The Dell™ PowerEdge™ Expandable RAID Controller (PERC) 5 family of controllers offers redundant array of independent disks (RAID) control capabilities. The PERC 5 Serial Attached Small Computer System Interface (SAS) RAID controllers support SAS devices and Dell-qualified SATA devices. The controllers provide reliability, high performance, and fault-tolerant disk subsystem management.
Scope of the User's Guide
This user's guide for the PERC 5 controllers documents the following topics:
Basic information about the PERC 5 controllers and RAID functionality
Information about PERC 5 controller features
Hardware installation and battery management
Installation procedures for operating system drivers
RAID configuration and management
Troubleshooting information
Regulatory information and notices
PERC 5 Controller Descriptions
The following list includes a description of each type of controller:
The PERC 5/E Adapter with two external x4 SAS ports and a transportable battery backup unit (TBBU)
The PERC 5/i Adapter with two internal x4 SAS ports with or without a battery backup unit, depending on the system
The PERC 5/i Integrated controller with two internal x4 SAS ports and a battery backup unit
Each controller supports up to 64 virtual disks. In addition, PERC 5/E controllers provide two ports that can connect to up to three enclosures each, for a total of six enclosures per controller. Each enclosure can contain up to 15 physical disks, meaning a controller can support up to 90 physical disks total in the six enclosures.
NOTE: PERC 5/i is limited by the configuration supported on the platform.
PCI Architecture
PERC 5 controllers support a PCI-E x8 host interface. PCI-E is a high-performance I/O bus architecture designed to increase data transfers without slowing down the Central Processing Unit (CPU). PCI-E goes beyond the PCI specification and is intended as a unifying I/O architecture for various systems such as, desktops, workstations, mobiles, server, communications, and embedded devices.
Operating System Support
The PERC 5 controllers support the following operating systems:
Microsoft® Windows® 2000 Server family
Windows Server® 2003 (includes Standard, Enterprise, and Small Business Servers)
Windows Server 2003 DataCenter
Windows XP
Windows Vista™
Red Hat® Enterprise Linux® 3, Red Hat Enterprise Linux 4, and Red Hat Enterprise Linux 5
SUSE® Linux Enterprise Server 9 and SUSE Linux Enterprise Server 10
NOTE: See driver release on www.dell.com for specific operating system service pack requirements.
NOTE: Microsoft Windows XP is supported with a PERC 5 controller only when the controller is installed in a Dell Precision™ workstation.
NOTE: See the system documentation located at the Dell Support website at support.dell.com for the latest list of supported operating systems and driver installation instructions.
RAID Description
RAID is a group of multiple independent physical disks that provide high performance by increasing the number of drives used for saving and accessing data. A RAID disk subsystem improves input/output (I/O) performance and data availability. The physical disk group appears to the host system as either a single storage unit or multiple logical units. Data throughput improves because several disks are accessed simultaneously. RAID systems also improve data storage availability and fault tolerance. Data loss caused by a physical disk failure can be recovered by rebuilding missing data from the remaining data or parity physical disks.
NOTICE: In the event of a physical disk failure, you cannot rebuild data on a RAID 0 virtual disk.
Summary of RAID Levels
RAID 0 uses disk striping to provide high data throughput, especially for large files in an environment that requires no data redundancy.
RAID 1 uses disk mirroring so that data written to one physical disk is simultaneously written to another physical disk. This is good for small databases or other applications that require small capacity, but complete data redundancy.
RAID 5 uses disk striping and parity data across all physical disks (distributed parity) to provide high data throughput and data redundancy, especially for small random access.
RAID 10, a combination of RAID 0 and RAID 1, uses disk striping across mirrored spans. It provides high data throughput and complete data redundancy.
RAID 50, a combination of RAID 0 and RAID 5, uses distributed data parity and disk striping and works best with data that requires high system availability, high request rates, high data transfers, and medium-to-large capacity.
RAID Terminology
Disk Striping
Disk striping allows you to write data across multiple physical disks instead of just one physical disk. Disk striping involves partitioning each physical disk storage space into stripes that can vary in size ranging from 8 KB to 128 KB, often referred to as stripe size. These stripes are interleaved in a repeated sequential manner. The part of the stripe on a single physical disk is called a strip.
For example, in a four-disk system using only disk striping (used in RAID level 0), segment 1 is written to disk 1, segment 2 is written to disk 2, and so on. Disk striping enhances performance because multiple physical disks are accessed simultaneously, but disk striping does not provide data redundancy.
With mirroring (used in RAID 1), data written to one disk is simultaneously written to another disk. If one disk fails, the contents of the other disk can be used to run the system and rebuild the failed physical disk. The primary advantage of disk mirroring is that it provides 100 percent data redundancy. Because the contents of the disk are completely written to a second disk, it does not matter if one of the disks fails. Both disks contain the same data at all times. Either of the physical disks can act as the operational physical disk.
Disk mirroring provides 100 percent redundancy, but is expensive because each physical disk in the system must be duplicated. Figure 1-2 shows an example of disk mirroring.
Parity creates a set of redundant data from two or more parent data sets. You can use the redundant data to rebuild one of the parent data sets. Parity data does not fully duplicate the parent data sets but that data can be used to reconstruct the data if lost. In RAID, this method is applied to entire physical disks or stripes across all the physical disks in a physical disk group.
The parity data is distributed across all the physical disks in the system. If a single physical disk fails, it can be rebuilt from the parity and the data on the remaining physical disks. RAID level 5 combines distributed parity with disk striping, as shown in Figure 1-3. Parity provides redundancy for one physical disk failure without duplicating the contents of entire physical disks. However, parity generation can slow the write process.
Figure 1-3. Example of Distributed Parity (RAID 5)
NOTE: Parity is distributed across multiple physical disks in the disk group.
