Struct JournalEventQueue 

pub struct JournalEventQueue { /* private fields */ }

Thread-safe unified journal event queue

Buffers all event types in memory and provides FIFO access. When the buffer is full, only SSE-only telemetry can be evicted. Correctness events are rejected if no telemetry slot can be reclaimed.

Implementations

impl JournalEventQueue

pub fn new(config: JournalQueueConfig) -> Self

Create a new journal event queue with specified configuration

pub fn push(&self, event: JournalEventMessage) -> Result<(), String>

Push an event to the queue

If the queue is at capacity, the oldest SSE-only event is dropped. A correctness event is never evicted: if the queue contains only correctness events, a new correctness event is rejected and new telemetry is dropped.

pub fn pop(&self) -> Option<JournalEventMessage>

Pop the next event from the queue (FIFO)

pub fn push_front(&self, event: JournalEventMessage) -> Result<(), String>

Re-queue an event at the front (used when send fails)

pub fn drain_batch(&self, max: usize) -> Vec<JournalEventMessage>

Drain up to N events from the queue for batch sending

pub fn peek_batch_run_ids(&self, max: usize) -> Vec<String>

Return the distinct run IDs represented by the next max live events. The background sender uses this to acquire ordering barriers before it removes those events from the queue.

pub fn hold_run(&self, run_id: &str)

Keep run_id’s queued events out of the periodic flush until release_run or an explicit per-run drain.

pub fn release_run(&self, run_id: &str) -> bool

Return the run to normal periodic flushing. Returns whether it was held.

pub fn is_held(&self, run_id: &str) -> bool

pub fn drain_batch_for_runs( &self, max: usize, allowed_run_ids: &HashSet<String>, ) -> Vec<JournalEventMessage>

Drain up to max events belonging to runs whose ordering barriers are already held. Events from other runs retain their relative FIFO order.

pub fn drain_all(&self) -> Vec<JournalEventMessage>

Drain all events from the queue

pub fn drain_run_events(&self, run_id: &str) -> Vec<JournalEventMessage>

Drain all events for a specific run_id from the queue. Events for other runs remain in the queue (order preserved).

pub fn len(&self) -> usize

Get current queue length

pub fn is_empty(&self) -> bool

Check if queue is empty

pub fn contains_run(&self, run_id: &str) -> bool

Whether this queue still contains an event for run_id.

Pull completion uses this while holding the run’s flush barrier: an empty result then means no flush for that run is queued or in flight.

pub fn record_sent(&self, is_sse_only: bool)

Record a successful event send (for metrics)

pub fn record_sent_batch(&self, count: usize, sse_only_count: usize)

Record batch of successful sends

pub fn record_error(&self)

Record an event send error (for metrics)

pub fn metrics(&self) -> JournalQueueMetrics

Get current metrics snapshot

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

Get current metrics as tuple (queued, sent, dropped, errors)

pub fn oldest_age(&self) -> Option<Duration>

Get age of oldest event in queue

pub fn batch_size(&self) -> usize

Get the configured batch size

pub fn flush_interval_ms(&self) -> u64

Get the configured flush interval in milliseconds

Trait Implementations

impl Clone for JournalEventQueue

fn clone(&self) -> Self

Returns a duplicate of the value.

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

Performs copy-assignment from source.