endocode-test

This repo contains:

• http-service, a small application written in Golang that serves few HTTP endpoints
• scripts (jenkinsfile + helm chart + terraform) to automate the deployment of the application in an existing kubernetes cluster
• helm charts to deploy and EKF + metricbeat stack for logging and monitoring

http-service

http-service is a simple golang application that serves HTTP requests.

OS Requirements

• macOs, Linux

Building and running with docker

This method requires a machine with docker. To run the service in a docker container, simply run make docker: dockerd will automatically build an image from the provided Dockerfile and run it with the option -d. By default, it listens to the port 8080. This value can be changed by setting the HOST_PORT variable. For example, if one would want the service to isten to the port 8081, the command would be:

make docker HOST_PORT=8081

local install

Requirements

• make
• golang 1.13 or newer
• git

To compile the source code run make http-service, the dependencies will be taken care of. To execute it, simply run make run; this will start the service, by default listening on the port 8080. To change it, export the variable LISTENING_PORT and set it equal to a port you'd liek the service to listen on.

Usage

The service accepts two endpoints:

• GET /helloworld - returns "Hello Stranger". It accepts one query parameter, name. If set, it returns "Hello $name", sliced by camel case
• GET /versionz - returns a JSON with the hash of the latest commit and the project name

For example, to call the first endpoint, you'd do:

    curl localhost:8080/helloworld

and the answer will be:

    Hello Stranger

To use the query funcionality, use this request

    curl localhost:8080/helloworld?name=MarcoRossi

and the answer will be:

    Hello Marco Rossi

The last endpoint can be contacted with

    curl localhost:8080/versionz

and the answer will be:

    {
        "git_commit": "2d23bd462aa5523a0bdcd272d4958700e3cc6eac",
        "project_name": "http-service"
    }

Deploy

This service can be deployed in an existing kubernetes cluster using the provided jenkinsfile, it relies on terraform. The main.tf contains all the instructions terraform needs to deploy the service inside the kubernetes cluster; it's based on an helm chart, http-service-chart that deploys the service plus an nginx ingress, that allows communication with the service from outside the cluster. By default, the host name to contact the service with is http-service.int, but to make the connection possible, the line $CLUSTER_IP http-service.int has to be appended to the host /etc/hosts file (N.B. $CLUSTER_IP is a placeholder, it stands in for the actual cluster ip, obtainable with the command kubectl cluster-info).

Logging

There is another terraform file, terraform/monitoring/main.tf, it contains instructions to install an EKF stack + metricbeat in the kubernetes cluster. It's based on the official elastic helm charts for Kibana, Elasticsearch and metricbeat. They require the official repo to be addedd:

    helm repo add elastic https://helm.elastic.co

and on a custom charts for Fluentd, fluentd-chart. It deploys also an nginx ingress that allows acces to the Kibana UI, by default at the address http://kibana.int. To access it, add the line $CLUSTER_IP kibana.int in the host /etc/hosts file.

Server stub

A configuraton file to generate a server stub is provided, openapi.yaml. To generate the stub run the commands:

    git clone https://github.com/openapitools/openapi-generator
    cd openapi-generator
    mvn clean package
    java -jar modules/openapi-generator-cli/target/openapi-generator-cli.jar generate \
    -i openapi.yaml \
    -g $LANGUAGE \
    -o /var/tmp/php_api_client

There are several $LANGUAGE choices, check the openapi-generator github page for a complete list.