Advertisement

Disk and Storage System Basics

  • Volker Herminghaus
Chapter

Abstract

Disk media are the entities that all persistent user data is eventually stored on. Because the surface of a disk medium can be permanently magnetized, disks can store information across reboots and power failures, when data residing in the computer's internal volatile memory is lost. Disks can not be replaced by any amount of volatile memory. After all, where would you put all that data after a shutdown? But a transition is slowly getting under way: A few months before work on this book was begun, Apple Inc. released a notebook computer that did not have a disk drive but used flash memory instead. EMC, a vendor of mass storage systems, announced a storage array that used flash. These events marked the beginning of a trend away from moving macroscopic mechanical spindles for storing data - an incredibly arcane concept when compared to light-based fibre-channel communications and memory cells holding only a few dozen electrons per bit.

However, flash is still much more expensive than disk storage, and even with prices falling and some problematic properties of flash being alleviated, disk based storage systems will be here for a long time. They will eventually be found at the back end of the storage chain, similar to tape reels in the times of the old mainframe computers. Disk storage will still need to be managed, and volume management software will still do that job. Emphasis will likely be shifting from performance towards reliability, as more people become aware of the fact that with the amount of data processed today, data errors will be a frequent problem very soon. Error rates looked extremely low a few years ago, but when multi-terabyte databases are processed at high speed around the clock, the seemingly low probability for errors that slip through all error checking and prevention mechanisms soon turns to certainty.

Keywords

Hard Disk File System Random Access Memory Device Driver Deadlock Avoidance 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer-Verlag Berlin Heidelberg 2009

Authors and Affiliations

  1. 1.Germany

Personalised recommendations