AMP-STACK

This repo shows a sample implementation of the L/AMP stack using Helm on a Kubernetes cluster on the Google Cloud Platform.

Specifics

The application uses the stable/lamp helm chart with some settings on top. The stack runs each service from separate containers in the pod. It uses the official images and the ones currently used are:

• httpd:2.4.29-alpine
• php:7-fpm
• mysql:5.7

Communication between the three services is realized over unix sockets.

A running instance

I have used this repository to create a demo of a running amp stack. It is exposed here http://35.204.68.180/

To see how I did this or to create your own, follow the instructions below.

Prerequisites

You need to have kubectl, Google Cloud SDK (for gcloud) and Helm installed and setup locally.

This part cannot be automated as the steps vary per environment/OS

The Cluster

To avoid mangling your own ~/.kube/config, the scripts use the kubectl config from config/kube-config. It has the context setup for my personal cluster. You will need to provide a cluster and authenticate to it (using gcloud), and replace this file before you continue onwards.

Setup Helm

Once you have the helm client installed locally, run the following command to install Tiller on the cluster:

$ ./setup-helm.sh

Install the AMP stack

$ ./init.sh

This will use helm to install the amp stack with our settings. It will also populate the web server with the initial contents from the html folder, which is just a Hello page.

When it is done it will display a URL for our working web site.

You can modify the scripts/files/html and create the stack again. Before executing the init.sh, you need to delete the old stack:

helm delete amp-endava --purge

Or you can modify the app_name variable in init.sh to create a new stack.

Updating the web content

To update the web site, add more files inside html and just run:

$ ./upload.sh