gitlab-org--gitlab-foss/doc/development/database/constraint_naming_convention.md

stage	group	info
Data Stores	Database	To determine the technical writer assigned to the Stage/Group associated with this page, see https://about.gitlab.com/handbook/engineering/ux/technical-writing/#assignments

Constraints naming conventions

The most common option is to let Rails pick the name for database constraints and indexes or let PostgreSQL use the defaults (when applicable). However, when defining custom names in Rails, or working in Go applications where no ORM is used, it is important to follow strict naming conventions to improve consistency and discoverability.

The table below describes the naming conventions for custom PostgreSQL constraints. The intent is not to retroactively change names in existing databases but rather ensure consistency of future changes.

Type	Syntax	Notes	Examples
Primary Key	pk_<table name>		pk_projects
Foreign Key	fk_<table name>_<column name>[_and_<column name>]*_<foreign table name>		fk_projects_group_id_groups
Index	index_<table name>_on_<column name>[_and_<column name>]*[_and_<column name in partial clause>]*		index_repositories_on_group_id
Unique Constraint	unique_<table name>_<column name>[_and_<column name>]*		unique_projects_group_id_and_name
Check Constraint	check_<table name>_<column name>[_and_<column name>]*[_<suffix>]?	The optional suffix should denote the type of validation, such as length and enum. It can also be used to disambiguate multiple CHECK constraints on the same column.	check_projects_name_length check_projects_type_enum check_projects_admin1_id_and_admin2_id_differ
Exclusion Constraint	excl_<table name>_<column name>[_and_<column name>]*_[_<suffix>]?	The optional suffix should denote the type of exclusion being performed.	excl_reservations_start_at_end_at_no_overlap

Observations

• Prefixes are preferred over suffices because they make it easier to identify the type of a given constraint quickly, as well as group them alphabetically;
• The _and_ that joins column names can be omitted to keep the identifiers under the 63 characters' length limit defined by PostgreSQL. Additionally, the notation may be abbreviated to the best of our ability if struggling to keep under this limit.