Distributed ID solution

Distributed id problem

It's mainly the problems caused by sub database and sub table . After the same business data table , Also ensure the data of different tables id only . Otherwise, business data of the same type , Data taken out id It's all the same , How to distinguish these data ？

Solution

Snowflake algorithm :
Unique generated based on timestamp and machine code id. Basic guarantee id Increasing . But when the server time calls back , There may be duplication id.
In response to this question , Baidu UIDGenerator With meituan Leaf Provides a solution , For business scenarios with high requirements , These two solutions can be used .

UUID
It's guaranteed to be unique , But it wastes too much performance .

redis&&Zookeeper
utilize redis or zookeeper Self increasing characteristics of

summary

The most recommended is the snowflake Algorithm . If not sensitive to time callback , Use the snowflake algorithm