I think The Manga Guide to Databases would be a good read for you.

Be sure to avoid places where you can get the book for free such as here.

The Lemmy app we are using on our phones needs to download content from Lemmy so it can be displayed to us. Lemmy might just have one big file full of links, but that’s annoying to have to write code to handle. Or it might have a folder full of files where each file is a post, but that’s also a bit annoying to write code to manage.

It (probably) uses a local SQLite database to store all of the cached posts.

Conceptually, a database is just a place to store things, just like a big text file. The database just handles a lot of the grunt work for you and makes it easier to search, organize, and filter the data.

So anywhere there is data, there could be a database.

not sure what you mean by ‘come together’… for users there would a ‘front end’ user interface, either a website or an application that includes access control and forms and fields, etc. on the ‘back end’ there would be an application server somewhere that connects to the database and lets certain users view/edit certain things as scripted.

ELI5: a database is the “memory” of a program.

Every piece of data that any software uses almost certainly comes from and goes to multiple databases.

Once the data is stored, you can execute “queries” to have powerful access to update many records at a time, read particular records based on their relationship to other records, and so much more.

Your bank balances, your purchase history, your emails, every part of your digital life is almost certainly spread across a constellation of databases.

Bonus Fediverse content:

Lemmy itself uses the Postgres database extensively. Posts, users, comments, votes and more are all individually stored in the database.

Mastodon also uses Postgres. If a post goes up on Lemmy, and a Mastodon server is federated with it, the Lemmy server will send out a HTTP request to the Mastodon server containing the contents of the post. The Mastodon server will use this information to write its own record of the post in its own database.

Regarding your question about VMs: You can run a database inside a VM, or give the VM access to an outside database via queries, or both! You might run SQLlite (a small and excellent embedded database) on the VM to track its local state, while also running queries against a large postgres database to synchronize with other services in the cluster.

Databases are collections of files sorted intelligently[1] and normally unintelligible to humans. But they are files in the end. Then, a system is located between users (humans or other systems) and those files. That system is called a Database Management System (think of the MySQL program).

Then, users can ask the Database Management System questions about the data or give commands that will cause new data or changes in the data, and the system will search the files and/or modify their contents to answer the questions or satisfy the commands.

You see, it is simply a collection of files and a program that can manage those files (so it needs access to them). In the real world, that tuple can be deployed in multiple ways.

Now, the concepts of “program,” “process,” and “files” are high-level concepts. They exist thanks to the Operational Systems, which allow us to speak in those terms. And that’s it: You need at least an Operational System to deploy a database solution.

In the real world, a single big server machine with a lot of space can be dedicated to a database solution. In the LAMP age, this was a normal practice. This example has no virtual machine or container; it is just a plain machine. That’s still an option.

There are more complex solutions that split these main concepts (files, programs/processes) into different “run time” environments (machines, virtual machines, clusters, etc.) where they can be fine-tuned to a specific task.

Now, the whole deal with the databases is to coordinate multiple users while serving as permanent memory. It is the central point where everything else orbits. They try to provide users with an ideal data model that is always updated at the quantum moment so that when users observe it, they can trust that it is the truth.

[1] With “intelligently” I mean that they are sorted in that way (very complicated way) for a reason: to facilitate typical operations, a database solution has to be performed, like searching for data and manipulation, taking advantage of schemas (if it is a database with schema).