eXtremeDB 关于mem_page_size与Disk_page_size

The mem_page_size and disk_page_size elements define the size of memory 
pages for conventional memory and persistent media devices, respectively. The 
mem_page_size can be calculated to optimize performance and will vary 
depending on the application dynamics, but as a rule of thumb, page size should 

be between 80 and 512 bytes; a 100 byte page size would be fairly good in most 

situations.

For all-in-memory databases disk_page_size will be set to zero by 

eXtremeDB, for persistent and hybrid databases both parameters should be set to 
a power of 2 with disk_page_size being at least 8 times the mem_page_size value 
to allow for efficient caching. It may be useful to map the disk_page_size to the 
blocking factor of the persistent media. This is because eXtremeDB performs I/O 
in units of disk_page_size (i.e. if you modify a record on a page that holds five 
records, eXtremeDB will write the entire page), and disks perform I/O in blocks, 

regardless of the size of I/O requested by the database system. In other words, if 

the disk blocking factor is 8192 bytes, the disk will always read or write 8192 
bytes, even if eXtremeDB makes a call to write 4096 bytes because its 
disk_page_size is 4096 bytes. Given that the disk will always read or write 8192 
bytes, eXtremeDB should, also.