Brocade Communications Systems SilkWorm 2100, Reference

Page: 10 / 32

1-2
SilkWorm 2100 Reference
Overview
1
The SilkWorm 2100 Advantage
The SilkWorm 2100 switch offers advantages in the following areas:
• Performance
• Reliability
• Manageability
• Flexibility
Performance
The SilkWorm 2100 switch offers enhanced performance:
Data Transfer A device arbitrates only among the devices in its looplet; frames are
transferred to the switch for forwarding to a destination device on
another looplet.
The non-blocking architecture of the SilkWorm 2100 switch
guarantees full-speed data delivery in all traffic conditions.
Cut-through routing ensures a maximum latency of 2 microseconds
from switch port to switch port, allowing parallel data transfers of
up to eight (one per loop) simultaneous transactions at a time.
Whereas traditional hub-based arbitrated loops are limited to a total
shared bandwidth of 100 MB/sec, the SilkWorm 2100 switch is
capable of transferring data at eight times (800 MB/sec) that rate.
Buffering Unlike a traditional loop environment where a device is blocked
from transferring data when another device on the loop has control,
the SilkWorm 2100 switch can buffer data within the switch
allowing a device to transfer multiple frames. This allows for
overlapping transfers and improves the overall throughput of
the SAN.
Reliability
The SilkWorm 2100 switch provides enhanced reliability:
Port Isolation The SilkWorm 2100 switch monitors each looplet to ensure proper
device operation. If errors are detected the looplet is not allowed to
initialize and the port is taken out of service. An alert is generated
and the port is monitored until the error condition is cleared. This
enables the loop to operate while the device is isolated and repaired,
thereby preventing a failing device from disabling the entire loop.
Status Reporting The SilkWorm 2100 switch provides extensive statistics, including
a full set of performance and error counters, to aid in diagnosis and
isolation of malfunctioning ports. These statistics are retrieved via
several interfaces such as SNMP, telnet, and an optional Web-based
management interface feature, WEB TOOLS.
Diagnostics Power On Self Tests (POSTs) and online diagnostics monitor
individual ports while the switch is running. Offline diagnostics
isolate problems enabling faster restoration of service.
Hardware Features Redundant fans and a hot-pluggable redundant power supply
option eliminate service down time related to failing power
supplies or fans.