Glossary: databases

← Databases

Databases carry jargon that often hides the point: driver, index, transaction, replication. Below is what the terms mean in the context we use them — design, migration and reporting — and why they matter the moment data has to move from one platform to another.

Updated on 28 August 2026

Basic terms

Driver (ODBC, JDBC)
The piece in between that lets a program talk to a specific database. Without the right driver a reporting tool can do nothing, however open the database is.
Index
A lookup aid alongside the table. Makes queries fast, costs space and slows writing. A damaged index can usually be rebuilt.
Transaction
A sequence of steps that may only succeed or fail together. This is what stops an invoice being half posted.
Schema
The layout of the database: which tables, which fields, which relations. In a migration this is the first thing that has to be right.

In migration and reporting

Data type
The kind of value in a field: text, integer, amount, date. Most migration errors originate right here.
Collation and character set
The rules for sorting and characters. Differences here explain why accented names end up in the wrong place after a migration.
Replication
Keeping two databases automatically in step. Useful for reporting, but no substitute for a backup: a mistake is copied across neatly.
Data warehouse
A separate database where data from several systems comes together, arranged for querying rather than editing.

Most database discussions end up being about definitions rather than technology: who counts as a customer, when is an order complete. That question always returns.

More on this →

Question not covered here?

Describe briefly what it is about and we will tell you what is possible.