Brocade Communications Systems SilkWorm 12000 Programming manual

Publication Number 53-0000251-02
Brocade SilkWorm 12000
Design, Deployment and
Management Guide

Copyright ©2002, Brocade Communications Systems, Incorporated.
ALL RIGHTS RESERVED. Publication Number 53-0000251-02
BROCADE, the Brocade B weave logo, Brocade: the Intelligent Platform for Networking Storage, SilkWorm, and
SilkWorm Express, are trademarks or registered trademarks of Brocade Communications Systems, Inc. or its
subsidiaries in the United States and/or in other countries. All other brands, products, or service names are or may be
trademarks or service marks of, and are used to identify, products or services of their respective owners.
Notice: This book was designed and written to provide information about storage area networking architectures. Every
effort has been made to make this book as complete and accurate as possible. However, the information in this book is
provided to you “AS IS,” without warranty of any kind, including, without limitation, any implied warranty of
merchantability, noninfringement or fitness for a particular purpose. Disclosure of information in this material in no way
grants a recipient any rights under Brocade's patents, copyrights, trade secrets or other intellectual property rights.
Brocade reserves the right to make changes to this document at any time, without notice, and assumes no responsibility
for its use.
IMPORTANT NOTICE
This document is the property of Brocade. This document is confidential to Brocade. It is intended solely as an aid for
installing and configuring Storage Area Networks constructed with Brocade switches. This document does not provide a
warranty to any Brocade software, equipment, or service, nor does it imply product availability. Brocade is not
responsible for the use of this document and does not guarantee the results of its use. Brocade does not warrant or
guarantee that anyone will be able to recreate or achieve the results described in this document. The installation and
configuration described in this document made use of third party software and hardware. Brocade does not make any
warranties or guarantees concerning such third party software and hardware.
The authors and Brocade Communications Systems, Inc. shall have no liability or responsibility to any person or entity
with respect to any loss, cost, liability or damages arising from the information contained in this book or the computer
programs that accompany it.
Export of technical data contained in this document may require an export license from the United States Government.
Brocade Communications Systems,
Incorporated
Corporate Headquarters
1745 Technology Drive
San Jose, CA 95110
T: [phone]
F: [phone]
European Headquarters
29, route de l-Aeroport
Case Postale 105
1211 Geneva 15,
Switzerland
T: [phone]
F: [phone]
[email protected]
Asia-Pacific Headquarters
The Imperial Tower 15th Floor
1-1-1 Uchisaiwaicho
Chiyoda-ku, Tokyo 100-0011
Japan
T: [phone]
F: [phone]
[email protected]

Publication Number 53-0000251-02 iii
Contents
Preface
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
Audience for This Document . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
Chapter 1 Introducing the SilkWorm 12000
Hardware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-2
High Availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3
Reliability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-4
Fault Monitoring and Diagnostics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-4
Intelligent Fabric Services Architecture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-5
Advanced Performance Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-5
Advanced Zoning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-5
Extended Fabrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-5
Fabric Watch. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-6
ISL Trunking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-6
QuickLoop/Fabric Assist (QLFA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-6
Chapter 2 SilkWorm 12000 Architecture and What Is New
Fabric OS 4.0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1
Dual Switch Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-2
Dual Control Processors For High Availability . . . . . . . . . . . . . . . . . . . . 2-3
Accessing the SilkWorm 12000 Switches . . . . . . . . . . . . . . . . . . . . . . . . 2-4
How Logical Switch Behavior Differs. . . . . . . . . . . . . . . . . . . . . . . . . . . 2-5
Port Addressing and Area Numbering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-5
Compatibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-10

iv Publication Number 53-0000251-02
Software High Availability Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-11
Failover Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-11
Failover Details. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-12
Chapter 3 SilkWorm 12000 Based SAN Designs
Scalability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-3
Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-4
ISL Over Subscription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-4
Device Attachment Strategies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-5
Attaching Nodes for Availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-5
Attaching ISLs For Availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-6
Attaching Nodes for Scalability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-7
Availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-8
SilkWorm 12000 Placement In The Fabric. . . . . . . . . . . . . . . . . . . . . . . . . . . 3-9
SilkWorm 12000 Based Fabric Topologies . . . . . . . . . . . . . . . . . . . . . . . . . . 3-10
The Continuum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-10
Single Chassis Topology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-14
Core/Edge Topology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-16
Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-16
Device Attachment Strategies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-16
Core & Edge Switch Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-18
Availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-19
Large Fabric Support Levels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-20
SilkWorm 12000 Reference Topologies. . . . . . . . . . . . . . . . . . . . . . . . . . 3-21
Single chassis topology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-21
Two Chassis/Four Switch Partial Mesh . . . . . . . . . . . . . . . . . . . . . . . . . . 3-22
Core/edge With Maximal Config
(current support levels). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-23
Chapter 4 Deploying the SilkWorm 12000
Deployment Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1

Publication Number 53-0000251-02 v
Unpacking and Installing the SilkWorm 12000 in the Rack. . . . . . . . . . . . . . 4-2
Unpacking the Switch. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-2
Site Planning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-2
Installing the Rack Mount Kit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-3
Reinstalling the Chassis Door and Cable Management Tray . . . . . . . . . . 4-4
Cable Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-5
Chapter 5 Configuring the SilkWorm 12000
Configuring the SilkWorm 12000. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1
Basic configuration steps:. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1
Logging into the SilkWorm 12000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-2
Using Diagnostic Tests to Verify Hardware (Optional) . . . . . . . . . . . . . . 5-2
Configuring IP Addresses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-3
Configuring the Switch Name . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-5
Setting the Domain ID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-5
Enabling Software Licenses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-5
Return Switches to Default Settings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-6
Chapter 6 SilkWorm 12000 Management Interfaces
Telnet. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-1
Web Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-8
Zoning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-10
Upload/Download. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-11
Port Setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-12
Configure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-13
Routing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-14
Extended Fabric . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-15
Fabric Manager 3.0. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-16
Console . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-17

vi Publication Number 53-0000251-02
Chapter 7 Maintaining the SilkWorm 12000
Hardware Maintenance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-6
Power Supply Maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-7
Identify A Faulty Power Supply. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-9
Steps For Installation and Removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-9
Power supply removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-9
Power supply installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-9
Blowers Maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-10
Identify A Faulty Blower Assembly. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-10
Blower Removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-12
Blower Install . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-12
Control Processor Maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-13
Identify A Faulty CP Card . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-14
CP Card Removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-15
CP Card Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-16
CP Card Verification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-17
16-Port Card Maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-17
Identifying A Faulty 16-Port Card . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-20
16-Port Card Removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-20
To Remove A 16-Port Card . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-20
Installing A 16-Port Card . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-22
16-Port Card Verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-22
Software Maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-23
Firmware Upgrade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-23
License upgrade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-25
License Verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-25

Publication Number 53-0000251-02 vii
Chapter 8 Troubleshooting/Support
Software Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-1
QuickLoop Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-2
Trunking Issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-2
Web Tools Issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-3
Zoning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-4
Other Possible Software Issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-4
Hardware Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-5
Software Command Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-6
Cable-side LEDs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-7
Non-cable Side LEDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-8
Miscellaneous Hardware Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-8
Diagnostic Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-9

viii Publication Number 53-0000251-02

Publication Number 53-0000251-02 vii
Preface
Introduction
This document discusses the practical aspects of designing, deploying, and maintaining a SilkWorm
12000 based SAN. The SilkWorm 12000 is a new product because it has: a bladed architecture, two
logical switches in one chassis, a new Fabric OS (version 4.0), and high availability/failover features.
Several SilkWorm 12000 features, while introduced in the SilkWorm 3800, require further discussion
within the context of the SilkWorm 12000, such as Trunking and 1-2 Gbit/sec auto-sensing ports. Other
considerations include designing and deploying SilkWorm 12000 based SANs and the integration of the
SilkWorm 12000 into existing SANs built with SilkWorm 2x00 and 3x00 technology.
This document addresses these new features and capabilities in a clear and concise manner, with liberal
use of examples, and is intended to be used in conjunction with SilkWorm 12000 manuals (see
References later in this section).
While working with the SilkWorm 12000, a multitude of decisions are necessary. This guide is intended
to identify the key decision points expected during the lifecycle of a SilkWorm 12000 deployment and
the advantages and disadvantages of adopting a particular approach. Also included in this guide are tips,
shortcuts, and suggestions for the efficient operation and maintenance of the SilkWorm 12000, which
are gathered from the engineers who developed and tested the SilkWorm 12000.
Audience for This Document
This guide is intended for technically focused personnel directly or indirectly responsible for the design,
deployment, and management of SilkWorm 12000 based SANs. The reader is expected to be familiar
with and have a working knowledge of SAN technology, Brocade SilkWorm switches, and Brocade
Fabric OS.
References
The following Brocade documentation is to be used in reference to this guide.
•Building SANs With Brocade Fabric Switches (Syngress Press) (ISBN: 1-928994-30-x)
•Brocade SAN Design Guide v2.2 (publication number: 53-0000231-03)
•SilkWorm 12000 Hardware Reference Manual –Beta Draft (publication number: 53-0000148-01)
•Brocade Fabric OS Reference Version 3.0/4.0 (publication number: 53-0000182-01)
•Brocade ISL Trunking User’s Guide (publication number: 53-0000189-01)
•Brocade Zoning User’s Guide v 3.0/4.0 (publication number: 53-0000187-01)

viii Publication Number 53-0000251-02
•Web Tools User’s Guide Version 4.0 (publication number: 53-0000185-01)
•Brocade Fabric Manager User’s Guide Version 3.0 (publication number: 53-0000204-0)
•Brocade MIB Reference Version 3.0/4.0 (publication number: 53-0000184-01)
•Brocade SilkWorm 12000 Core Migration User’s Guide v1.1 (53-0000477-02)

Publication Number 53-0000251-02 1-1
1
Introducing the SilkWorm 12000
This chapter includes the following sections:
•Hardware on page 1-2
•High Availability on page 1-3
•Reliability on page 1-4
•Fault Monitoring and Diagnostics on page 1-4
•Intelligent Fabric Services Architecture on page 1-5
The Brocade SilkWorm 12000 core fabric switch represents the next generation of advanced Fibre
Channel switches used to intelligently interconnect storage devices, hosts, and servers in a Storage Area
Network (SAN). It is a revolutionary product: a dual 64-port Fibre Channel switch that delivers
unprecedented performance, scalability, flexibility, functionality, reliability and availability. Figure 1-1
shows two views of the SilkWorm 12000. Several key features of the SilkWorm 12000 and the Fabric
operating system (Fabric OS) are detailed below:
•The dual switch capability allows either one or two 64-port switches per chassis. The switches may
be interconnected together to create a single high port count fabric, or they may be used in a highly
available redundant fabric SAN. Dual redundant control processors (CP) provide high availability
within the chassis. The control processors are located on the CP cards.
•The SilkWorm 12000 is based on Brocade’s third generation technology, which supports 1 and 2
Gbit/sec auto-sensing Fibre Channel ports. Trunking technology groups up to four ports together to
create high performance 8 Gbit/sec ISL trunks between switches.
•Universal ports self-configure as E-ports, F-ports, or FL-ports.
•Small Form-Factor Pluggable (SFP) optical transceivers support any combination of Short
Wavelength (SWL) and Long Wavelength (LWL) optical media on a single switch module
(hereafter called 16-port card).
•Fully networkable, the SilkWorm 12000 offers forward and backward compatibility with all
Brocade SilkWorm switches.
•High availability, redundant design, extensive diagnostics, and system monitoring capabilities
integrated with Fabric OS management tools deliver unprecedented Reliability, Availability, and
Serviceability (RAS).
•The SilkWorm 12000 offers a highly available platform for mission-critical SAN-designed
applications.
•Extensible and multi-protocol to support 1 Gbit/sec, 2 Gbit/sec and 10 Gbit/sec Fibre Channel, IP,
and InfiniBand protocols.

1-2 Publication Number 53-0000251-02
Introducing the SilkWorm 12000
1
Figure 1-1 Cable and non-cable side views of the SilkWorm 12000
Hardware
The SilkWorm 12000 features a modular and scalable mechanical construction that allows a wide range
of flexibility in switch installation, Fabric design, and maintenance. Using a 14U (rack unit) mechanical
design, up to three SilkWorm 12000 chassis may be mounted in a standard 42U rack, supporting as many
as 384 Fibre Channel ports in a single rack. As shown in Figure 1-2, the modular multi-card assembly
chassis of the SilkWorm 12000 consists of the following:
•Up to eight hot-swappable 16-port cards, delivering up to two separate 64-port Fibre Channel
switches in a single chassis. Each 64-port switch uses four 16-port cards.
•Two slots for Control Processor cards.
-A single active CP card can control both 64-port switches in the chassis.
-A redundant CP card can assume control of a single or dual switch configuration in the event of
an active CP failure.
•Modular hot-swappable Field Replaceable Units (FRUs):
-16-Port Card
-Control Processor Card (CP Card)
-Small Form-Factor Pluggable (SFP) optical transceivers
-Blower assembly
-Power supply
Figure 1-2 identifies the components as described above.
Cable Side (Front) Non-Cable Side (Rear)

Publication Number 53-0000251-02 1-3
Introducing the SilkWorm 12000 1
Figure 1-2 SilkWorm 12000 With Identified Components
High Availability
High availability is an all-encompassing term, and this term normally includes attributes such as
reliability and availability. If a system is under continuous operation, it is accessible 7 days a week and
24 hours per day by users and the system manager. Availability is designated in terms of 9s. For example,
the architecture of the SilkWorm 12000 is designed for availability in excess of 99.999%.
The following features contribute to the high availability design of the SilkWorm 12000:
•Redundant, hot-swappable components
•Redundant power and cooling subsystems
•No single point of failure
•Enhanced data integrity on all data paths
•Fabric Shortest Path First (FSPF) automatic rerouting around failed links
•Integration with SNMP managers
•Automatic Control Processor fail over
Port 15
Port 0
Slot 1
16-port cards
1-4
Slot numbers
7-10
Power supply #4
Power supply #3
Power supply #2
Power supply #1
Slot 10
2 CP cards
Slot numbers 5 and 6

1-4 Publication Number 53-0000251-02
Introducing the SilkWorm 12000
1
The SilkWorm 12000 high availability software architecture provides a common framework for all
applications that reside on the system and allows global and local states to be maintained such that any
component failure is fully manageable. High availability elements consist of the High Availability
Manager, the heartbeat, the fault/health framework, the replicated database, initialization, and software
upgrade. The software high availability model is discussed in more detail later in this section.
The power supplies (four total) support two 64-port switches in a chassis with the ability to tolerate the
failure of as many as two power supplies. The power supplies are hot-swappable, without taking the
switch down and without incurring any outage.
The blower assemblies (three total) are also hot-swappable and the chassis can operate fully with only
two blower assemblies in place allowing for the failed blower assembly to be replaced with no outage.
The recommended systems networks for high availability include the use of redundant fabrics and dual
paths from hosts to storage devices in a SAN. When dual fabrics are used, one switch, one link, or an
entire fabric can go down, but data traffic will be re-routed to the alternate path ensuring that the SAN
remains operational.
Reliability
In addition to being available, the system must be reliable. This means that some, if not all, of its internal
state must be maintained. In a reliable system, a user is not aware of the internal state of the chassis
components and will experience continued system service with zero degradation of service.
The SilkWorm 12000 provides the following features to ensure reliability. All data inside the switch is
protected by the following Error Detection and Correction mechanisms as follows:
•Power-on self test (POST)
•Error detection and fault isolation (EDFI), such as cyclic redundancy checking (CRC), parity
checking, checksum, and illegal address checking
•Dual control processors, with each control processor containing two serial ports and one Ethernet
port. Offline Control Processor diagnostics and remote diagnostics make troubleshooting
straightforward.
•I2C monitoring and control of environmental conditions
Fault Monitoring and Diagnostics
Fault monitoring, diagnostic tests, and system status indicators simplify management and ensure
availability of the SilkWorm 12000.
Diagnostic testing occurs in three areas: Power-On Self Test (POST), switch level testing, and
manufacturing tests. The Power-On Self Test is card oriented and ensures that the switch is ready for use
during power up. Switch level testing is done at the user port level. These tests rely on the standard
Fabric OS support to provide routing and port setup. Manufacturing support includes long duration
testing.
The WWN card located on the non-cable side of the switch summarizes the system status of each 16-port
card, each Control Processor Card, and each power supply module. LEDs on the blowers show the status
of the blower assemblies.

Publication Number 53-0000251-02 1-5
Introducing the SilkWorm 12000 1
Brocade’s Fabric Watch exposes enhanced status reporting capabilities of the SilkWorm switches
through all the standard management interfaces, including SNMP, the Fabric Access Layer API, Brocade
Web Tools, Fabric Manager, and the command line interface.
Intelligent Fabric Services Architecture
Fabric OS v4.0 (required for the SilkWorm 12000) provides a wide variety of Advanced Fabric Services
that are designed to improve the investment protection, security, performance, scalability, and efficiency
of Brocade SAN fabrics.
Features of Fabric OS v4.0 include Trunking, Advanced Zoning, and Advanced Performance
Monitoring. These features, some of which are also available in previous versions of Fabric OS, are
discussed in the following pages.
Advanced Performance Monitoring
Advanced Performance Monitoring (available on 2 Gbit/sec switches) can monitor performance
characteristics on any attribute within the first 64 bytes of a Fibre Channel frame. Predefined graphs
measure end-to-end performance; port, switch, and AL-PA bandwidth utilization; SCSI commands
(read, write, and read/write); and protocol comparisons (SCSI versus IP). As a result, performance
monitoring provides the foundation for performance tuning, resource optimization, service level
agreement compliance reporting, and bill-back applications.
Advanced Zoning
Advanced Zoning software limits access to data by segmenting a fabric into virtual private SANs. On
1 Gbit/sec and 2 Gbit/sec switches, software-enforced zoning prevents hosts from discovering
unauthorized target devices. Hardware-enforced zoning prevents a host from accessing a device that is
unauthorized. This provides the most secure zoning available. In addition, Advanced Zoning on
2 Gbit/sec switches enables hardware enforcement for devices identified by World Wide Name (WWN).
This is new functionality that was not available in the SilkWorm 2000 series switches, which could only
do soft WWN zoning. With WWN zoning, zone enforcement adjusts automatically, even if a device
moves to another port. This new zoning model allows for the continued flexibility that traditional
software-enhanced zoning provides plus garners the security benefits of legacy hardware-enforced
zoning.
Extended Fabrics
Extended Fabrics software enables long distance (100km) ISLs over dark fiber or Dense Wave Division
Multiplexing (DWDM) connections at full bandwidth.

1-6 Publication Number 53-0000251-02
Introducing the SilkWorm 12000
1
Fabric Watch
Fabric Watch software enables organizations to set thresholds and ranges for SAN fabrics, and raise
management alerts when performance or errors vary outside predefined ranges.
ISL Trunking
ISL Trunking increases the performance and availability of links between 2 Gbit/sec switches and
minimizes the SAN management effort. Up to four 2 Gbit/sec ISLs between two switches can be
combined into a single trunk, or logical ISL, at 8 Gbit/sec. Traffic is load balanced across all the links
(i.e. any traffic can go across any available trunk link). This is an improvement over current static routing
and load sharing where servers are allocated individual dedicated links.
QuickLoop/Fabric Assist (QLFA)
QuickLoop/Fabric Assist (QLFA) connects private loop hosts to the SAN fabric for better performance
and fault management, while protecting investments in legacy loop devices. Because many legacy
devices are designed for FC-AL configurations, Fabric OS translative mode protects investments by
supporting private loop target devices. The SilkWorm 12000 running Fabric OS v4.0 supports translative
mode. Therefore switches that do support QuickLoop or Fabric Assist can be connected to a SilkWorm
12000, even though the SilkWorm 12000 does not support QuickLoop or Fabric Assist directly. It is
possible to connect devices that are accessed by QuickLoop/Fabric Assist devices to the SilkWorm
12000. This means that any type of target device may be attached to a switch running Fabric OS v4.0 and
may be included in a QuickLoop Fabric Assist zone that has its private host attached to a switch running
QuickLoop and Zoning. QuickLoop and Zoning are pre-requisites for QLFA, on Fabric OS v2.3 or later
(SilkWorm 2xxx) or v3.0.1 or later (SilkWorm 3800/3200).

Publication Number 53-0000251-02 2-1
2
SilkWorm 12000 Architecture and What Is New
This chapter includes the following sections:
•Fabric OS 4.0 on page 2-1
•Dual Switch Model on page 2-2
•Port Addressing and Area Numbering on page 2-5
•Compatibility on page 2-10
•Software High Availability Model on page 2-11
Note: All topics in this section establish a foundation for further discussions regarding the design,
operation, and management of the SilkWorm 12000 and SilkWorm 12000 based SANs. Where
appropriate, detail is provided in this section to identify key changes between the SilkWorm
12000 and previous SilkWorm switch models or to highlight key architectural features.
The SilkWorm 12000 utilizes embedded Linux as its underlying operating system, however all SAN
management is still performed on the Fabric OS level. While the impact to previous users of SilkWorm
switches is nominal, it is important to note what has changed, and what prompted these changes. The
dual switch model introduces several new concepts that are important to understand for the design,
deployment, and maintenance of the SilkWorm 12000 and SilkWorm 12000-based SANs. The 16-port
card design of the SilkWorm 12000 introduces a new “slot” operand to many commands, and a new
model for port identification that should be understood for effective operation of the SilkWorm 12000.
The SilkWorm 12000 is compatible with all Brocade switch models and is interoperable with switches
from vendors such as McData. To enable this compatibility, you must change certain configuration
settings before connecting other switches to the SilkWorm 12000. Finally, the software high availability
architecture is discussed.
Fabric OS 4.0
Fabric OS v4.0 is built upon a real-time version of Linux version 2.4. Linux was chosen due to industry
wide support for hardware and software, portability, and scalability. Figure 2-1 shows the screen output
when booting the switch. Notice the references to Linux.

2-2 Publication Number 53-0000251-02
SilkWorm 12000 Architecture and What Is New
2
Figure 2-1 SilkWorm 12000 Boot Up With Fabric OS v4.0
Fabric OS 4.0 is a superset of previous versions of Fabric OS, so most commands used with previous
versions as well as several new commands are available.
Warning: The SilkWorm 12000 is equipped with a Root account intended for diagnostics and
debugging purposes solely by the trained engineers of the equipment vendor. Improper use of the
functionality made available through the Root account, such as treating the SilkWorm 12000 like
a standard Linux system and using the Linux functions and commands, can cause significant
harm and disruption to the operation of the SAN fabric. During normal operation, log in to the
switch as the “admin” user. The “admin” user utilizes a restricted shell with access to Fabric OS
commands only.
Dual Switch Model
The SilkWorm 12000 houses two separate logical switches within a single chassis. Each switch is
capable of scaling to 64 ports by adding up to four 16-port cards to the respective logical switch. It is
possible to interconnect the two switches inside a chassis to form a fabric. Each logical switch has its
own unique domain ID, WWN, and IP address. Each switch in the chassis is an entity accessible through
telnet and other methods, using the unique IP address of that switch. The switches are numbered zero and
one, as shown in Figure 2-2. You can also access a switch using a serial connection for installation, such
as for setting the IP address of the switch or for diagnostic purposes.
The system is coming up, please wait...
Checking system RAM - press any key to stop test
Checking memory address: 08000000
System RAM check complete
Press escape within 4 seconds to enter boot interface.
Entry point at 0x00400000 ...
Loading Initial RAM disk
Uncompressing Linux... done.
read_silkworm_bdinfo: silkworm->board = 10, silkworm->board_rev =2
id mach(): done
Physical Memory: 0x08000000 Used memory = 0x07f00000
Linux version 2.4.2-mvista_010329 (swrel@nermal) (gcc version 2.95.3 20010112
(prerelease)) #1 Tue Dec 11 00:39:11 PST 2001

Publication Number 53-0000251-02 2-3
SilkWorm 12000 Architecture and What Is New 2
Figure 2-2 SilkWorm 12000 Dual-Switch Model
Dual Control Processors For High Availability
The dual control processors (CP) operate in an active/standby model. The CP cards are named “cp0” (the
CP card on the left, when looking at the chassis) and “cp1” (the CP card on the right when looking at the
chassis), as shown in Figure 2-2. Each CP card is assigned an IP address for maintenance and diagnostic
purposes. Cp0 is located in slot 5 and cp1 is located in slot 6. One CP card is active and manages both
switches. The other CP card is in standby mode and will become active when a failover is required or a
forced failover occurs. Both CP cards should run the same version of Fabric OS. When configured for a
single switch, meaning only one switch is populated with 16-port cards, it is still necessary to have two
CP cards to maintain availability should one CP card fail. Normally, cp0 is given preference to become
the active CP card and cp1 operates as the standby CP card. In the initial release of Fabric OS v4.0, when
a failover occurs, it takes approximately thirty seconds before the standby CP card becomes active. The
use of a dual fabric solution can mitigate or eliminate the impact of this failover period. During the
failover period, I/O is not possible and attached devices must log back into the fabric and re-authenticate.
In properly designed high availability SANs, the real effect on I/O can be limited to as little as four
seconds. The next release of Fabric OS v4.1 is planned to support completely non-disruptive updates to
minimize path interruption.
Logical
Switch 0 Logical Switch 1
cp0 cp1

2-4 Publication Number 53-0000251-02
SilkWorm 12000 Architecture and What Is New
2
The dual CP card model and the concept of logical switches is a change from past SilkWorm switch
implementations. The SilkWorm 2000 and 3000 families of 8-port and 16-port switches all had a static
relationship between the processor and the switch. Now the switch and the processor are de-coupled.
One implication of this model is that instead of downloading firmware to a switch, it is necessary to
download firmware to a CP card.
Note: The time it takes to activate the standby CP, when a failover occurs, will be considerably less
when using Fabric OS 4.1 and greater.
Accessing the SilkWorm 12000 Switches
When accessing SilkWorm 12000 switches, it is possible to access either switch by its respective IP
address or by using a serial connection to the active CP card. It is possible to access a CP card by telnet
using the respective CP card’s IP address or by connecting a serial cable to the CP card. Access to the CP
card should be limited to: installation purposes, setting a switch’s IP address, doing firmware
maintenance (i.e., downloads), or for diagnostic purposes.
When telnetting to an inactive CP card, the user is entered into a limited access environment where no
access to a switch is possible. When a user accesses an active CP card, that user will have access to the
full Fabric OS environment. To determine a CP card state, whether inactive or active, use the command
haShow (see Figure 2-3). When telnetting to a SilkWorm 12000 switch, the user will encounter a login
prompt from the active CP card. This may seem confusing, since the destination is a switch and not the
CP card; however, once logged in, the user is then placed into the target switch environment, as shown in
Figure 2-4.
Note: While it is possible to access the switches via the active CPcard, either using an Ethernet address
or serial connection, primary access to the switches should be via Ethernet to the switch and not
the CP card.
Figure 2-3 Determining a CP Card State Using the haShow Command. CP1 is Active.
Figure 2-4 SilkWorm 12000 Log In
sun1# telnet sw0_156_22
Trying 192.168.156.22...
Connected to sw0_156_22.
Escape character is '^]'.
Fabric OS (cp1)
cp1 login: admin
Password:
sw0_156_22:admin> hashow
Local CP (Slot 6, CP1): Active
Remote CP (Slot 5, CP0): Standby
HA Enabled, Heartbeat Up
sun1# telnet sw1_156_23
Trying 192.168.156.23...
Connected to sw1_156_23.
Escape character is '^]'.
Fabric OS (cp1)
cp1 login: admin
Password:
sw1_156_23:admin>