Why all Developers most know SQL

Even with NoSQL alternatives, developers should learn SQL

Originally published at https://fek.io.

I spent last Saturday going to the 2022 SQL Saturday event here in Jacksonville, FL. You can think of this event as a conference in a day, where system admins, developers and Devops workers get together to learn more about Microsoft SQL Server.

Microsoft SQL Server was the first enterprise relational database I worked with as a programmer. Sql Server like many other relational database systems share something in common, and that is the Structured Query Language.

Back in 2003 I went to a big box store to buy a laptop that was powerful enough to run SQL Server. This was long before Docker. The salesman at the store asked me what I was planning on using the laptop for, and I told him I was planning on using it to run SQL Server. He told me that he had recently graduated from a college with a degree in computer science, and now he wanted to learn SQL. I was shocked that they did not teach any SQL databases in his course work, but I have heard this from other computer science students since.

No matter what type of development you wind up doing, their is a good chance that you will have to work with some type of database. Data does not have much value if it can not be persisted. The statistic I heard at the time was that over 50% of development jobs required the use of some sort of database.

Relational Database Drawbacks

As great as relational databases are, they do have some drawbacks. One of the drawbacks is that they can be difficult to scale. Scaling most relational databases used to mean scaling vertically, meaning that because of having data consistency, all data had to be persisted on some sort of disk or volume. There are tricks that some vendors and projects have used to make these databases scale.

One of the methods is to use database mirroring or clustering, where data can be copied from disk to another for the same database. While this method offers failover reliability it still means using a server or service that can be a single point of failure.

Another method that can be used is database sharding. Sharding a database can be as simple as…