Intel RAID Controllers - Best Practices white paper

Intel® RAID Controllers Best Practices White Paper

Revision 1.0

RAID set rebuilds are executed for true drive failures. Error recovery is usually not

offered on desktop drives due to the cost. A desktop drive often responds too slowly to

preserve the RAID and the drive is flagged for replacement. The RAID set then

operates in a slower degraded mode until the drive is replaced. Complete loss of data is

possible if another drive times out while the drive is in a degraded mode. With higher

MTBF and error recovery, enterprise drives offer greater reliability than desktop drives in

the server environment.

Hard drives last longer when used in the application for which they were designed. Desktop

drives often lack workload management to lower thermal stresses, have a lower tolerance for

the normal rotational vibration found in a server environment, are not designed to run 24 hours a

day, 7 days a week; and may fail prematurely when installed in a server.

To get the best performance and avoid drive failures, Intel recommends using enterprise drives

for server applications. SAS and SATA drives should not be mixed in the same enclosure.

Please refer to the Enterprise-class versus Desktop-class Hard Drives white paper available on

the http://support.intel.com/support/motherboards/server/

Web site.