    pub admin_bearer_token: Option<String>,
    pub admin_listen: String,
    pub batch_spool_dir: Option<String>,
    pub batch_max_pending_bytes: usize,
    pub per_key_rate_limits: Option<HashMap<String, (f64, f64)>>,
}

Configuration for the OxiLLaMa API server.

Fields

host: String

Host address to bind to.

port: u16

Port number.

max_concurrent: usize

Maximum concurrent requests.

timeout_secs: u64

Request timeout in seconds.

cors_enabled: bool

Enable CORS headers.

api_keys: Vec<String>

API keys for authentication (empty = no auth).

rate_limit_capacity: f64

Rate limit: maximum burst capacity (0.0 = no limit).

rate_limit_rate: f64

Rate limit: tokens per second refill rate.

body_limit_bytes: usize

Maximum request body size in bytes (0 = no limit).

metrics_enabled: bool

Enable the /metrics Prometheus endpoint.

structured_tracing: bool

Enable structured request tracing middleware.

router_capacity: usize

Maximum number of concurrently loaded models (0 = 1, single-model mode).

router_mem_budget_mb: usize

Memory budget for the model pool in MiB (0 = unlimited).

router_preload: Vec<String>

Model IDs to pre-load at startup.

admin_bearer_token: Option<String>

Bearer token required for all /admin/* routes.

None = token-less mode (admin only accessible from loopback).

admin_listen: String

Address the admin interface is expected to listen on. Used for the startup safety check: non-loopback + no token → fatal error.

batch_spool_dir: Option<String>

Directory for disk-spooled batch jobs. Defaults to $TMPDIR/oxillama_batch_spool.

batch_max_pending_bytes: usize

Maximum pending bytes across all queued batch jobs.

per_key_rate_limits: Option<HashMap<String, (f64, f64)>>

Per-key override map: api_key → (capacity, rate_per_second).

When a request carries an API key that appears in this map, the override (capacity, rate) pair is used instead of the server defaults. Keys absent from this map use rate_limit_capacity and rate_limit_rate as their bucket parameters.

None (the default) disables per-key rate limiting entirely.

Trait Implementations

impl Clone for ServerConfig

fn clone(&self) -> ServerConfig

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for ServerConfig

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

Formats the value using the given formatter.

impl Default for ServerConfig

fn default() -> Self

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for ServerConfig

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for ServerConfig

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.