Introduction

Twitter's Snowflake is a ID generation scheme that tackles all of the requirements below:

ID fits under 64 bits
ID will be used with distribution in mind (horizontal scale SQL, Cassandra, etc.)
Roughly sortable (challenging in a distributed system)

Scheme

The composition is as follows:

\([timestamp] - [worker] - [sequence]\)

Timestamp (41 bits): Millisecond precision
Worker Number (10 bits): Also known as the machine ID, in a distributed database we may have many, many workers. Each number represents a unique machine. Worker numbers can be chosen at startup with something like Zookeeper
Sequence Number (12 bits): This can be thought of as the transaction ID on the worker machine itself. Therefore, this number is based on a per-thread basis. Keep in mind that 12 bits allows for 4096 entries (\(2^{12} = 4096\)), so the sequence number should naturally reset back to 0 once it passes the 4096 mark.