Struct TraceQueries 

pub struct TraceQueries { /* private fields */ }

High-performance query patterns for Delta Lake trace storage.

Time predicates are always applied FIRST to enable Delta Lake partition pruning. span_cache provides sub-millisecond repeat reads for trace detail clicks. metrics_cache provides sub-millisecond repeat reads for dashboard metric charts.

Implementations

impl TraceQueries

pub fn new(ctx: Arc<SessionContext>) -> Self

pub async fn get_trace_spans( &self, trace_id_bytes: Option<&[u8]>, service_name: Option<&str>, start_time: Option<&DateTime<Utc>>, end_time: Option<&DateTime<Utc>>, limit: Option<usize>, ) -> Result<Vec<TraceSpan>, TraceEngineError>

Get all spans for a trace, reconstructed as a tree with hierarchy fields populated.

Arguments

• trace_id_bytes - Raw 16-byte trace ID
• service_name - Optional service filter
• start_time - Optional lower time bound (applied FIRST for partition pruning)
• end_time - Optional upper time bound
• limit - Optional row limit

When trace_id_bytes is 16 bytes, results are cached for 5 minutes — repeat detail clicks (common in the UI) return in <1µs without hitting Delta Lake.

pub async fn query_spans( &self, trace_id_bytes: Option<&[u8]>, service_name: Option<&str>, start_time: Option<&DateTime<Utc>>, end_time: Option<&DateTime<Utc>>, limit: Option<usize>, ) -> Result<Vec<TraceSpan>, TraceEngineError>

Execute the actual DataFusion query without cache logic.

pub async fn get_trace_metrics( &self, service_name: Option<&str>, start_time: DateTime<Utc>, end_time: DateTime<Utc>, bucket_interval: &str, attribute_filters: Option<&[String]>, entity_uid: Option<&str>, ) -> Result<Vec<TraceMetricBucket>, TraceEngineError>

Get trace metrics over a time range, bucketed by the given interval string.

bucket_interval must be a valid DataFusion DATE_TRUNC precision unit: "second", "minute", "hour", "day", "week", "month", "year".

Matches Postgres logic: trace duration = MAX(end_time) - MIN(start_time) across all spans of a trace (not per-span duration_ms). Root service is the service of the span where parent_span_id IS NULL. Service filter applies to root spans only.

attribute_filters is a list of "key:value" strings OR-matched against search_blob. entity_trace_ids is an optional pre-resolved list of binary trace IDs (16 bytes each).

pub async fn query_spans_from_trace_filters( &self, filters: &TraceFilters, ) -> Result<Vec<TraceSpan>, TraceEngineError>

Look up traces from the summary table that match aggregate-level filters (e.g. entity_uid, queue_uid, has_errors) and return spans for the most-recent matching trace. The entire pipeline runs as a single DataFusion JOIN — no intermediate collection, no Postgres round-trip.