-
-
Data warehouse, the nature of the Hive in fact equivalent to the files stored in HDFS already had a two-shot relationship Mysql in order to facilitate the use HQL query management
-
For data analysis, washing, Hive suitable for data analysis and off-line cleaning, high latency
-
Based on HDFS, MapReduce, Hive data is still stored on DataNode, written in HQL statement will eventually be converted into MapReduce code execution
-
-
HBase
-
Database, is stored in the column non-oriented relational database
-
For storing structured and unstructured data table suitable for storing a single non-relational data, not suitable relational query, similar to other operations JOIN
-
Avatar based on the HDFS, data is stored in persistent hFile, stored in DataNode in ResionServer be managed in the region
-
Low latency, access to online business use, the face of a large number of enterprise data, HBase can be linear single table to store large amounts of data, while providing efficient data access speed
-
Hive