bStats

The parent repository of the bStats project.

Repository structure

This repository contains all bStats repositories as submodules. These are

bstats-backend - The backend of bStats.
bstats-web - The (new) frontend of bStats.
bstats-legacy - The original backend and frontend of bStats.
bstats-metrics - The Java-based metrics classes of bStats.

Currently bStats is in a transition phase from the original backend + frontend to a complete rewrite. The "live" version of bStats is currently running with the old frontend and a combination of old and new backend. Which routes are served by which backend can be found in the nginx configuration at /prod/volumes/nginx/nginx.conf.

This repository also provides the development environment for bStats.

Clone repository

When cloning the repository, it is recommended to use the --recursive flag to clone the repository with all submodules.

git clone --recursive git@github.com:Bastian/bStats.git

Alternatively, you can also clone the submodules after cloning this parent repository by running

git submodule init && git submodule update

Start Development

There are some prerequisites for development:

Docker
Linux. WSL2 is recommended for Windows. MacOS should work but is untested
GNU Make (should be available on most Unix systems)

Additionally, you need a Firebase account place the firebase-config.json file and the service-account-file.json file inside the dev directory.

Then you can start all services by running make. The services will run in Docker containers, thus no additional dependencies like Node.js and npm are required (even though it is recommended to have them installed for your IDE tooling).

Start Production

The prod directory contains the production environment for bStats. Similar to the dev environment, you have to place the firebase-config.json file and the service-account-file.json file inside the prod directory. Additionally, you have to update the prod/volumes/bstats-legacy/config.json file with your Recaptcha secrets and replace the sessionSecret with a random (long) string of your choosing.

You can then start the services by running make start-prod.

Backups

Redis data is backed up to AWS S3. AWS credentials have to be provided in a AWS.env file in the prod directory. Afterwards, cronjobs can be set up with the add-backup-cronjobs.sh script in the prod directory.

A lifecycle configuration for S3 is given in the prod/s3-lifecycle-config.json file.

Disaster Recovery

Assuming the worst happens and the current bStats server explodes. To get back up running, start the prod environment on a new server once and stop it after it has successfully started. Then place the backup Redis dumps (dump.rdb) in the prod/volumes/redis/node-<x>m directory. Make sure that the dump file is the correct one for the current node. You can look at the prod/volumes/redis/node-<x>m/nodes.conf file to find out for which key-range the node is responsible. You can then start the prod again and everything should work again (if it is a new server with a new IP, you ofc also have to update the DNS settings in Cloudflare).

License

This project is licensed under the MIT License.