Hosting Django Docker Postgres on DigitalOcean Droplets-CodePudding

I am new to deploying to DigitalOcean. I recently created a Droplet, into which I will put my Django API to serve my project. I saw that the droplet I rent has 25 GB of memory, and for the actual size of my project, it could be a good fit for the database. I run my entire code with Docker Compose, which handles Django and PostgreSQL. I was wondering if it is a good idea to store all the data in the PostgreSQL created by Docker Image, and stored in the local memory of the droplet, or do I need to set up an external Database in Digital Ocean?

Technically I think that could work, but I don't know if it is safe to do, if I will be able to scale the droplets once the size scales, or if I will be able to migrate all the data into an external database, on digital ocean, a managed PostgreSQL all my data without too many problems.

And if I save it all locally, what about redundancy? Or can I clone my database on a daily basis in some way?

If someone knows about that and can help me, please I don't know where to start

CodePudding user response：

Very good question!

So your approach is perfectly fine, even though I wouldn't use docker compose for your production environment. The reason being that you will definitely need to rebuild and redeploy your application container multiple times and it's better to decouple it from your database container.

In order to do that, just run something similar to this in your server:

docker run -d \
    --name my_db \
    -p 5432:5432 \
    -e POSTGRES_PASSWORD='mypassword' \
    -e POSTGRES_USER="postgres" \
    -v /var/run/postgresql:/var/run/postgresql \
    --network=mynetwork \ 
    --restart=always \
    postgres

Make sure to define a custom network and to register in it your db container as well as your web container and any other standalone service that you need (for example Redis or Celery)

The advantage of the Digital Ocean databases is that they are fully managed. Meaning that you can use their infrastructure to handle backups, and that it would run on a separate server that you can scale up and down based on need.

If I were you I would start with a PostgreSQL container in your server, making sure to define a volume so that even if your server instance crashes the data will be stored on file and will be there waiting for you when you restart the server and db.

Then as your needs evolve you can migrate your data on a more reliable and flexible system. This can be a Digital Ocean database or AWS RDS or a number of others. DB migrations can be a bit scary but nowadays there are plenty of tools to help you with those too.