The original article titled HowSAPHANAIsSuchaFastDatabase, however, the author's opinion is that HANA's speed mainly stems from the development of hardware, and HANA is not suitable for all application scenarios. However, what I pay attention to is what is beyond the conclusion. Changes in data access modes improve storage hardware, from physical disks to SSD, memory, and corresponding database query methods
The original article titled How sap hana Is Such a Fast Database, but the author's opinion Is that HANA's speed mainly stems from the development of hardware, and HANA Is not suitable for all application scenarios. However, what I pay attention to is what is beyond the conclusion. Changes in data access modes improve storage hardware, from physical disks to SSD, memory, and corresponding database query methods
The original article titled How sap hana Is Such a Fast Database, but the author's opinion Is that HANA's speed mainly stems from the development of hardware, and HANA Is not suitable for all application scenarios.
However, what I pay attention to is what is beyond the conclusion.
Changes in Data Access Modes
The increase in storage hardware, from physical disks to SSD, memory, and the corresponding database query methods have also changed.
When a database uses a traditional physical disk, if you read 3 columns from a table with 10 columns, you must also read all 10 columns.
If the table is 100 mb and 1 million rows, you must read the entire table to complete the query. When multi-table Join is involved, the situation is more serious.
If the database is stored in SSD or memory, you only need to read three columns if you need three columns.
Elimination of data exchange
Traditional databases stored on disks need to spend a lot of time on Swapping, which reads data into the cache, and then writes the data back to the disk after processing ,:).
Memory databases eliminate this kind of data exchange (Swapping). Because all data is operated in the memory, the only purpose of storage is to ensure data persistence and backup.
In fact, this explains that even if all the data is cached to the memory, it is still not as fast as the memory database, whether it is a column-based database such as HANA or a row-based database such as TimesTen.
Conclusion
A column-based database has a speed advantage, but it is Not a omnipotent advantage. Although SAP wants to use HANA as an integrated and applicable platform, One Size Does Not Fit All is the principle of this database, HANA is very suitable for analytical applications such as BW. For OLTP applications such as ERP, we will wait and see how S/4 HANA works.
Reference
- How sap hana Is Such a Fast Database-SCM Focus 'sap HANA Blog