Metrics API

GET /metrics

Returns Prometheus text format metrics from the gateway process.

curl http://localhost:8080/metrics

The endpoint is implemented by crates/nexis-gateway and is registered beside GET /health.

Registered Metric Families

Family	Purpose
nexis_build_info	Gateway package version and commit label
nexis_http_requests_total	HTTP request count by method and path
nexis_http_responses_total	HTTP response count by method, path, and status
nexis_http_latency_seconds	HTTP request latency histogram
nexis_connections_active	Active WebSocket connections
nexis_connections_total	Total WebSocket connections established
nexis_connection_errors	Connection errors by type
nexis_pool_connections_active	Active sharded-pool connections
nexis_pool_connections_peak	Peak sharded-pool connection count since start
nexis_pool_connections_total	Total sharded-pool connections created
nexis_pool_messages_sent_total	Broadcast messages sent through the pool
nexis_pool_messages_dropped_total	Broadcast messages dropped by the pool
nexis_messages_received_total	Messages received
nexis_messages_sent_total	Messages sent
nexis_messages_by_type	Messages grouped by type
nexis_message_latency_seconds	Message processing latency histogram
nexis_message_size_bytes	Message size histogram
nexis_rooms_active	Active rooms
nexis_rooms_created_total	Rooms created since start
nexis_room_members	Member count by room
nexis_operation_throughput_total	Domain operation throughput
nexis_operation_errors_total	Domain operation errors
nexis_operation_latency_seconds	Domain operation latency histogram
nexis_ai_requests_total	AI provider request count
nexis_ai_errors_total	AI provider errors
nexis_ai_latency_seconds	AI provider latency histogram
nexis_ai_tokens_total	AI provider token usage

Dashboards and Alerts

Use the operations guide for Prometheus, Grafana, and alert rule setup:

• Monitoring
• Troubleshooting