Ultra3 -to- Ultra3
External RAID Controller
Overview
The newest member of the NS-RC series
is designed with Infortrend's 133MHz chipsets and PowerPC
750CXe CPU for fast, high-throughput RAID operation. I/Os
are rapidly processed and distributed to array hard drives
via a high-speed 64-bit data path in the maximum bandwidth
of 1056MB/second. The result is a controller ready for
future applications that require both high sustained data
rates and fast I/O turnaround.
High Flexibility
All SCSI channels can be defined as either host or drive.
This allows integrators a flexible deployment for connecting
multiple host systems through independent channels. Firmware
upgrade can be performed on single or redundant controllers
during active services. Controllers can shift I/Os temporarily
allowing firmware on both controllers to be updated. In
the unlikely event of controller failure, a failed unit
can be replaced with one running later version of firmware.
64 array partitions and 128 LUN's per controller are
supported.
High Data Availability
The controller's advanced capabilities ensure the highest
level of availability and fault tolerance. The capabilities
include: automatic drive failure detection and background
rebuild, bad sector reassignment, auto failback/failover,
multi-level RAID, to name a few.
When the controller detects a hard drive failure in
a RAID 1, 3, or 5 logical drive, the disk array instantly
replaces the failed disk with a hot-spare. Data is reconstructed
into the hot-spare drive without interference to the
host. In the absence of a hot-spare drive, users can
hot-swap the failed drive with a new one and an automatic
rebuild will begin.
The battery backup option protects write-back cache
data from sudden power interruptions. The battery supplies
auxiliary power to maintain cached data for up to 72
hours.
Bad Media Management
The controller is capable of detecting bad sectors on
SCSI drives during disk reads or writes, recovering
the data from parity fault, and performing bad sector
reassignments. Bad sector management is fully transparent
to host.
The controller supports active-active redundancy with
a synchronized write-back cache.
Each controller has a dedicated high-bandwidth channel
for an exact replica of the counterpart's cached data.
I/Os belonged to a failed controller can be automatically
transferred to the surviving controller, and vice versa
to the replacement making failover and failback totally
transparent to host.
Zero Down-Time Technology
The controller is designed from top to bottom to be
hot-swappable. The docking connectors and special canister
mechanism ensure joining and separating main board free
of signal glitches and power surge.
With automatic failover/failback, auto rebuild, synchronized
cache, disk array operation is thoroughly protected
from possible causes of downtime.
Board-Only Solutions for OEMs
NS-RC controllers are the ideal solutions for
OEM customers who design their own backplanes. The controller
board can be readily integrated with a cableless and
hot-swappable enclosure design via the connection of
all supported signals over docking connectors.
|
