RAID

Aug 21, 2025

Overview

This lecture explains RAID (Redundant Array of Independent Disks), covering different RAID levels, how they provide data redundancy and performance, and clarifies that RAID is not a form of data backup.

Introduction to RAID

  • Hard drives store large amounts of data but can fail due to moving parts.
  • RAID is used to maintain data availability and uptime when a drive fails.
  • RAID is not a backup method; it's for redundancy and system availability.
  • RAID stands for Redundant Array of Independent (formerly Inexpensive) Disks.

RAID Levels Overview

  • RAID levels refer to different ways drives can be combined for redundancy or performance.
  • Not all RAID levels provide redundancy, so choose carefully when setting up RAID.

RAID 0 (Striping)

  • RAID 0 splits data across at least two drives, called striping.
  • This increases performance since each drive stores part of the data.
  • No redundancy: if one drive fails, all data is lost.
  • RAID 0 provides zero redundancy.

RAID 1 (Mirroring)

  • RAID 1 duplicates the data on two drives, creating exact copies.
  • Uses twice the storage space since data is duplicated.
  • Provides redundancy: if one drive fails, data remains available on the other.

RAID 5 (Striping with Parity)

  • RAID 5 distributes data across multiple drives with one drive used for storing parity information.
  • Parity lets you rebuild data if a single drive fails.
  • More efficient use of storage compared to mirroring.
  • Parity is spread across all drives, improving recovery.
  • Rebuilding data from parity can cause a performance hit due to CPU processing.

RAID 10 (RAID 1+0 / Stripe of Mirrors)

  • Combines RAID 0 (striping) and RAID 1 (mirroring).
  • Data is striped across multiple drives, and each stripe is mirrored.
  • Provides high performance and redundancy.
  • System remains available even if multiple drives fail, as long as one from each mirrored pair survives.

Key Terms & Definitions

  • RAID — Redundant Array of Independent Disks; combines multiple drives for redundancy or performance.
  • Redundancy — Having extra copies or ways to recover data if hardware fails.
  • Striping — Splitting data across multiple drives to increase performance.
  • Mirroring — Duplicating data across drives for redundancy.
  • Parity — Extra calculated data used to recover lost information after a drive failure.

Action Items / Next Steps

  • Identify the RAID level suitable for your needs before setup.
  • Remember to implement a separate backup strategy, as RAID is not a backup solution.

Full transcript