[HBZ sharing] InnoDB principle of MySQL

When the primary key is not configured Mysql Of InnoDB how ？

1. Mysql I will use my own rowid A primary key

InnoDB What is the underlying data structure of ？

1. B+Tree

BTree Characteristics (MyISAM)

1. Nonclustered indexes , namely 【 Indexes 】 and 【 Corresponding data 】 Are two separate files .
2. Find the corresponding data , Node key With data below data, It's a 16 It's binary , This data It means that key The corresponding position in the data file , Then I will take this 16 It's binary data Go to the data file to find the corresponding data

B+Tree Characteristics (InnoDB)

1. Clustered index , namely 【 Indexes 】 and 【 Corresponding data 】 It's together , In a file
2. Find the corresponding by index key after , The content of this data is actually the same as key It's together , Just retrieve the data directly , No need to hold data Go to another file to find , This is a little
3. B+Tree The leaf node of key, The primary key id Is ordered , From small to arranged , In other words, it is also very good for range search
4. Range lookup , such as id > 15, Leaves key=15 All the following data are required , because id Are arranged in order ,15 Back node , Must be greater than 15 Of

Why? InnoDB The engine needs a primary key , And it is recommended to use integer ？

1. B+Tree The bottom layer may perform some value comparison operations , At this time, the integer will highlight the advantage , such as id > 15.
2. If you do not use integers , While using uuid, If range query is used later , That can only pass through time , Cannot pass primary key id