Emulex 370 User Manual

Page 4

Improved Serviceability
One of the greatest problems with legacy
loop–based architectures is the ability to quickly
isolate and diagnose problems. This situation
is now eliminated with InSpeed’s built–in
advanced diagnostics and statistics tools.
Problems can now be identified more quickly
and while storage systems are still active to
end users, minimizing system downtime and
lowering field service costs. Furthermore,
InSpeed Embedded Storage Switches deliver
hot–pluggable interface ports that make
removing and replacing drive shelves a simple
effort, as opposed to having to go through
the significant rewiring efforts loop–based
architectures require. Overall, InSpeed–based
storage switches can reduce the total cost
to service storage in the field by as much as
15%, and allow less sophisticated personnel to
perform system diagnosis for faster repair.
Higher System Performance While Scaling
Another primary benefit to storage system
switched back ends is the ability to scale
the number of disk drive spindles behind an
individual storage controller and not encounter
performance degradation. InSpeed Embedded
Storage Switches deliver up to three times the
performance of shared loop architectures as
the storage system scales to higher numbers of
drives. The number of IOPs per MB of storage
can now stay constant, regardless of drive and
capacity scale. Based on a wire speed non–
blocking Crossbar switch core that dynamically
handles multiple simultaneous conversations
at 2 or 1Gb/s performance across all ports, the
switch has a total bandwidth of 80Gb/s. Because

InSpeed delivers multiple direct port–to–port
connections, congestion typically found in shared
loop architectures is eliminated and performance
is made consistent especially as storage systems
scale to additional drives and capacities, and
when storage systems are running advanced
processes like remote mirroring, remote
replication, remote backup or rebuilding activities.
InSpeed Embedded Storage Switches are
packed with standard advanced performance
while scaling enabling features. Fairness and
Prioritization ensure devices communicating
with each other have guaranteed access
to all devices in the system, or explicitly
gives prioritized access over other devices.
And Automatic Trunking fully multiplies
load–balanced bandwidth between InSpeed
Embedded Storage Switches and/or InSpeed
SOCs in fully switched infrastructures and
provides dynamic path failover protection. Up
to 5x performance while scaling improvements
over shared loop infrastructures are possible.
Faster Time to Market
InSpeed Embedded Storage Switches
deliver a 6 to 18 month time to market
improvement over proprietary in–house
developed architectures using port bypass
circuits because InSpeed is a drop–in solution
that has been tested and field deployed by
numerous leading storage system providers.
Testing and validation has been performed
in numerous configurations using InSpeed
storage switches to ensure they are fully
reliable and market ready.

��

��

�

��

��

�

��