Struct Distribution 

pub struct Distribution { /* private fields */ }

Exponential histogram distribution for high-performance metric recording.

Cells are pre-allocated and indexed by thread_id & mask, so recording is lock-free with no per-call TLS bookkeeping.

Implementations

impl Distribution

pub fn new(shard_count: usize) -> Self

Create a new Distribution with shard_count shards.

The count is rounded up to the next power of two for fast modulo.

pub fn record(&self, value: u64)

Record a value.

pub fn count(&self) -> u64

Cumulative count across all shards.

pub fn sum(&self) -> u64

Cumulative sum across all shards.

pub fn sum_and_count(&self) -> (u64, u64)

Sum and count across all shards in a single pass.

Cheaper than calling Self::sum and Self::count separately: each shard’s ExpBuckets is visited once, and sum + count share a cache line per shard so the combined pass roughly halves the cache traffic of two separate scans.

pub fn min(&self) -> Option<u64>

Approximate minimum from bucket boundaries.

pub fn max(&self) -> Option<u64>

Approximate maximum from bucket boundaries.

pub fn mean(&self) -> Option<f64>

Mean value, or None if no values have been recorded.

pub fn buckets_snapshot(&self) -> ExpBucketsSnapshot

Merge bucket counts from all shards into a single snapshot.

Trait Implementations

impl ClickHouseExport for Distribution

fn export_clickhouse( &self, batch: &mut ClickHouseMetricBatch, name: &str, description: &str, time_unix_nano: u64, )

impl Debug for Distribution

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

Formats the value using the given formatter.

impl Default for Distribution

fn default() -> Self

Returns the “default value” for a type.

impl DistributionSnapshot for Distribution

fn count(&self) -> u64

fn sum(&self) -> u64

fn min(&self) -> Option<u64>

fn max(&self) -> Option<u64>

fn zero_count(&self) -> u64

fn visit_positive_buckets(&self, visitor: &mut dyn FnMut(i32, u64))

impl DogStatsDExport for Distribution

fn export_dogstatsd( &self, output: &mut String, name: &str, tags: &[(&str, &str)], )

Export distribution as |d samples for Datadog percentile computation.

impl OtlpExport for Distribution

fn export_otlp( &self, metrics: &mut Vec<Metric>, name: &str, description: &str, time_unix_nano: u64, )

Distribution exports as a native OTLP ExponentialHistogram (scale 0, base-2).

impl PrometheusExport for Distribution

fn export_prometheus(&self, output: &mut String, name: &str, help: &str)

Export distribution as summary (count + sum only, no quantiles).