What is a UUID and what is it used for?

A UUID (Universally Unique Identifier) is a 128-bit identifier designed to be globally unique without the need for central coordination. It is commonly used in databases, distributed systems, APIs, and backend applications to identify records, sessions, transactions, or resources. Thanks to the vast number of possible combinations, a UUID practically eliminates the risk of identifier collisions even in very large systems.

UUID v1

Generated based on time and the device's MAC address. Provides uniqueness but may reveal system information.

UUID v3

A deterministic UUID generated based on a namespace and name using the MD5 hash function.

UUID v4

The most popular version — generated randomly using a cryptographic random number generator.

UUID v5

Similar to v3, but uses SHA-1 instead of MD5. Often used when determinism and greater security are needed.

UUID Structure

Segment	Length	Description	Example
`time_low`	8 characters	The first part of the identifier; in v1, it contains part of the timestamp.	550e8400
`time_mid`	4 characters	The second part of the timestamp.	e29b
`time_hi_and_version`	4 characters	Contains the UUID version.	41d4
`clock_seq`	4 characters	Clock sequence preventing collisions.	a716
`node`	12 characters	Node identifier (e.g., MAC in v1 or random data in v4).	446655440000

Where are UUIDs used

UUID

Databases

UUIDs often replace auto-increment IDs in distributed systems, where multiple servers generate records simultaneously.

UUID

APIs and microservices

Allows for secure identification of resources without revealing the number of records or the database structure.

UUID

Distributed systems

UUIDs allow identifiers to be generated independently on multiple machines without conflicts.

UUID

Tokens and sessions

Often used as a session ID, transaction token, or file identifier.

UUID

Event logging

Helps track individual requests or operations in monitoring systems and logs.

UUID

Data import / synchronization

Facilitates combining data between different systems without identifier conflicts.

Tip: The most commonly used version is UUID v4 because it is fast to generate and does not reveal information about the system or generation time.

Frequently Asked Questions (FAQ)

Is a UUID really unique?

A UUID does not guarantee absolute mathematical uniqueness, but the probability of collision is extremely low. In version v4, there are approximately 3.4×10^38 possible combinations, which in practice eliminates the risk of duplicates.

Why is a UUID better than an auto-increment ID?

UUIDs can be generated independently on multiple servers without a central database. This makes them excellent for microservices architecture and distributed systems.

Is a UUID safe to use in a URL?

Yes. A UUID does not reveal the order of records or the number of users, so it is safer than incremental identifiers in public APIs or links.

Why is a UUID so long?

A UUID has 128 bits to provide a massive identifier space and minimize the risk of collisions even in very large systems.

Does a UUID affect database performance?

It can. A UUID as a primary key tends to be slower than integers because it is larger and less ordered. Many systems use UUIDs in binary format or so-called UUIDv7/ULID for better indexing.

Can data be read from a UUID?

It depends on the version. UUID v1 contains time and device MAC information, while UUID v4 is completely random and contains no readable data.

UUID Generator

What is a UUID and what is it used for?

UUID Structure

Where are UUIDs used

Frequently Asked Questions (FAQ)