Database Glossary

Use this page as a reference when a term interrupts your reading. You do not need to memorize it before starting SQL. The groups follow the learning path: data, queries, transactions, storage, distributed systems, and analytics.

Use a term in a concrete example

In orders.customer_id, table means the collection named orders, column means the field named customer_id, and value means a particular identifier such as 1. The schema declares that this field is required and references an existing customer. A row is one order containing that value alongside its identifier and date.

A primary key identifies the order itself. A foreign key checks its customer reference. A join uses the matching values to assemble a result. An index offers a way to find matching rows with less search work. These are four different mechanisms, even when they involve the same identifier.

If a term remains unclear, use the examples in the introduction, technology comparisons, DBMS walkthrough, or modeling exercise. The tables below are a lookup aid, not a replacement for those explanations.

Data and structure

For examples, see data models, normalization, and data integrity.

SQL and queries

Start with the runnable SQL introduction. For injection prevention, see SQL injection.

Transactions and concurrency

See transactions before concurrency control.

Storage and performance

See row and column storage, pages, and indexing.

Distribution and database models

See database types and distributed databases.

Analytics, operations, and tools

Product names are introduced in the engine chapters, where their capabilities have context. Start data warehousing after practicing aggregates and joins.

Work through missing values instead of memorizing the definition

Run this standalone SQLite example in a fresh database:

CREATE TABLE glossary_contacts (
    contact_id INTEGER PRIMARY KEY,
    phone TEXT
);
INSERT INTO glossary_contacts VALUES (1, '555-0100'), (2, NULL), (3, '');

SELECT COUNT(*) AS rows,
       COUNT(phone) AS supplied_phone_values
FROM glossary_contacts;

COUNT(*) counts every row. COUNT(phone) counts non-null phone values, so the empty string counts while NULL does not. A non-null value is not automatically a usable phone number.

SELECT contact_id
FROM glossary_contacts
WHERE phone IS NULL
ORDER BY contact_id;

This returns contact 2. IS NULL tests the missing-value marker; phone = NULL does not provide the same test because ordinary comparison with null produces an unknown result. Contact 3 has a present but empty string.

Technology names and their roles

A product name tells you which implementation to investigate; it does not stand in for a model definition or a performance guarantee.

The Couchbase document guide explains its key/document distinction. AWS's database decision guide identifies the cloud-service families; the repository's engine chapters develop them further.

Check your understanding

1. Why are a key, a constraint, a join, and an index not interchangeable terms?
2. Which glossary example shows that an empty string differs from null?
3. What is the difference between a document model and the MongoDB product?
4. Why does a Java persistence API not tell you whether the backend is relational?
5. Which term describes entry age, and which describes removal for memory pressure?