Light Gateway Tutorial

We have recently combined both light-proxy and light-router to create a new light-gateway project. Like the http-sidecar, which is based on the same codebase and used as a sidecar in a Kubernetes cluster, the light-gateway can be used in Docker, Windows or Linux services.

The following tutorials demonstrate the usage of the light gateway. We use two demo services to complete the full cycle: market service and petstore service.

Both services are located in the light-example-4j project. You need to check out the repo and build both locally.

Also, we need to check out the light-gateway project in your workspace.

cd ~/networknt
git clone https://github.com/networknt/light-example-4j.git
git clone https://github.com/networknt/light-gateway.git

Build market service

cd ~/networknt/lgight-exmaple-4j/rest/market
mvn clean install -Prelease

Build petstore service

cd ~/networknt/light-example-4j/rest/petstore-maven-single
mvn clean install -Prelease

Build light-gateway

cd ~/networknt/light-gateway
mvn clean install

Multiple Rule Actions

So far, we know how to use the yaml-rule-plugin to perform the request & response transformation in the chain for most of the use cases. However, if there are complex senarios that need more transform steps or plugins, we can configure the light-gateway values.yml and rules.yml to chain multiple rule actions together. Here is the log in openapi-petstore terminal. 2024-11-19T10:33:37.862 [XNIO-1 task-2] test-correlation INFO c.n.p.handler.FlowersPostHandler:38 handleRequest - requestHeaders = {accept-encoding=[gzip, deflate, br], x-traceability-id=[test-traceability], x-forwarded-port=[8443], x-forwarded-for=[0:0:0:0:0:0:0:1], vnd. Read More »

Request Response Transformation

In this tutorial, we will demo the request and response transformation based on the rule engine in the light-gateway. The same handlers/interceptors can be used in the http-sidecar and any application built with the light-4j. The demo will be based on three real use cases from our customers. We will show you the configuration and run the test with the petstore backend API to show you how the light-gateway works on the request and response chain. Read More »

Client Proxy Server Proxy

After we have done the server-proxy, we can set up the market API to use the client-proxy to access the petstore API. Assume that the petstore API and the server-proxy still up and running. The following command can confirm the petstore is working and accessible from the server-proxy. curl -k https://localhost:7443/v1/pets And the result should be. [{"id":1,"name":"catten","tag":"cat"},{"id":2,"name":"doggy","tag":"dog"}] Client Proxy Let’s start another light-gateway instance for the client-proxy. The market API will call the client-proxy, the client-proxy will call the server-proxy and then the server-proxy will call the petstore API. Read More »

Server Proxy

This tutorial assumes that we build the petstore API with .NET, Python, or Nodejs without any cross-cutting concerns. Let’s put the light-gateway instance on the localhost to access the petstore API to address all the cross-cutting concerns. Petstore Serivce First, let’s start the Petstore API and ensure it is working. The default config should have the petstore service started at 9443 with HTTPS/2. cd ~/networknt/light-example-4j/rest/petstore-maven-single java -jar target/server.jar Second, let’s send a request to the petstore server to ensure it is working. Read More »