Cache manager

As I already said, the main bottleneck of databases is disk I/O. To improve performance, modern databases use a cache manager.

Instead of directly getting the data from the file system, the query executor asks for the data to the cache manager. The cache manager has an in-memory cache called buffer pool. Getting data from memory dramatically speeds up a database. It’s difficult to give an order of magnitude because it depends on the operation you need to do:

• sequential access (ex: full scan) vs random access (ex: access by row id),
• read vs write

and the type of disks used by the database:

• 7.2k/10k/15k rpm HDD
• SSD
• RAID 1/5/…

but I’d say memory is 100 to 100k times faster than disk.

But, this leads to another problem (as always with databases…). The cache manager needs to get the data in memory BEFORE the query executor uses them; otherwise the query manager has to wait for the data from the slow disks.