Nginx ingress monitoring dashboards and metrics - Enterprise Distributed Application Service

When you route traffic through NGINX Ingress controllers in Kubernetes, you need visibility into request volume, latency, error rates, and controller health to detect issues and optimize performance. Enterprise Distributed Application Service (EDAS) provides built-in monitoring for NGINX Ingress resources across five dashboards, so you can diagnose traffic problems without setting up a separate monitoring stack.

Monitoring data comes from two sources:

Application Real-Time Monitoring Service (ARMS) Prometheus Service data source: Provides infrastructure and request metrics by default.
Loki log data source: Provides gateway log analysis. To use this data source, install a later version of the Prometheus component.

Prerequisites

Before you begin, make sure that you have:

An EDAS-managed Kubernetes cluster with NGINX Ingress controllers deployed
Access to the EDAS console

Access NGINX Ingress monitoring

Log on to the EDAS console.
In the left-side navigation pane, choose Traffic Management > Application Routing.
On the Application Routing (Kubernetes Ingress) page, find the target Ingress and click Monitor in the Operation column.
On the Ingress monitoring page, select a region and click any of the five monitoring tabs.

To enable the latest monitoring version, go to the Overview tab, click Go to the opening, and then click Enable in the Latest Version of Monitoring Not Enabled dialog box.

To disable monitoring, click Disable Monitoring in the upper-right corner of the Ingress monitoring page.

Dashboard overview

Five dashboards cover different monitoring scenarios:

Dashboard	When to use
Infrastructure Monitoring	Check cluster-wide request volume, connection counts, HTTP status codes, latency percentiles, configuration reload status, and controller resource usage.
Overview	Inspect per-Ingress traffic breakdown: request counts, inbound/outbound traffic, status code distribution, and latency percentiles (P50, P95, P99, P9999).
TopN	Identify hotspots by ranking the top 10 services, Ingresses, regions, browsers, and devices by request volume, failure rate, latency, or traffic.
URL Analysis	Drill into per-URL metrics: QPS, P95 latency, failure rate, and error details. Filter by cluster, namespace, Ingress, and service.
Log Analysis	Inspect individual request logs with full field details, including trace IDs for end-to-end troubleshooting in the ARMS console.

Two additional tabs, NGINX Monitoring (Community Edition) and Request Processing Performance (Community Edition), display community-version dashboards. The data on these tabs is for reference only.

Infrastructure Monitoring

The Infrastructure Monitoring tab integrates with the ARMS Prometheus Service basic monitoring panel. It provides cluster-wide metrics across all NGINX Ingress controllers, helping you view overall request statistics and identify issues through URL Analysis and Log Analysis. Metrics on controller load and configuration reloads help with resource management and anomalous configuration detection.

Requests

Metric	Description
Total number of requests	Overall QPS across all controllers of the NGINX Ingress controller service.
Number of requests	Requests processed by the selected Ingresses in the last 2 minutes.
Number of requests (by Ingress)	Line chart of HTTP request count per Ingress.
ReqPS(by Ingress)	Line chart of requests per second per Ingress.
QPS	Line chart of requests per second per pod of the NGINX Ingress controller.
ReqPS (by Path)	Line chart of failed requests per second per path. A request is considered failed when the HTTP status code is 4xx or 5xx.

Connections

Metric	Description
Number of connections	Line chart of pod connections based on an NGINX Ingress controller.
Number of new connections	New connections to pods of the selected NGINX Ingress controller service.
Total connections	Total connections to pods of the selected NGINX Ingress controller service.

Status

Metric	Description
HTTP 1/2xx	Count of HTTP 1xx or 2xx status codes in the selected time period.
HTTP 3xx	Count of HTTP 3xx status codes in the selected time period.
HTTP 4xx	Count of HTTP 4xx status codes in the selected time period.
HTTP 5xx	Count of HTTP 5xx status codes in the selected time period.
Success rate	Percentage of successful requests processed by the selected Ingresses.
HTTP status code	Line chart of HTTP status codes over time.
Request failure rate	Line chart of the ratio of 404 or 5xx status codes.
Error rate (by Path)	Line chart of failure rate per path. A request is considered failed when the HTTP status code is 4xx or 5xx.

Config

Metric	Description
Number of times Reload is configured	Number of NGINX configuration file reloads in the last minute.
Number of Pods that failed to configure	Whether the last reload succeeded. Displays the number of controller pods if the last reload failed, or N/A if it succeeded.

Latency

Metric	Description
delay time	Line chart of percentile latencies for all requests.
Delayed heat map	Heatmap of request latencies for the selected Ingresses in the selected time period.
Request processing delay (by Path)	Line chart of request latencies per path.
Upstream service delay (by Path)	Line chart of upstream service latencies per path.

Controller load

Metric	Description
Network I/O	Line chart of inbound and outbound traffic for the NGINX Ingress controller.
CPU	Line chart of CPU utilization for the NGINX Ingress controller.
Memory	Line chart of memory usage for the NGINX Ingress controller.

Prometheus metrics reference

The following Prometheus metrics are exposed by the NGINX Ingress controller. Use these metrics to build custom dashboards or alerting rules.

Request metrics

Metric	Description
`nginx_ingress_controller_request_duration_seconds_bucket`	Performance analysis of latencies of monitoring requests.
`nginx_ingress_controller_response_duration_seconds_bucket`	Performance analysis of latencies of upstream services.
`nginx_ingress_controller_request_duration_seconds_count`	Number of requests based on the following labels: request latency, status code, path, method, and host.
`nginx_ingress_controller_request_size_sum`	Size of request packets based on the following labels: request latency, status code, path, method, and host.
`nginx_ingress_controller_response_size_bucket`	Size of the request response packet.
`nginx_ingress_controller_response_size_count`	Number of request responses based on the following labels: request latency, status code, path, method, and host.
`nginx_ingress_controller_response_size_sum`	Size of response packets based on the following labels: request latency, status code, path, method, and host.
`nginx_ingress_controller_requests`	Number of requests.

Controller metrics

Metric	Description
`nginx_ingress_controller_config_hash`	Hash value in the NGINX configuration file.
`nginx_ingress_controller_config_last_reload_successful`	Whether the NGINX configuration file is reloaded.
`nginx_ingress_controller_success`	Number of times the NGINX configuration file is reloaded.
`nginx_ingress_controller_ingress_upstream_latency_seconds_sum`	Latency of upstream service requests processed by Ingresses.

NGINX process metrics

Metric	Description
`nginx_ingress_controller_nginx_process_connections`	Number of connections of NGINX services.
`nginx_ingress_controller_nginx_process_cpu_seconds_total`	CPU utilization of the NGINX service.

Overview

The Overview tab shows traffic metrics for a specific Ingress, filtered by Ingress name and namespace.

Metric	Description
PV	Total request count in the selected time range.
Inflow	Total inbound traffic in the selected time range.
outflow	Total outbound traffic in the selected time range.
Status code distribution	Pie chart of HTTP status codes.
Method distribution	Pie chart of request methods (GET, POST, PUT, etc.).
Distribution of Ingress request	Pie chart of controllers used to process requests.
PV (line chart)	Request count over time.
Failure rate	Line chart of request failure rate. Excludes requests that return HTTP 1xx--3xx status codes.
Request delay	Line chart of P50, P95, P99, and P9999 latencies.

TopN

The TopN tab ranks services, Ingresses, regions, and client types by key metrics to help identify hotspots and outliers.

Request statistics

Metric	Description
Service request Top10	Top 10 services by request count.
Service failure rate Top10	Top 10 services by failure rate. A request is considered failed when the HTTP status code is 4xx or 5xx.
Service delay Top10	Top 10 services by response latency.
Service traffic Top10	Top 10 services by traffic volume.
Ingress Top10	Top 10 Ingresses by request count, success rate, average latency, or inbound/outbound traffic.
Service Top10	Top 10 services by request count, success rate, average latency, or inbound/outbound traffic.

Geographical statistics

Metric	Description
Country (region) request Top10	Top 10 countries and regions by request count.
City Request Top10	Top 10 cities by request count.

Equipment statistics

Metric	Description
Browser request Top10	Top 10 browsers by request count.
Device type request Top10	Top 10 device types by request count.
Operating System Request Top10	Top 10 operating systems by request count.

URL Analysis

The URL Analysis tab provides per-URL metrics. Enter a URL or HTTP status code in the Search field to filter results by specific traffic characteristics.

Metric	Description
QPS	Queries per second per URL. Filterable by cluster, namespace, Ingress, service, and time range. Supports URL aggregation.
Duration - P95	P95 latency per URL. Filterable by cluster, namespace, Ingress, service, and time range. Supports URL aggregation. Sampling period: 5 minutes.
Failure rate	Request failure rate per URL. Filterable by cluster, namespace, Ingress, service, and time range. Supports URL aggregation. Sampling period: 1 minute.
Error details	Failure rate per URL and HTTP status code. Filterable by cluster, namespace, Ingress, service, and time range. Supports URL aggregation. Sampling period: 1 minute.
Recent request log	Request logs in the selected time range.

Log Analysis

The Log Analysis tab displays aggregated Ingress request logs. To trace a specific request, expand a log record by clicking the Angle bracket icon icon, then click the URL next to TraceId to open the trace analysis page in the ARMS console. For more information, see Trace query.