Kong NGINX Connections Test

In Kong API Gateway, NGINX connections refer to the network connections managed by the underlying NGINX server, which powers Kong’s request handling. These include active, reading, writing, and waiting connections between clients and upstream services. Monitoring these connections helps assess traffic load, concurrency levels, and server performance. Sudden spikes or high numbers of idle or open connections can indicate issues like traffic surges, slow upstreams, or resource bottlenecks, affecting API responsiveness and requiring administrative attention.

Monitoring NGINX connections helps administrators detect traffic spikes, manage concurrency, and identify potential bottlenecks. It ensures optimal resource utilization, prevents overload, and supports maintaining consistent API performance and availability under varying traffic conditions.

This test monitors NGINX Connections and reports key metrics related to connection rates and number of connections. Administrators can review these metrics and trends to identify any drop in performance and can take corrective actions before it affects the business application.

Target of the test : A Kong API Gateway

Agent deploying the test : An internal/remote agent

Outputs of the test : One set of results for the target Kong API Server being monitored.

Configurable parameters for the test

Parameter	Description
Test Period	How often should the test be executed.
Host	The IP address of the host for which this test is to be configured.
Port	Specify the port at which the specified Host listens. By default, this is 8001.
SSL Enabled	If the Kong API Gateway uses the HTTPS protocol, set the SSL ENABLED test parameter to Yes. By default, this parameter is set to No.

Measurements made by the test

Measurement	Description	Measurement Unit	Interpretation
Accepted connections Rate	Indicates the number of accepted connections per minute to this instance of Kong API Gateway.	Requests/min	Number of connections accepted by the NGINX server per minute during the last measurement period.
Handled connections Rate	Indicates the number of connections successfully handled per minute by this instance of Kong API Gateway.	Requests/min	Number of connections successfully handled by NGINX per minute during the last measurement period. This typically matches the accepted connections rate.
Percentage connections handled	Indicates the percentage of accepted connections that were handled by this instance of Kong API Gateway.	Percentage	A high percentage means most accepted connections were processed successfully. A lower percentage may indicate dropped or failed connections. Normally, accepted and handled connections should be equal.
Total active connections	Indicates the total number of active connections during the last measurement period to this instance of Kong API Gateway.	Number	Number of currently active connections, including those in reading, writing, and waiting states.
Reading connections	Indicates the number of connections where NGINX engine of this Kong API Gateway is reading from the client.	Number	Connections in which NGINX is currently reading the request header from the client.
Writing connections	Indicates the number of connections where NGINX engine of this Kong API Gateway is writing to the client	Number	Connections in which NGINX is sending a response back to the client.
Waiting connections	Indicates the number of idle/keep-alive connections to this Kong API Gateway waiting for the next request	Number	Connections where NGINX is idle, keeping the connection alive while waiting for the next client request.