Struct Config 

pub struct Config {

    pub storage: StorageConfig,
    pub dimensions: u16,
    pub distance_metric: DistanceMetric,
    pub flush_interval: Duration,
    pub split_threshold_vectors: usize,
    pub merge_threshold_vectors: usize,
    pub split_search_neighbourhood: usize,
    pub max_pending_and_running_rebalance_tasks: usize,
    pub rebalance_backpressure_resume_threshold: usize,
    pub max_rebalance_tasks: usize,
    pub chunk_target: u16,
    pub query_pruning_factor: Option<f32>,
    pub metadata_fields: Vec<MetadataFieldSpec>,
}

Configuration for a VectorDb instance.

Fields

storage: StorageConfig

Storage backend configuration.

Determines where and how vector data is persisted. See StorageConfig for available options including in-memory and SlateDB backends.

dimensions: u16

Vector dimensionality (immutable after creation).

All vectors written to this database must have exactly this many f32 values. Common dimensions: 384 (MiniLM), 768 (BERT), 1536 (OpenAI).

distance_metric: DistanceMetric

Distance metric for similarity computation (immutable after creation).

flush_interval: Duration

How often to flush data to durable storage (in seconds).

split_threshold_vectors: usize

Number of vectors in a centroid’s posting list that triggers a split.

merge_threshold_vectors: usize

Number of vectors below which a centroid’s posting list triggers a merge.

split_search_neighbourhood: usize

Number of neighboring centroids to scan for reassignment candidates after a split.

max_pending_and_running_rebalance_tasks: usize

The maximum number of centroids that require rebalancing before which backpressure is applied by pausing ingestion of new vector writes.

rebalance_backpressure_resume_threshold: usize

After backpressure is applied, ingestion resumes after the total number of centroids requiring rebalance drops below this value.

max_rebalance_tasks: usize

The maximum number of rebalance tasks that the rebalancer will run concurrently.

chunk_target: u16

Target number of centroids per chunk.

query_pruning_factor: Option<f32>

Query-aware dynamic pruning epsilon (ε₂ from SPANN paper).

When set, a posting list is searched only if its centroid’s distance to the query satisfies dist(q, c) <= (1 + epsilon) * dist(q, closest). This reduces query latency by skipping distant posting lists while preserving recall.

Typical values: 0.1 to 0.5. None disables pruning (all nprobe posting lists are searched).

metadata_fields: Vec<MetadataFieldSpec>

Metadata field schema.

Defines the expected attribute names and types. Writes with unknown attribute names or type mismatches will fail. If empty, any attribute names are accepted with types inferred from the first write.

Trait Implementations

impl Clone for Config

fn clone(&self) -> Config

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for Config

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

Formats the value using the given formatter.

impl Default for Config

fn default() -> Self

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for Config

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for Config

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

where __S: Serializer,

Serialize this value into the given Serde serializer.