How Oracle creates a self-increasing primary key
Oracle primary keys are commonly divided into UUID and There are two types of self-increasing int. Let’s briefly talk about their respective advantages and disadvantages:
Advantages of UUID
1. It is easy to generate, whether it is through UUID can be easily created using sys_guid() or java's uuid.
2. Suitable for insert and update operations in batch data.
3. Cross-server data merging is very convenient.
Advantages of INT self-increasing
1. Small space occupied
2. Good performance, UUID is not on the same level as int
3. Easy to remember
Their respective advantages are each other’s shortcomings
Recommended "mysql video tutorial"
Applicable Scope:
Generally, UUID is used as the only primary key in a distributed environment. As for other projects, I strongly recommend using int as the primary key.
PS: Recently, one of my distributed projects also had UUID duplication during the peak weekend access period, so I did not write down the advantage of UUID as a globally unique ID, but we can use Zookeeper as the primary key to generate it. The operator creates a unique primary key of type int, which is completely fine.
So to sum up, I still recommend that all primary keys should use the int type. The advantages of the int type are particularly obvious when it comes to tens of millions of data.
Oracle steps to create self-increasing:
Create self-increasing sequence
-- 创建自增长序列 create sequence seq_on_chance_contract increment by 1 -- 每次加1 start with 1 -- 从1开始计数 nomaxvalue -- 不设置最大值 nocycle -- 一直累加 nocache;
Create data table
The primary key can Use INTEGER and NUMBER. If the system uses Hibernate as the ORM framework, it is not recommended to use NUMBER
create table TBL_CHANCE_CONTRACT ( ID INTEGER PRIMARY KEY, CHANCE_SUCCESS_ID VARCHAR2(50) not null, CONTENT CLOB, CREATE_USER_ID VARCHAR2(50), CREATA_USER_NAME VARCHAR2(80), CREATE_DATE TIMESTAMP(6) )
Create a primary key trigger so that you do not need to manage the ID primary key when adding new records
-- 创建主键触发器 CREATE OR REPLACE TRIGGER tg_on_id_chance_contract BEFORE INSERT ON TBL_CHANCE_CONTRACT FOR EACH ROW WHEN (new.ID is null) BEGIN SELECT seq_on_chance_contract.nextval into:new.ID from dual; END;
Some tips ：
View all sequences and tables
-- 查找所有SEQUENCE、TABLE select * from user_objects ubs; -- 查找所有SEQUENCE select * from user_objects ubs where ubs.OBJECT_TYPE='SEQUENCE';
Check the current sequence number
select seq_on_chance_contract.nextval from dual; select seq_on_chance_contract.currval from dual;
Reset sequence
General reset The sequence needs to be deleted and then rebuilt, but this is more troublesome. There are two ways to do it without deleting the sequence:
1. Use the step parameter to first find out the nextval of the sequence. Remember, put Incrementally change the value to negative (and vice versa), and then change it back again.
Assume that the sequence name that needs to be modified is seq_on_chance_contract
-- 重置序列 select seq_on_chance_contract.nextval from dual; -- 假设得到结果 n alter sequence seq_on_chance_contract increment by -2; -- 注意是 -(n-1) select seq_on_chance_contract.nextval from dual; -- 再查一遍，走一下,重置为1了 alter sequence seq_on_chance_contract increment by 1; -- 还原
2. Use the stored procedure to implement (v_seqname)
create or replace procedure seq_reset(v_seqname varchar2) as n number(10); tsql varchar2(100); begin execute immediate 'select '||v_seqname||'.nextval from dual' into n; n:=-(n-1); tsql:='alter sequence '||v_seqname||' increment by '|| n; execute immediate tsql; execute immediate 'select '||v_seqname||'.nextval from dual' into n; tsql:='alter sequence '||v_seqname||' increment by 1'; execute immediate tsql; end seq_reset;
Then call the stored procedure:
exec seq_reset('v_seqname');
This article comes from PHP Chinese website, mysql tutorial column, welcome to learn!
The above is the detailed content of How to create an auto-increasing primary key in Oracle. For more information, please follow other related articles on the PHP Chinese website!