You all know that when it comes to deploying an application, Linux is one of the best places to do it. But what if you build an application using MongoDB and the next day, when you try to deploy it on Linux, you find out that MongoDB isn't compatible with your Linux environment, and you cannot install another version of Linux to fix this compatibility issue?
So, what should one do?
Here's what I did.
I was using Ubuntu 24.04 LTS with the Linux 7.0.0-38-generic kernel. When I tried to run MongoDB 8.0, it failed to start because of a kernel compatibility issue related to TCMalloc.
I also found out that MongoDB 7.0 does not officially support Ubuntu 24.04 LTS. According to MongoDB's official documentation, MongoDB 7.0 supports Ubuntu 20.04 and 22.04, while support for Ubuntu 24.04 starts with MongoDB 8.0.
So, what could I do without changing my Linux installation?
I thought, why not try Docker?
I pulled the MongoDB 7.0 image from Docker Hub and tried to run it inside a Docker container.
And you know what?
It worked!
🤔 But why did it work?
The reason is that Docker containers share the host machine's Linux kernel, but they have their own userspace environment, including libraries and other dependencies. They don't need to use the host's installed MongoDB packages or libraries.
This means I could run MongoDB 7.0 inside a container without installing it directly on Ubuntu 24.04. The container provided an isolated environment with the dependencies required by MongoDB 7.0, while still using the host's Linux kernel.
That's when I realized why Docker is so useful for deploying applications. Instead of configuring everything on a single machine and worrying about compatibility issues, we can use Docker images to run specific versions of software in isolated environments.
This experience helped me understand why Docker makes deployment easier and why using images can save us from a lot of configuration headaches.
🐹 Thank you for reading
