Advertisement
Advertisement
UUID Versions Explained
A UUID (Universally Unique Identifier) is a 128-bit number in the format 8-4-4-4-12 hex characters. There are 5 versions, each generated differently for different use cases.
UUID Versions at a Glance
- Version 1: time-based + MAC address. Contains creation timestamp. Used in Cassandra databases
- Version 4: fully random. Most commonly used. ~5.3 × 10²⁶ possible values — collision probability is negligibly small
- Version 5: namespace + SHA-1 hash. Deterministic — same input always produces the same UUID
- Version 7 (modern): time-ordered random UUID — designed for sortable database primary keys
When to Use UUIDs
- Database primary keys in distributed systems where sequential IDs conflict
- API resource identifiers that should be globally unique
- Session tokens, file names, or reference numbers
- Anywhere you need uniqueness across multiple systems without coordination
Can two UUIDs ever be the same?
Theoretically yes, but the probability is astronomically small. UUID v4 has 2¹²² possible values (about 5.3 × 10³⁶). If you generated 1 billion UUIDs per second for 100 years, the probability of a collision would still be less than 1 in 10 billion. In practice, UUIDs are treated as globally unique for all engineering purposes.
Should I use UUID or ULID for database primary keys?
UUID v4 is random and causes index fragmentation in B-tree databases because new inserts can't be appended in order. ULID and UUID v7 solve this by embedding a timestamp prefix, making inserts sort naturally and improving index performance. For high-performance databases with heavy write loads, ULID or UUID v7 are preferred over v4.