IBM Enterprise Storage Server 2105 E10, Введение и руководство по планированию

Страница: 38 / 156

Note: For updated and additional information not included in this publication, see
the README file on the IBM Subsystem Device Driver compact disk
recordable (CDR) or visit the Subsystem Device Driver Web site at:
www.ibm.com/storage/support/techsup/swtechsup.nsf/support/sddupdates/
IBM Enterprise Storage Server availability features
This section describes the following enhanced data availability features that the
ESS offers for hosts with UNIX, OS/400, Windows NT, and Windows 2000
operating systems and for System/390 hosts:
v
Redundant array of independent disks (RAID)
v
Predictive failure analysis
v
Failover protection
v
Sparing
v
Fibre-channel support (on supported host systems)
v
Storage area network (SAN) support
v
Cache
v
Nonvolatile storage
v
Concurrent maintenance
v
FlashCopy (not supported for OS/400)
v
Peer-to-Peer Remote Copy (not supported for OS/400)
v
IBM Subsystem Device Driver
This section also describes data availability features that are available only for:
v
Open host systems. See “UNIX, Windows NT, and Windows 2000 availability
features” on page 18.
v
System/390 hosts. See “System/390 availability features” on page 19.
Enhanced data availability for some host systems is also available with the following
software. The software automatically uses a secondary path (from a secondary
server) if a failure occurs.
v
HACMP, for mirroring, for RS/6000 host systems
v
Similar software for other non-IBM host systems
v
System/390 hosts running TPF can use routines that ensure volume redundancy
Redundant array of independent disks implementation
The IBM Enterprise Storage Server protects all storage with redundant arrays of
independent disks (RAID). An array is an ordered collection of physical devices
(disk drives) that are used to define logical volumes or devices.
The ESS attaches disk drive modules (DDMs) in RAID-5 configurations. The
implementation of RAID-5 configurations distributes (stripes) parity across all DDMs
in the array. The ESS requires two spare DDMs per loop. With two spares, a loop
can have two DDM failures and still allow you to access data.
Implementation of RAID-5 data striping supports independent accesses to the
DDMs. The dual-loop device interface card provides the RAID-5 protection on the
ESS.
All DDMs on a single loop must be of the same capacity.
14
ESS Introduction and Planning