Reluctant DBA Part 2: Terminology
This week, let’s nail down some SQL Server terminology - just to make sure we’re all on the same page with what these words mean in the SQL Server universe.
SQL Server is referred to as an RDBMS. What’s that?
A relational database management system, or RDBMS, is usually a server-based product that manages relational databases.
On one level, the server must manage the same types of things any server product manages: Security, client connectivity, data backup and restore, and so forth. On another level, the server must execute instructions (usually in a variant of the SQL language) that retrieves data, modifies data, adds data, and removes data from one or more databases.
The relational part of RDBMS comes from the fact that the database can relate various parts of data to each other. For example, your data might include a list of customers and a list of orders they have placed; the RDBMS provides the means to connect these two, so that you can retrieve orders placed by a specific customer. This is in contrast to a flat file database, which would only be able to contain a single list of data - just customers, for example, or just orders. A comma-separated values (CSV) file is a good example of a flat file database.
Today, most any database product - including desktop products such as Microsoft Access or Apple’s Bento - have relational capability, and many products don’t even use the RDBMS term any more, or make any reference to “relational,” since it’s more or less assumed that the relational capability is there.
So what, then, is a database? What’s a table?
Simply put, a database is a collection of tables. A database usually contains all of the data related to a particular application. An order-processing application might include data such as customers, orders, products, and more; all of that data would be contained within a single database.
In some products, like Microsoft Access, a database is held within a single file (either an .MDB file or an .ACCDB file, depending on the version). In other products, like Microsoft SQL Server, the database might be spread across multiple files (with .MDF, .LDF, and .NDF filename extensions). Spreading across multiple files is often a way to improve performance by involving multiple disks, or to meet other goals such as backup performance.
In many products, including SQL Server, certain configuration options are made at the database level. For example, SQL Server offers the option to automatically close a database file when it is not in use; this would affect the entire database and all of its files.
A table is simply a list of data, not unlike a spreadsheet in some respects.
Each column in the list is referred to as a domain, and contains a particular type of information: A name, address, price, and so forth. Tables are usually represented visually in a grid, and each column is usually topped by the column name.
Each row in the list is referred to as an entity, a row, or a record. A row represents all of the information for a given thing. For example, in a table of customer information, a single row represents a single customer, with columns for their name, address, city, country, phone number, and so forth.
Tables are contains within a database, and are managed as part of that database. Because some tables relate to each other - for example, a list of customers would have an obvious relationship to a list of orders placed by customers - tables are usually not managed individually, but rather as a set - all the tables in the database comprising that set.
Related Posts
Rate this Post
(No Ratings Yet)
Loading ...
