Webwatcher

Webwatcher is a web application that monitors your websites and e-mail you when one of them becomes unavailable.

How it works?

Webwatcher is a standard web application. It has two parts:

• The backend, a REST API written in Python that relies on a MySQL database
• The frontend is a website written using HTML, CSS and vanilla JavaScript

Webwatcher has been conceived to be easily deployed on Google App Engine (see "Deployment on Google App Engine" section below).

The web interface displays the status of the websites you are monitoring and allows you to set the followings:

• The websites you want to check
• The people who will be e-mailed in case of a failure
• The global settings (SMTP server, option to avoid cache)

Webwatcher does not check the websites automatically. The websites are checked only when a request is made to the special URL /api/check. To automate this check, you need to set up a cron that will call this URL periodically. For example, you can use the following entry in the crontab of the server to check the websites every 5 minutes:

*/5 * * * * curl http://localhost:8080/api/check

Requirements

You will need the following dependencies:

• Python 3 with PIP and venv
• Node.js with NPM
• An SQL database

Database

The application relies on the following environment variables to connect to the database:

• DB_HOST: the host (default localhost)
• DB_PORT: the port (default 3306)
• DB_SOCKET: the socket, can be used instead of DB_HOST and DB_PORT (no default value)
• DB_USER: the user (default root)
• DB_PASSWORD: the password (default root)
• DB_NAME: the name of the database (default webwatcher) Set these variables only if you would like to customize their values.

The database will be automatically initialized by the application itself.

Sessions

The website uses client side "sessions" (using cookies). The cookies are encrypted using a key. By default, a random key is generated each time the application starts. This does not work for the Google App Engine environment, where multiple instances of the application can run at the same time. The environment variable COOKIE_SECRET allows choosing the key that will be used to encrypt the cookie.

Local usage

The first step is to install the requirements described in the Requirements section.

The back-end

First, install the dependencies:

python3 -m venv .venv
source .venv/bin/activate
pip3 install -r requirements.txt

Then, launch the back-end:

python3 main.py

The front-end

First, install the dependencies;

npm install

Then, start the front-end:

npm start

Finally, open your browser on http://localhost:8080.

Server usage

The first step is to install the requirements described in the Requirements section.

You will also need a production-ready WSGI server such as Gunicorn and a front server such as Nginx (acting as reverse proxy for the back-end and a classic HTTP server for the front-end). Install them with the following command:

sudo apt install nginx python3-gunicorn

Set up the project

On the server, checkout the code of the project in a folder, for example in /opt/webwatcher. Then, in this folder, execute the following commands to install the project dependencies and generate the front-end bundle:

python3 -m venv .venv
source .venv/bin/activate
pip3 install -r requirements.txt
npm install
npm run build

Set up a systemd service

Copy the file resources/webwatcher.service to the remote path /etc/systemd/system/webwatcher.service. Then, run the following command to enable and start the service:

sudo systemctl enable --now webwatcher.service

Set up Nginx

Copy the Nginx configuration file resources/webwatcher in /etc/nginx/sites-available/webwatcher and enable this configuration with a symbolic link:

sudo ln -s /etc/nginx/sites-available/webwatcher /etc/nginx/sites-enabled/webwatcher

Then, reload Nginx:

sudo nginx -s reload

Check websites automatically

Do not forget to set up the cron that will check the websites regularly. You can add the following entry in your crontab to check the websites status every 5 minutes:

*/5 * * * * /usr/bin/curl http://localhost:8000/api/check

Deployment on Google App Engine

This application is ready to be deployed on Google App Engine. The best is to create a new Google Cloud Platform project dedicated to the application.

You will need the following dependencies:

• Google Cloud SDK (follow the instructions to log in to your project)
• Node.js with NPM

Create the database instance

In your project, use the menu Databases > SQL to create a Google SQL instance. Make sure to configure the instance properly:

• Use MySQL version 8.0
• Choose a strong password and note it
• In Customize your instance > Connections, allow the access though a public IP (you can disable the access through a private IP)

Wait for the instance to be created. Back to the main page of your instance, get the connection name of the instance (visible in the block Connect to this instance under Connection name). It should be something like project-id:us-west1:webwatcher. The name of the connection must then be prefixed with /cloudsql/ to get the final name of the connection (/cloudsql/project-id:us-west1:webwatcher with the previous example).

Then, you must create the database itself. From the main page of your instance, select Databases in the left menu and create a database named webwatcher (or any other name that pleases you). Keep default "Character set" and "Collation".

Allow the Google App Engine service account to use the Cloud SQL API

A Google App Engine application relies on a Google Cloud Platform service account. You must ensure this account has the appropriate Cloud SQL role and permissions as described here.

Here is the simplified process:

• Use the menu IAM & Admin > IAM
• Identify the service account used by Google App Engine (its name should be App Engine default service account)
• Edit the user and add the role Cloud SQL Client

Configure the application

Next step is to configure the application:

• In the file environment.yaml, add your custom database connection settings (the name of the connection, the password, and possibly the name of the database if you used a custom name) and generate a strong key for the cookies
• In the file cron.yaml, you can change how often websites are checked

Build the front-end

First, install the dependencies;

npm install

Then, build the front-end:

npm run build

Deploy

Finally, deploy the application with the Google Cloud CLI:

gcloud app deploy

Tips

Access your Cloud SQL database from anywhere

In your project, use the menu Databases > SQL and select your Google SQL instance. Then, select Connections and add your IP in the section Authorized networks.