The value of a data-storage system should not be underestimated. This article will explain how tech company use data storage in a professional context.

Hard drive: The basic storage component

Data storage starts with a hard drive. Hard drives are classified by various factors, including storage size, reading and writing speeds, and latency. Latency is the time it takes the computer to start to read or to write a certain data block.

The basic hard drive types are classic and SSD. A classic drive uses a magnet. An SSD (Solid State Drive) uses flash memory, which has better speed and latency than a classic drive.

RAID principle

As we talk about professional needs, the first point that’s come to mind is the reliability of a storage device. The reliability of a storage device can be enhanced with the RAID (Redundant Array of Independent/Inexpensive Disks) principle. The principle involves saving data in parallel on a number of hard drives to increase the writing and reading speeds and overall reliability. Such a system ensures data are not lost when one drive crashes.

Disk arrayHard drives usually are stored in a disk array, which is a hardware element that goes into data centers. Disk arrays also are used in:

  • Multiple power supply equipment, which ensures a steady power supply, regardless if part of the system fails.
  • Inverter, which provides battery power until power can be provided by other equipment, such as a fuel-powered generator.

Cache battery, which is a small piece of electronic equipment that allows the disk array that gives a brief boost to the amount of data a hard drive usually is able to write. The equipment fosters high performance despite hard drive limitations.Disk arrays might be expensive, but there is an advantage: each component can be changed out without turning off the system. This is the only physical way to get high availability for your data.

SAN (Storage Area Network)

There are two main ways of using a disk array. The first is SAN, or Storage Area Network. SAN, in effect, splits the disk array into blocks and attaches each block to a computer, usually a server. The blocks are recognized as a hard drive. The interesting part in this architecture is that the connections between the server and the disk array are redundant too, usually with two optical fibers.

NAS (Network Area Storage)

The second way of using a disk array is via the NAS principle, which is a bit more involved than the SAN principle. With NAS, there is not just one computer/server that accesses a data block, but many computers that access a shared file system. Application of the principle becomes clear in database contexts and services.

Which to chose?Managing server infrastructure has become a specialty, with many existing solutions of varying costs. A large company with hundreds servers may decide on creating an in-house data center, but there are drawbacks: technical issues and not being as competitive as companies that specialize in data storage.

Any questions?

This article was more an introduction on how things work at your hosting provider. A future article will discuss how to examine your data-storage needs in determining whether to use online storage or a hosting company.