KSUID

KSUID, short for K-Sortable Unique Identifier, is a globally unique identifier designed for distributed systems that require local generation without coordination and with time-based ordering. A KSUID embeds creation time in the identifier so that IDs produced earlier sort before those produced later, facilitating time-based querying and ordering.

The KSUID format is 20 bytes: a 4-byte timestamp and a 16-byte random payload. The timestamp is a big-endian unsigned integer representing the number of seconds since a fixed epoch (commonly May 13, 2010 UTC). The remaining 16 bytes provide a cryptographically strong random value to ensure uniqueness. The 20-byte binary value is encoded as a 27-character base62 string, which yields a URL-safe, human-friendly representation.

Key properties include time-sortability, high probability of global uniqueness, and collision resistance. Because the timestamp occupies the most significant bytes, KSUIDs generated at earlier times sort before those generated later when compared as strings or binary data. Within the same second, the 128-bit random payload provides a wide space to minimize collisions. KSUIDs are well suited as database keys, log identifiers, event IDs, and other distributed system identifiers where order by creation time is beneficial.

There are libraries implementing KSUID in multiple languages, including Go, Python, JavaScript, and Java, often with bindings to the official design. The approach contrasts with other sortable ID schemes in encoding choices and timestamp granularity, but shares the goal of providing compact, sortable, globally unique identifiers for distributed applications.