Apache hadoop sequencefile provides a way to save data to HDFS in the form of binary key-value pairs. And other key-Value
Compared with data structures (such as B-tree), sequencefile cannot modify or delete data, nor insert data into data. Sequencefile can only end
Data.
Sequencefile provides three storage formats:
1. uncompressed format: no compression format 2. Record compressed: Record compression format 3. Block compressed: block compression format
The headers of the three storage formats are the same. Their headers contain the following information, which will be read by sequencefile. Reader.
1. the header part includes the class type of key and value, sequencefile. reader will read these two attributes and instantiate them through reflection; 2. version includes the version information of sequencefile; 3.is compressed is used to specify whether it is record compressed; 4.is block compressed is used to specify whether it is block compressed
5. Metadata is a key-Value List structure that can be used to store metadata of files. Key-value pairs can only be of the <text, text> type. And the information
It is written in the construction method of sequencefile. Writer. Cannot be modified.
6.
Not complete...