From 119fc489bc3074fd3bd081d97937ba23945fbcf5 Mon Sep 17 00:00:00 2001 From: Brian Ehlert Date: Wed, 25 Jan 2023 10:36:01 -0800 Subject: [PATCH] Clarify deep service insight (#3453) (cherry picked from commit 01d817ffea7c003165e9e421ad762de90245bcb0) --- .../logging-and-monitoring/service-insight.md | 18 ++++++++++-------- 1 file changed, 10 insertions(+), 8 deletions(-) diff --git a/docs/content/logging-and-monitoring/service-insight.md b/docs/content/logging-and-monitoring/service-insight.md index 9073e633b1..d75dbcb618 100644 --- a/docs/content/logging-and-monitoring/service-insight.md +++ b/docs/content/logging-and-monitoring/service-insight.md @@ -11,14 +11,16 @@ docs: "DOCS-000" --- -The Ingress Controller exposes an endpoint and provides host statistics for Virtual Servers (VS) and Transport Servers (TS). +The Ingress Controller exposes an endpoint and provides host statistics for services exposed using the VirtualServer (VS) and TransportServer (TS) resource objects. It exposes data in JSON format and returns HTTP status codes. The response body holds information about the total, down and the unhealthy number of -upstreams associated with the hostname. -Returned HTTP codes indicate the health of the upstreams (service). +upstream pods associated with the configured hostname. +Returned HTTP codes indicate the health of the service. -The service is not healthy (HTTP response code different than 200 OK) if all upstreams are unhealthy. -The service is healthy if at least one upstream is healthy. In this case, the endpoint returns HTTP code 200 OK. +The service is indicated as not healthy (HTTP response code different than 200 OK) if all upstreams (pods) are determined unhealthy by NGINX Plus. +The service is healthy if at least one upstream pod is healthy as determined by NGINX Plus. In this case, the endpoint returns HTTP code 200 OK. + +NGINX Plus determination of healthy can be tuned using advanced health checks, and also dynamically relate to pods responses and responsiveness. See Upstream Healthcheck https://docs.nginx.com/nginx-ingress-controller/configuration/virtualserver-and-virtualserverroute-resources/#upstream ## Enabling Service Insight Endpoint @@ -39,9 +41,9 @@ If you're using *Helm* to install the Ingress Controller, to enable Service Insi The Service Insight provides the following statistics: -* Total number of VS and TS -* Number of VS and TS in 'Up' state -* Number of VS and TS in 'Unhealthy' state +* Total number of VS or TS pods +* Number of VS or TS pods in 'Up' state +* Number of VS or TS pods in 'Unhealthy' state These statistics are returned as JSON: