fix(rustproxy): prune stale per-route metrics, add per-route rate limiter caching and regex cache, and improve connection tracking cleanup to prevent memory growth

rust/crates/rustproxy-metrics/src/collector.rs

@@ -1,5 +1,6 @@
 use dashmap::DashMap;
 use serde::{Deserialize, Serialize};
+use std::collections::HashSet;
 use std::sync::atomic::{AtomicU64, Ordering};
 use std::sync::Mutex;
 
@@ -196,6 +197,12 @@ impl MetricsCollector {
                 if val <= 1 {
                     drop(counter);
                     self.ip_connections.remove(ip);
+                    // Evict all per-IP tracking data for this IP
+                    self.ip_total_connections.remove(ip);
+                    self.ip_bytes_in.remove(ip);
+                    self.ip_bytes_out.remove(ip);
+                    self.ip_pending_tp.remove(ip);
+                    self.ip_throughput.remove(ip);
                 }
             }
         }
@@ -342,6 +349,17 @@ impl MetricsCollector {
         }
     }
 
+    /// Remove per-route metrics for route IDs that are no longer active.
+    /// Call this after `update_routes()` to prune stale entries.
+    pub fn retain_routes(&self, active_route_ids: &HashSet<String>) {
+        self.route_connections.retain(|k, _| active_route_ids.contains(k));
+        self.route_total_connections.retain(|k, _| active_route_ids.contains(k));
+        self.route_bytes_in.retain(|k, _| active_route_ids.contains(k));
+        self.route_bytes_out.retain(|k, _| active_route_ids.contains(k));
+        self.route_pending_tp.retain(|k, _| active_route_ids.contains(k));
+        self.route_throughput.retain(|k, _| active_route_ids.contains(k));
+    }
+
     /// Get current active connection count.
     pub fn active_connections(&self) -> u64 {
         self.active_connections.load(Ordering::Relaxed)
@@ -633,6 +651,42 @@ mod tests {
         assert!(collector.ip_connections.get("1.2.3.4").is_none());
     }
 
+    #[test]
+    fn test_per_ip_full_eviction_on_last_close() {
+        let collector = MetricsCollector::with_retention(60);
+
+        // Open connections from two IPs
+        collector.connection_opened(Some("route-a"), Some("10.0.0.1"));
+        collector.connection_opened(Some("route-a"), Some("10.0.0.1"));
+        collector.connection_opened(Some("route-b"), Some("10.0.0.2"));
+
+        // Record bytes to populate per-IP DashMaps
+        collector.record_bytes(100, 200, Some("route-a"), Some("10.0.0.1"));
+        collector.record_bytes(300, 400, Some("route-b"), Some("10.0.0.2"));
+        collector.sample_all();
+
+        // Verify per-IP data exists
+        assert!(collector.ip_total_connections.get("10.0.0.1").is_some());
+        assert!(collector.ip_bytes_in.get("10.0.0.1").is_some());
+        assert!(collector.ip_throughput.get("10.0.0.1").is_some());
+
+        // Close all connections for 10.0.0.1
+        collector.connection_closed(Some("route-a"), Some("10.0.0.1"));
+        collector.connection_closed(Some("route-a"), Some("10.0.0.1"));
+
+        // All per-IP data for 10.0.0.1 should be evicted
+        assert!(collector.ip_connections.get("10.0.0.1").is_none());
+        assert!(collector.ip_total_connections.get("10.0.0.1").is_none());
+        assert!(collector.ip_bytes_in.get("10.0.0.1").is_none());
+        assert!(collector.ip_bytes_out.get("10.0.0.1").is_none());
+        assert!(collector.ip_pending_tp.get("10.0.0.1").is_none());
+        assert!(collector.ip_throughput.get("10.0.0.1").is_none());
+
+        // 10.0.0.2 should still have data
+        assert!(collector.ip_connections.get("10.0.0.2").is_some());
+        assert!(collector.ip_total_connections.get("10.0.0.2").is_some());
+    }
+
     #[test]
     fn test_http_request_tracking() {
         let collector = MetricsCollector::with_retention(60);
@@ -650,6 +704,35 @@ mod tests {
         assert_eq!(snapshot.http_requests_per_sec, 3);
     }
 
+    #[test]
+    fn test_retain_routes_prunes_stale() {
+        let collector = MetricsCollector::with_retention(60);
+
+        // Create metrics for 3 routes
+        collector.connection_opened(Some("route-a"), None);
+        collector.connection_opened(Some("route-b"), None);
+        collector.connection_opened(Some("route-c"), None);
+        collector.record_bytes(100, 200, Some("route-a"), None);
+        collector.record_bytes(100, 200, Some("route-b"), None);
+        collector.record_bytes(100, 200, Some("route-c"), None);
+        collector.sample_all();
+
+        // Now "route-b" is removed from config
+        let active = HashSet::from(["route-a".to_string(), "route-c".to_string()]);
+        collector.retain_routes(&active);
+
+        // route-b entries should be gone
+        assert!(collector.route_connections.get("route-b").is_none());
+        assert!(collector.route_total_connections.get("route-b").is_none());
+        assert!(collector.route_bytes_in.get("route-b").is_none());
+        assert!(collector.route_bytes_out.get("route-b").is_none());
+        assert!(collector.route_throughput.get("route-b").is_none());
+
+        // route-a and route-c should still exist
+        assert!(collector.route_total_connections.get("route-a").is_some());
+        assert!(collector.route_total_connections.get("route-c").is_some());
+    }
+
     #[test]
     fn test_throughput_history_in_snapshot() {
         let collector = MetricsCollector::with_retention(60);