Why I Revisited MS SQL Server Basics: A Deep Dive into String Data Types

Today, when I raised PR with my manager (It was regarding store procedure), my manager asked me a couple of questions about that.

Like, Why do you declare the data type of string as varchar(n) instead of nvarchar(n) and other scenario-based questions?

You might also have faced a similar situation with seniors or with teachers in college.

Although I gave answers to him, that debate raised questions for me on my basics.

And I decided to revise from the start.

While revising I noticed a few points, which I thought to share with you all, so I am writing this article.

Let's start from the start.

When we create a database 2 files get generated behind the scene.

1. Data file: This has a (.mdf) extension and is known as a primary data file.
2. Log file: This has a (.ldf) extension and is known as a transactional log file.

Although, we can configure how many files should be generated. And if we configure more than 2 files there is 3rd category which is :

• User-defined files: and those known as secondary data files having the extension (.ndf)

Whenever we perform any DB transaction it first gets stored in the log file and once the transaction is completed it gets updated into the Primary data file.

The default and the lowest size of each file is 8 MB and maximum there is no limit. And for the database, according to Microsoft, one database can be up to 524 PB.

Select vs Print query

If you have used the "select" query then, you might have heard about the "Print" query also.

Let's understand when to use which.

Basically "select" query returns the table data in grid format whereas the "Print" query returns data as text.

So whenever we need to perform an operation on table data we use the "select" query and mostly to print logs or errors we use the "print" query.

Now, Let's discuss the most favorite interview question.

Char vs Varchar vs Nvarchar?

Char:

Char store max 8000 characters, and hold 1 byte per character space.

• If we don't specify size, by default SQL assigns 1 byte to it and holds only one character.

• If we give more input than the defined size, it auto-truncates the data.

• Char follows the static memory allocation, so if we allocate a 20-byte size and store 10 characters, SQL still allocates all 20-byte storage and the remaining 10 bytes will be considered a waste of space.

Varchar:

• Varchar can store max 2GB of data, and hold 1 byte per character space. Although it stores up to 8000 bytes of data in the table anything beyond that is stored in the text file.

• So you might have noticed its limit as 8000 bytes in multiple articles.

• To declare a column to store 2GB data, we can use varchar(max).

• Varchar follows the dynamic memory allocation. So if we allocate a 20-byte size and store 10 characters, SQL will allocate only 10-byte storage and the remaining 10 bytes will get saved.

Nvarchar:

• Nvarchar follows the same properties as varchar, the only difference is that it uses 2 bytes to store one character.

• because it follows Unicode standard data, to support multiple languages.
  And char, varchar follows ASCCI standard data.

There is one more datatype called Text, but it has been deprecated, so you can ignore it.

There is one article that I found very useful during my research. It has detailed information about all data types.

You can go through it for more information.

If you think I have missed something or need to update feel free to comment. It will help me as well as our community.