Struct ClientMetricsSnapshot 

pub struct ClientMetricsSnapshot {

    pub requests_started: u64,
    pub requests_succeeded: u64,
    pub requests_failed: u64,
    pub requests_timed_out: u64,
    pub requests_cancelled: u64,
    pub broker_errors: u64,
    pub retries: u64,
    pub buffered_records: u64,
    pub max_buffered_records: u64,
    pub produced_records: u64,
    pub produce_batches: u64,
    pub consumed_records: u64,
    pub request_bytes: u64,
    pub response_bytes: u64,
    pub in_flight_requests: u64,
    pub total_latency: Duration,
    pub max_latency: Duration,
    pub request_latency_buckets: [u64; 13],
}

Point-in-time values from ClientMetrics.

Fields

requests_started: u64

Request roundtrips that have started.

requests_succeeded: u64

Request roundtrips that completed successfully.

requests_failed: u64

Request roundtrips that returned an error, including timeouts.

requests_timed_out: u64

Failed request roundtrips caused by the configured request timeout.

requests_cancelled: u64

Request futures dropped before producing a result.

broker_errors: u64

Non-zero Kafka error codes observed in decoded broker responses.

retries: u64

Additional high-level operation attempts after an initial attempt.

buffered_records: u64

Buffered producer records accepted and not yet completed.

max_buffered_records: u64

Highest observed number of outstanding buffered producer records.

produced_records: u64

Records successfully acknowledged by Produce operations.

produce_batches: u64

Successful topic-partition Produce chunks.

consumed_records: u64

Records returned to callers by consumer poll and fetch operations.

request_bytes: u64

Kafka request payload bytes, excluding the four-byte frame length.

response_bytes: u64

Kafka response payload bytes, excluding the four-byte frame length.

in_flight_requests: u64

Request roundtrips currently awaiting completion.

total_latency: Duration

Sum of completed and cancelled request latency.

max_latency: Duration

Highest observed completed or cancelled request latency.

request_latency_buckets: [u64; 13]

Approximate request latency histogram, using the documented upper-bound buckets from Self::latency_percentile.

Implementations

impl ClientMetricsSnapshot

pub fn latency_percentile(&self, percentile: u8) -> Option<Duration>

Returns an upper-bound latency estimate for a percentile in the range 0..=100.

The result is approximate because the client records counts in fixed upper-bound buckets instead of retaining every request duration. It returns None when no request has completed or when percentile is greater than 100.

Trait Implementations

impl Clone for ClientMetricsSnapshot

fn clone(&self) -> ClientMetricsSnapshot

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for ClientMetricsSnapshot

impl Debug for ClientMetricsSnapshot

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

Formats the value using the given formatter.

impl Default for ClientMetricsSnapshot

fn default() -> ClientMetricsSnapshot

Returns the “default value” for a type.

impl Eq for ClientMetricsSnapshot

impl PartialEq for ClientMetricsSnapshot

fn eq(&self, other: &ClientMetricsSnapshot) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl StructuralPartialEq for ClientMetricsSnapshot