Difference between traditional relational database and columnar database

Columnar or column-based storage (Columnar or column-based) is relative to row-based storage (Row-basedstorage) of traditional relational databases. Simply put, the difference between the two is how to organize the table.1) Row storage writes are done once. If this write is built on the operating system's file system, the success or failure of the write process can be guaranteed and the integrity of the data can therefore be determined.

2) Column storage due to the need to split a row of records into a single column to save, the number of writes is significantly more than row storage (meaning that the number of times the magnetic head scheduling, and the magnetic head scheduling is time-consuming, usually in 1ms ~ 10ms), plus the head needs to be moved on the disk and positioning time spent, the actual time consumption will be even greater. Therefore, row storage has a great advantage in writing.

3) There is also data modification, which is actually a write process. The difference is that data modification is to mark the record on disk for deletion. Row storage is to write once in a specified location, column storage is to locate the disk to multiple columns to write separately, this process is still row storage columns times. So, data modification is also dominated by row storage.