Disks and Files | Database Systems

cs186berkeley.net · 2,758 words · saved by 1 readers

Whenever a database uses data, that data must exist in memory. Accessing this data is relatively fast, but once the data becomes very large, it becomes impossible to fit all of it within memory. Disks are used to cheaply store all of a database’s data, but they incur a large cost whenever data is accessed or new data is written. The basic API for disks includes READ and WRITE which stands for transferring “pages” of data from disk to RAM and transferring “pages” of data from RAM to disk respectively. Note that both API calls are very slow due to the structure of magnetic disks. Platters usually spin at around 15000 rpm. The arm assembly moves in or out to position a head on a desired track which is under heads and makes a “cylinder”. Only one head reads/writes at any one time The block/page size is a multiple of (fixed) sector size. Times to access (read/write) a disk block are listed as below: • seek time (moving arms to position disk head on track); 2-3 ms on average • rotational de

Disks and Files - Database Systems Disks and Files | Database Systems Search Menu Expand Document Database Systems Save this note as PDF Memory and Disk Whenever a database uses data, that data must exist in memory. Accessing this data is relatively fast, but once the data becomes very large, it becomes impossible to fit all of it within memory. Disks are used to cheaply store all of a database’s data, but they incur a large cost whenever data is accessed or new data is written. Disk API The basic API for disks includes READ and WRITE which stands for transferring “pages” of data from disk to

Disks and Files | Database Systems

related reading