Is it a bad idea to run a database in Docker for production?

For business production environments, I'd recommend dedicated solutions like Azure SQL or running a SQL cluster on virtual machines. This gives you control over high availability and vendor support. However, for testing and home labs, Docker can be pretty convenient since it's less critical if you accidentally delete something.

One key point to consider is that Docker containers are designed to be stateless. While you can mount a folder on your host to keep data, you might run into scaling and backup challenges later on. It’s okay for temporary environments like development or testing, but for production, it's usually better to keep your database separate from containerized applications.

This is a really common question! What did you find when you looked it up online? Is there something specific about the risks that you’d like clarification on?

Containerized databases can be a bit risky for production. The main concern is that if the entire process fails, you could lose data or experience downtime. Unless you're set up with a highly available system with automated backups and performance tuning, you might want to stick to managed database services or run your DB on a container orchestration platform instead.

Honestly, your friend might be a bit behind the times. There are plenty of successful setups using Docker for databases today. Performance can depend on various factors, including how the database is accessed. In a well-isolated environment or a Kubernetes cluster, running a database in Docker can work without many risks.