Note:
This topic has been translated from a Chinese forum by GPT and might contain errors.Original topic: tidb 分区表
|
username: Raymond
TiDB already has a distributed architecture, and the stored data has already been scattered. What is the greatest benefit of using partitioned tables in this case?
Facilitates operations management, such as partitioning by time range. To delete historical data, you only need to drop the partition, which is much more efficient than deleting it.
I don’t know if there is any difference in management convenience and access efficiency between TiDB partition tables and MySQL partition tables.
In fact, partitioning in other databases is also mainly used for data lifecycle management. Partitioning itself does not cause a significant difference in read/write performance, although there may be a slight difference.
The main purpose is for archiving and managing data. This is the biggest feature. After all, delete does not release space.
Indeed, MySQL partition tables are the same, making management easier and avoiding large transactions.
It depends on the ultimate purpose of using partitioned tables. If it is for archiving to manage data under certain rules, it is quite good.
Oracle is not just for the convenience of later maintenance. If the data volume of the table is particularly large, creating partition tables with corresponding partition keys for specific business scenarios can significantly improve query efficiency.
Besides ease of management, are there also performance improvements? Because regions of 96M~144M are already scattered, similar to partitions, right?
This topic was automatically closed 1 minute after the last reply. No new replies are allowed.
|
username: 
|
username: 
|
username: 
|
username: 
|
username: 
|
username: 
|
username: