A method and system for generating a unique identifier

Abstract

The invention discloses a method and system for generating a unique identifier. The steps of the generating method include: a number issuer receives a registration identification factor request sent by a number taker and parameter information of the number taker, and records the parameters of the number taker information and send the identification factor to the number obtainer; the number obtainer receives the identification factor sent by the number sender, and the buffer unit of the number obtainer buffers the identification factor; the generation unit of the number obtainer obtains the identification factor from the buffer unit, and according to The identification factor generates a unique identifier. The design principle of the present invention is reliable and the structure is clear and definite, the double-buffer queue adopted ensures high availability, and the generated unique identifier does not depend on time, so it has the prospect of wide application.

Description

technical field [0001] The present invention relates to the technical field of identifiers, in particular to a method and system for generating a unique identifier. Background technique [0002] In a distributed system, it is often necessary to use a globally unique identifier to define corresponding different types of data, such as: unique identifiers for orders, unique identifiers for messages, and unique identifiers for data links. The unique identifier can facilitate indexing, occupies relatively little space, and is convenient for storage. [0003] At present, there are many ways to obtain a unique identifier in the existing technology, and there are mainly the following solutions: [0004] (1) UUID, the implementation of UUID is relatively simple. It generates a series of unordered unique identifiers based on the MAC address and timestamp of the network card, and there is no sorting, so the trend cannot be guaranteed to increase. The storage space is large and the que...

AI Technical Summary

Problems solved by technology

[0004] (1) UUID, the implementation of UUID is relatively simple. It generates a series of unordered unique identifiers based on the MAC address and timestamp of the network card, and there is no sorting, so the trend cannot be guaranteed to increase. The storage space is large and the query efficiency is low;

[0005] (2) The MySQL database is incremented by using MySQL's auto_increment (digital self-increment attribute). To be suitable for a distributed system architecture, at least two MySQL databases are required to generate odd and even numbers by distinguishing the initial value and step size of auto_increment , using the digital self-increment sequence mechanism of the database to improve reliability, but the disadvantage is also obvious, that is, the entire system is unavailable when the database is abnormal;

[0006] (3) The snowflake algorithm (SnowFlake) uses a 64-bit long number as a globally unique identifier; it is widely used in distributed systems, where 1 bit of inapplicable bit indicates positive or negative, and then 41 bits of it are used as The number of milliseconds (the length can be used for 69 years), 10 bits are used as the working machine identification (up to 1024 nodes are deployed in the branch), and 12 bits are used as the serial number (it is the counting sequence number, and each node can generate 4096 identification numbers per millisecond) ), the advantage is that it is stored in the database in an orderly manner according to time, the indexing efficiency is high, and the generation does not depend on the database. The system time is called back or changed, which may cause id conflicts or duplication

Images