Module metrics 

Monitoring and metrics infrastructure for hybrid search.

This module provides comprehensive observability for the search system:

• Prometheus metrics for latency, throughput, and error tracking
• Structured logging with tracing
• Performance monitoring and alerting

Metrics Exposed

• maproom_search_query_latency_seconds - End-to-end search latency histogram
• maproom_search_fusion_time_seconds - Score fusion computation time histogram
• maproom_search_cache_hit_rate - Cache effectiveness gauge (0.0-1.0)
• maproom_search_result_count - Number of results returned per query histogram
• maproom_search_errors_total - Counter for errors by type
• maproom_search_queries_total - Total number of search queries counter

Alert Thresholds

• p95 latency > 50ms: Performance degradation warning
• p95 latency > 100ms: Critical performance issue
• Error rate > 1%: Reliability issue warning
• Error rate > 5%: Critical reliability issue
• Cache hit rate < 50%: Cache effectiveness warning

Usage

use maproom::metrics::{SearchMetrics, get_metrics};
use std::time::Instant;

#[tokio::main]
async fn main() {
    let metrics = get_metrics();

    let start = Instant::now();
    // ... perform search ...
    let duration = start.elapsed();

    metrics.record_query_latency(duration.as_secs_f64(), "code", true);
    metrics.record_result_count(10, "code");
    metrics.increment_queries("code", true);
}

Re-exports

pub use cache_metrics::CacheMetrics;

pub use performance::get_performance_metrics;

pub use performance::PerformanceMetrics;

pub use prometheus::init_metrics_server;

pub use prometheus::metrics_handler;

pub use search_metrics::get_metrics;

pub use search_metrics::get_registry;

pub use search_metrics::SearchMetrics;

Modules

cache_metrics

Cache metrics tracking for embedding deduplication.

performance

Performance metrics collection for indexing, memory, and throughput.

prometheus

Prometheus metrics endpoint and HTTP server.

search_metrics

Search metrics collection and tracking.