Observability

Configurations in Observability.

Configurations

• Access Log Fields — Logs only server errors with a lean field set
• Access Logging — Structured JSON logging with sampling; logs 10% of requests. request_id ensures each log line has a correlation ID. access_log emits method, path, status, and timing
• Cloud Events — Publishes a bounded, best-effort CloudEvents 1.0 response event after the upstream response completes
• Errors Total — Prometheus counter for proxy errors classified by cause, covering filter rejections, timeouts, unreachable upstreams and internal faults
• Grpc Access Log — Logs how each gRPC call ended
• Grpc Health Check — Probes upstream endpoints with grpc.health.v1.Health/Check instead of an HTTP GET
• Http Active Requests — Prometheus gauge for HTTP requests currently in flight per listener
• Logging — request_id — ensures every request has a correlation ID
• Metric Label Sets — Selectively disables individual label dimensions on Prometheus metrics to bound total time-series cardinality
• Process Logging — Non-blocking process logs written to a file
• Route Templates — Collapses dynamic path segments in the Prometheus route label into stable templates, so high-variety URLs share one series instead of widening cardinality
• Tcp Access Log — Structured JSON logging of TCP connection events (connect and disconnect)
• Tcp Active Connections — Prometheus gauge for TCP connections currently open per listener
• Tcp Byte Counters — Prometheus counters for bytes forwarded over TCP connections in each direction, labeled by listener
• Tcp Connection Metrics — Prometheus histogram for TCP connection duration
• Tcp Connections Total — Prometheus counter for total accepted TCP connections per listener
• Trace Context — W3C Trace Context header propagation
• Tracing Otlp — Exports distributed tracing spans to an OpenTelemetry Collector via OTLP/gRPC
• Upstream Requests Total — Prometheus counter for requests that reached an upstream endpoint, labeled by cluster, endpoint and status class