Difference between traditional row storage and (HBase) column storage

Column storage differs from traditional relational databases in that data is stored by rows in a table. One of the important benefits of the column approach is that the entire database is automatically indexed because the selection rules in a query are defined by columns. Aggregated storage of data per field stored by columns can greatly reduce the amount of data read when the query requires only a few fields, and aggregated storage of data for one field, that makes it easier to design better compression/decompression algorithms for this aggregated storage.

The difference between traditional (Oracle) row storage and (Hbase) column storage

Here to write the picture depicta

1, the data is stored in rows

2, the query without indexing uses a lot of I/O

3, the creation of the indexes and materialization of the view requires a lot of time and resources

4, the face of the Query demand, the database must be heavily inflated to meet the performance requirements

Here to write a picture description

1, the data is stored in columns - each column is stored separately

2, the data is the index

3, only access to the query involved in the column - a large reduction in system IO

4, each column is handled by a thread - concurrent processing of queries

5, data types are consistent, data characteristics are similar - efficient compression