Technologies/Envoy/envoy.http.downstream.request.by_status

EnvoyMetric

envoy.http.downstream.request.by_status

Requests by status code class

Dimensions:None

Available on:

Prometheus (1)Native (1)

Interface Metrics (2)

Prometheus

envoy_http_downstream_rq_xx

Total number of HTTP requests by response code class (2xx, 3xx, 4xx, 5xx)

Dimensions:None

Native

http.downstream_rq_xx

Total number of HTTP responses by status code class (1xx, 2xx, 3xx, 4xx, 5xx)

Dimensions:None

Sources

envoy_http_downstream_rq_xxwww.envoyproxy.io

http.downstream_rq_xxwww.envoyproxy.io

Technical Annotations (44)

Configuration Parameters (5)

envoy.reloadable_features.http2_use_oghttp2recommended: false

Workaround to disable oghttp2 and prevent connection hangs in v1.29

common_http_protocol_options.max_requests_per_connectionrecommended: 9500

Set below upstream connection limit to avoid GOAWAY race condition

spec.rules.backendRefs.name

Must reference an existing Kubernetes Service name in the same or allowed namespace

spec.rules.backendRefs.port

Must match an exposed port on the referenced backend service

retry.perRetry.backOff.baseIntervalrecommended: 1s or greater

0s causes proto validation failure and complete route rejection

Error Signatures (15)

5xxhttp status

503http status

504http status

status=~"5.."http status

408http status

max_duration_timeoutlog pattern

HTTP/1.1 500 Internal Server Errorhttp status

BackendNotFounderror code

direct_responselog pattern

Service default/backend-does-not-exist not foundlog pattern

route_not_foundlog pattern

Proto constraint validation failedlog pattern

value must be greater than 0slog pattern

RouteConfigurationValidationErrorexception

RetryBackOffValidationError.BaseIntervalexception

CLI Commands (4)

sum by (endpoint) (rate(http_requests_total{status=~"5.."}[5m]))diagnostic

kubectl get httproute/backend -o yamldiagnostic

egctl x status all -Adiagnostic

kubectl logs -n envoy-gateway-system -l gateway.envoyproxy.io/owning-gateway-namespace=default,gateway.envoyproxy.io/owning-gateway-name=eg -c envoy | grep start_time | jq

diagnostic

Technical References (20)

http_requests_totalcomponentoghttp2componentstream_timeoutconfigurationHTTP/2protocolGOAWAY frameprotocollast_stream_idprotocolupstream_local_addressconceptAWS ALBcomponentHTTPRoutecomponentResolvedRefscomponentdirect_responseconceptbackendRefscomponentBackendTrafficPolicycomponentdynamic_route_configscomponentxDSprotocolRouteConfigurationcomponentenvoy-gatewaycomponentRequestHeaderModifiercomponentX-Ray service mapcomponentfault rateconcept

Related Insights (9)

High 5xx error rate indicates upstream instability or misconfigurationcritical

▸

503/504 errors disrupt SLAs due to no healthy upstreams or timeoutscritical

▸

HTTP 5xx error rate exceeds 1% threshold causing user-facing failurescritical

▸

HTTP/2 POST requests hang with 408 max_duration_timeout after v1.29 upgradecritical

▸

HTTP/2 streams ignored after GOAWAY not retried causing 503swarning

▸

HTTPRoute backend service not found causes HTTP 500 errorscritical

▸

BackendTrafficPolicy with baseInterval: 0s causes all routes to return 404critical

▸

Incorrect HTTPRoute filters cause all routes to return 404critical

▸

Service showing 3-4% fault rate in X-Ray service mapwarning

▸