BOLTDB Source Analysis-Data structure-1

This is a creation in Article, where the information may have evolved or changed.

BOLTDB Data structure

boltdbThe data exposed to the user is less conceptual, with only the following:

• Optionsboltdbthe relevant configuration selection at initialization;
• DBThe entire boltdb holder, and boltdb related operations are to be initiated by invoking its method, which is boltdb an abstraction;
• StatsDB.Stats()the data structure returned by the calling method contains some boltdb internal count information that can be viewed by the user;
• BucketSimilar to 表 a concept, in the BOLTDB related data must exist Bucket under a different Bucket data isolation, each under a Bucket monotonically increasing Sequence , similar to the self-incrementing primary key;
• BucketStatsBucketSome of the statistical information;
• TxboltdbTransaction data structure, in boltdb which all the operations on the data must occur within a transaction, and boltdb the concurrent reads are implemented here;
• Cursoris an iterator to the internal b-tree, which is used to traverse the database, provide First / Last / Seek and other operations;

There are also internal data structures that help implement some internal logic:

• node is used to store each bucket Part of Key-value, each bucket inode records the data structure of the key-value pair, each node contains a inode array, inode is a k-v memory-cached record that is recorded as leafpageelement ;
• leafpageelement disk to record the specific Key-value index;
• page user records important information for each region in the persisted file, while page is divided into a variety of Code class= "Highlighter-rouge" >page stores different data information;