Dell Brocade 6505, Brocade 5100, Brocade 5300 User Manual

Download

Page 1

53-1003141-01

27 June 2014

Fabric OS

Troubleshooting and Diagnostics Guide

Supporting Fabric OS v7.3.0

Page 2

ADX, AnyIO, Brocade, Brocade Assurance, the B-wing symbol, DCX, Fabric OS, ICX, MLX, MyBrocade, OpenScript, VCS, VDX, and Vyatta are registered trademarks, and HyperEdge, The Effortless Network, and The On-Demand Data Center are trademarks of Brocade Communications Systems, Inc., in the United States and/or in other countries. Other brands, products, or service names mentioned may be trademarks of their respective owners.

Notice: This document is for informational purposes only and does not set forth any warranty, expressed or implied, concerning any equipment, equipment feature, or service of fered or to be offered by Brocade. Brocade reserves the right to make changes to this document at any time, without notice, and assumes no responsibility for its use. This informational document describes features that may not be currently available. Contact a Brocade sales office for information on feature and product availability. Export of technical data contained in this document may require an export license from the United States government.

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.

The product described by this document may contain “open source” software covered by the GNU General Public License or other open source license agreements. To find out which open source software is included in Brocade products, view the licensing terms applicable to the open source software, and obtain a copy of the programming source code, please visit http://www.brocade.com/support/oscd.

Brocade Communications Systems, Incorporated

Corporate and Latin American Headquarters Brocade Communications Systems, Inc. 130 Holger Way San Jose, CA 95134 Tel: 1-408-333-8000 Fax: 1-408-333-8101 E-mail: info@brocade.com

European Headquarters Brocade Communications Switzerland Sàrl Centre Swissair Tour B - 4ème ét age 29, Route de l'Aéroport Case Postale 105 CH-1215 Genève 15 Switzerland Tel: +41 22 799 5640 Fax: +41 22 799 5641 E-mail: emea-info@brocade.com

Asia-Pacific Headquarters Brocade Communications Systems China HK, Ltd. No. 1 Guanghua Road Chao Yang District Units 2718 and 2818 Beijing 100020, China Tel: +8610 6588 8888 Fax: +8610 6588 9999 E-mail: china-info@brocade.com

Asia-Pacific Headquarters Brocade Communications Systems Co., Ltd. (Shenzhen WFOE) Citic Plaza No. 233 Tian He Road North Unit 1308 – 13th Floor Guangzhou, China Tel: +8620 3891 2000 Fax: +8620 3891 2111 E-mail: china-info@brocade.com

Page 3

Document History

Title Publication number Summary of changes Date

Fabric OS Troubleshooting and Diagnostics Guide

53-0000853-01 First released edition. March 2008

53-1001187-01 Added support for Virtual Fabrics, fcPing,

pathInfo, and additional troubleshooting tips.

53-1001340-01 Added support for checking physical

connections, updated commands, removed obsolete information, and moved the FCIP and FICON chapters into their respective books.

53-1001769-01 Added support for the Rolling Reboot

Detection feature and the Superping tool; added enhancements for supportSave and spinFab; updated commands; transferred the iSCSI chapter into its respective book.

53-1002150-01 Added Frame Viewer and Diagnostics port

features.

53-1002150-02 Updated the Diagnostics port feature. June 2011

53-1002751-01 Updated for Fabric OS v7.1.0 December 2012

53-1002751-02 Corrected errors and omissions in the

guide.

53-1002930-01 Updated for Fabric OS v7.2.0. July 2013

53-1003141-01 Updated for Fabric OS v7.3.0. June 2014

November 2008

July 2009

March 2010

April 2011

March 2013

Fabric OS Troubleshooting and Diagnostics Guide iii 53-1003141-01

Page 4

iv Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 5

Contents

About This Document

How this document is organized . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

Supported hardware and software . . . . . . . . . . . . . . . . . . . . . . . . . . xii

What’s new in this document. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

Document conventions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

Text formatting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

Command syntax conventions . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

Command examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xiv

Notes, cautions, and warnings . . . . . . . . . . . . . . . . . . . . . . . . . . xiv

Key terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiv

Additional information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

Brocade resources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

Other industry resources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

Getting technical help. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

Document feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvi

Chapter 1 Introduction

Troubleshooting overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Network Time Protocol. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Most common problem areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

Questions for common symptoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

Gathering information for your switch support provider. . . . . . . . . . . 5

Setting up your switch for FTP. . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

Using the supportSave command. . . . . . . . . . . . . . . . . . . . . . . . . 5

Capturing output from a console . . . . . . . . . . . . . . . . . . . . . . . . . 7

Capturing command output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Building a case for your switch support provider . . . . . . . . . . . . . . . . 7

Basic information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Detailed problem information. . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

Gathering additional information . . . . . . . . . . . . . . . . . . . . . . . .10

Chapter 2 General Troubleshooting

Licenses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11

Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

Frame Viewer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12

Switch message logs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Fabric OS Troubleshooting and Diagnostics Guide v 53-1003141-01

Page 6

Switch boot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .14

Rolling Reboot Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .14

FC-FC routing connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Generating and routing an ECHO . . . . . . . . . . . . . . . . . . . . . . . . 16

Superping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18

Routing and statistical information . . . . . . . . . . . . . . . . . . . . . . 21

Performance issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .22

Chapter 3 Connectivity

Port initialization and FCP auto-discovery process. . . . . . . . . . . . . .23

Link issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25

Connection problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25

Checking the physical connection . . . . . . . . . . . . . . . . . . . . . . . 25

Checking the logical connection . . . . . . . . . . . . . . . . . . . . . . . . .26

Checking the Name Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27

Link failures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28

Determining a successful speed negotiation. . . . . . . . . . . . . . .28

Checking for a loop initialization failure . . . . . . . . . . . . . . . . . . .29

Checking for a point-to-point initialization failure . . . . . . . . . . . 29

Correcting a port that has come up in the wrong mode . . . . . .30

Marginal links . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30

Troubleshooting a marginal link . . . . . . . . . . . . . . . . . . . . . . . . . 31

Device login issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .33

Pinpointing problems with device logins . . . . . . . . . . . . . . . . . .33

Media-related issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35

Testing the external transmit and receive path of a port . . . . . 36

Testing the internal components of a switch . . . . . . . . . . . . . . .36

Testing components to and from the HBA . . . . . . . . . . . . . . . . . 36

Segmented fabrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

Reconciling fabric parameters individually . . . . . . . . . . . . . . . . 37

Downloading a correct configuration . . . . . . . . . . . . . . . . . . . . .38

Reconciling a domain ID conflict . . . . . . . . . . . . . . . . . . . . . . . .38

Reconciling incompatible software features . . . . . . . . . . . . . . .39

Port mirroring. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

In-Order Delivery. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .40

Port mirroring considerations . . . . . . . . . . . . . . . . . . . . . . . . . . .40

Supported platforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

Maximum mirror connections . . . . . . . . . . . . . . . . . . . . . . . . . . .42

Configuring a port to be a mirror port. . . . . . . . . . . . . . . . . . . . .43

Adding a port mirror connection . . . . . . . . . . . . . . . . . . . . . . . . .43

Deleting a port mirror connection. . . . . . . . . . . . . . . . . . . . . . . .43

Displaying port mirror connections. . . . . . . . . . . . . . . . . . . . . . .44

Chapter 4 Configuration

Configuration upload and download issues . . . . . . . . . . . . . . . . . . .45

Gathering additional information . . . . . . . . . . . . . . . . . . . . . . . .48

vi Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 7

Brocade configuration form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .48

Chapter 5 Firmware Download Errors

Blade troubleshooting tips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

Firmware download issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .52

Troubleshooting with the firmwareDownload command . . . . . . . . .55

Gathering additional information . . . . . . . . . . . . . . . . . . . . . . . .55

USB error handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .56

Considerations for downgrading firmware . . . . . . . . . . . . . . . . . . . .56

Preinstallation messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

Blade types. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59

Firmware versions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59

Platform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .60

Routing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .60

Chapter 6 Security

Passwords . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

Password recovery options . . . . . . . . . . . . . . . . . . . . . . . . . . . . .62

Device authentication. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

Protocol and certificate management . . . . . . . . . . . . . . . . . . . . . . . .63

Gathering additional information . . . . . . . . . . . . . . . . . . . . . . . .63

SNMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

Gathering additional information . . . . . . . . . . . . . . . . . . . . . . . .64

FIPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

Chapter 7 Virtual Fabrics

General Virtual Fabrics troubleshooting . . . . . . . . . . . . . . . . . . . . . .65

Fabric identification issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66

Logical Fabric issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66

Base switch issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .67

Logical switch issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .68

Switch configuration blade compatibility. . . . . . . . . . . . . . . . . . . . . .69

Gathering additional information. . . . . . . . . . . . . . . . . . . . . . . . . . . .70

Chapter 8 ISL Trunking

Link issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71

Buffer credit issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

Getting out of buffer-limited mode . . . . . . . . . . . . . . . . . . . . . . .72

Fabric OS Troubleshooting and Diagnostics Guide vii 53-1003141-01

Page 8

Chapter 9 Zoning

Overview of corrective action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .73

Verifying a fabric merge problem . . . . . . . . . . . . . . . . . . . . . . . .73

Verifying a TI zone problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . .73

Segmented fabrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

Zone conflicts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75

Resolving zoning conflicts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

Correcting a fabric merge problem quickly . . . . . . . . . . . . . . . . 77

Changing the default zone access . . . . . . . . . . . . . . . . . . . . . . .77

Editing zone configuration members . . . . . . . . . . . . . . . . . . . . .77

Reordering the zone member list . . . . . . . . . . . . . . . . . . . . . . . .78

Checking for Fibre Channel connectivity problems . . . . . . . . . .78

Checking for zoning problems. . . . . . . . . . . . . . . . . . . . . . . . . . . 79

Gathering additional information. . . . . . . . . . . . . . . . . . . . . . . . . . . .80

Chapter 10 Diagnostic Features

Fabric OS diagnostics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81

Diagnostic information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

Power-on self-test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .82

Disabling POST. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84

Enabling POST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84

Switch status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84

Viewing the overall status of the switch . . . . . . . . . . . . . . . . . . .85

Displaying switch information . . . . . . . . . . . . . . . . . . . . . . . . . . .85

Displaying the uptime for a switch . . . . . . . . . . . . . . . . . . . . . . .87

Using the spinFab and portTest commands . . . . . . . . . . . . . . . . . . . 87

Debugging spinFab errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88

Clearing the error counters . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89

Enabling a port. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

Disabling a port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .90

Port information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .90

Viewing the status of a port . . . . . . . . . . . . . . . . . . . . . . . . . . . .90

Displaying the port statistics. . . . . . . . . . . . . . . . . . . . . . . . . . . .91

Displaying a summary of port errors for a switch . . . . . . . . . . . 92

Equipment status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .93

Checking the temperature, fan, and power supply . . . . . . . . . .94

Checking the status of the fans . . . . . . . . . . . . . . . . . . . . . . . . .94

Checking the status of a power supply. . . . . . . . . . . . . . . . . . . . 94

Checking temperature status . . . . . . . . . . . . . . . . . . . . . . . . . . .95

System message log . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .95

Displaying the system message log with no page breaks. . . . . 96

Displaying the system message log one message at a time . .96

Clearing the system message log . . . . . . . . . . . . . . . . . . . . . . . .96

Port log. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .96

Viewing the port log . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

viii Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 9

Syslogd configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98

Configuring the host . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .99

Configuring the switch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

Automatic trace dump transfers . . . . . . . . . . . . . . . . . . . . . . . . . . .100

Specifying a remote server . . . . . . . . . . . . . . . . . . . . . . . . . . . .101

Enabling the automatic transfer of trace dumps. . . . . . . . . . .101

Setting up periodic checking of the remote server . . . . . . . . .101

Saving comprehensive diagnostic files to the server . . . . . . .101

Appendix A Switch Type and Blade ID

Appendix B Hexadecimal Conversion

Hexadecimal overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

Example conversion of the hexadecimal triplet Ox616000 . . . . . .107

Decimal-to-hexadecimal conversion table. . . . . . . . . . . . . . . . . . . .108

Index

Fabric OS Troubleshooting and Diagnostics Guide ix 53-1003141-01

Page 10

x Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 11

About This Document

In this chapter

•

How this document is organized . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

•

Supported hardware and software. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii

•

What’s new in this document . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

•

Document conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

•

Additional information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

•

Getting technical help . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

•

Document feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvi

How this document is organized

The document contains the following chapters:

•

Chapter 1, “Introduction,” gives a brief overview of troubleshooting the Fabric OS, and provides

procedures for gathering basic information from your switch and fabric to aid in troubleshooting.

•

Chapter 2, “General Troubleshooting,” provides information on licensing, hardware, and syslog

issues.

•

Chapter 3, “Connectivity,” provides information and procedures on troubleshooting various link

issues.

•

Chapter 4, “Configuration,” provides troubleshooting information and procedures for

configuration file issues.

•

Chapter 5, “Firmware Download Errors,” provides procedures for troubleshooting firmware

download issues.

•

Chapter 6, “Security,” provides procedures for user account and security issues.

•

Chapter 7, “Virtual Fabrics,” provides procedures for troubleshooting Virtual Fabrics.

•

Chapter 8, “ISL Trunking,” provides procedures for resolving trunking issues.

•

Chapter 9, “Zoning,” provides preparations and procedures for performing firmware

downloads, as well troubleshooting information.

•

Chapter 10, “Diagnostic Features,” provides procedures for the use of the diagnostics

commands for the chassis, ports, and other chassis equipment, as well as providing information on the system messages.

•

Appendix A, “Switch Type and Blade ID,” provides reference information to guide you in

understanding switch output.

•

Appendix B, “Hexadecimal Conversion,” provides reference information for translating

hexadecimal output.

Fabric OS Troubleshooting and Diagnostics Guide xi 53-1003141-01

Page 12

Supported hardware and software

In those instances in which procedures or parts of procedures documented here apply to some switches but not to others, this guide identifies which switches are supported and which are not.

Although many different software and hardware configurations are tested and supported by Brocade Communications Systems, Inc. for Fabric OS v7.3.0, documenting all possible configurations and scenarios is beyond the scope of this document.

The following hardware platforms are supported by this release of Fabric OS:

•

Brocade 300 switch

•

Brocade 5100 switch

•

Brocade 5300 switch

•

Brocade 5410 embedded switch

•

Brocade 5424 embedded switch

•

Brocade 5430 embedded switch

•

Brocade 5431 embedded switch

•

Brocade 5432 embedded switch

•

Brocade 5450 embedded switch

•

Brocade 5460 embedded switch

•

Brocade 5470 embedded switch

•

Brocade 5480 embedded switch

•

Brocade M6505 embedded switch

•

Brocade 6505 switch

•

Brocade 6510 switch

•

Brocade 6520 switch

•

Brocade 6547 embedded switch

•

Brocade 6548 embedded switch

•

Brocade 7800 extension switch

•

Brocade 7840 extension switch

•

Brocade VA-40FC

•

Brocade Encryption Switch

•

Brocade DCX Backbone family:

Brocade DCX

Brocade DCX-4S

•

Brocade DCX 8510 Backbone family:

Brocade DCX 8510-4

Brocade DCX 8510-8

xii Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 13

What’s new in this document

The following information is added:

•

Support for Brocade 7840 extension switch.

•

Support for FC16-64 port blade.

•

POST2 section is updated.

For further information about documentation updates for this release, refer to the release notes.

Document conventions

This section describes text formatting conventions and important notice formats used in this document.

Text formatting

The narrative-text formatting conventions that are used are as follows:

bold text Identifies command names

Identifies the names of user-manipulated GUI elements Identifies keywords and operands Identifies text to enter at the GUI or CLI

italic text Provides emphasis

Identifies variables Identifies paths and Internet addresses Identifies document titles

code

text Identifies CLI output

Identifies command syntax examples

Command syntax conventions

For readability, command names in the narrative portions of this guide are presented in mixed lettercase: for example, switchShow. In actual examples, command lettercase is often all lowercase. Otherwise, this manual specifically notes those cases in which a command is case-sensitive. Command syntax in this manual follows these conventions:

command Commands are printed in bold.

option, option Command options are printed in bold.

argument, arg Arguments.

[ ] Optional element.

variable Variables are printed in italics. In the help pages, values are underlined

enclosed in angled brackets < >.

Fabric OS Troubleshooting and Diagnostics Guide xiii 53-1003141-01

Page 14

NOTE

ATTENTION

CAUTION

DANGER

... Repeat the previous element, for example “member[;member...]”

value Fixed values following arguments are printed in plain font. For example,

show WWN

| Boolean. Elements are exclusive. Example:

show -mode egress | ingress

Command examples

This book describes how to perform configuration tasks using the Fabric OS command line interface, but does not describe the commands in detail. For complete descriptions of all Fabric OS commands, including syntax, operand descriptions, and sample output, refer to the Fabric OS Command Reference.

Notes, cautions, and warnings

The following notices and statements are used in this manual. They are listed below in order of increasing severity of potential hazards.

A note provides a tip, guidance, or advice, emphasizes important information, or provides a reference to related information.

An Attention statement indicates potential damage to hardware or data.

A Caution statement alerts you to situations that can be potentially hazardous to you or cause damage to hardware, firmware, software, or data.

A Danger statement indicates conditions or situations that can be potentially lethal or extremely hazardous to you. Safety labels are also attached directly to products to warn of these conditions or situations.

Key terms

For definitions specific to Brocade and Fibre Channel, refer to the Brocade Glossary.

For definitions of SAN-specific terms, visit the Storage Networking Industry Association online dictionary at:

http://www.snia.org/education/dictionary

xiv Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 15

Additional information

This section lists additional Brocade and industry-specific documentation that you might find helpful.

Brocade resources

To get up-to-the-minute information, go to http://my.brocade.com and register at no cost for a user ID and password.

White papers, online demonstrations, and data sheets are available through the Brocade website at:

http://www.brocade.com/products-solutions/products/index.page

For additional Brocade documentation, visit the Brocade website:

http://www.brocade.com

Release notes are available on the MyBrocade website and are also bundled with the Fabric OS firmware.

Other industry resources

For additional resource information, visit the Technical Committee T11 website. This website provides interface standards for high-performance and mass storage applications for Fibre Channel, storage management, and other applications:

http://www.t11.org

For information about the Fibre Channel industry, visit the Fibre Channel Industry Association website:

http://www.fibrechannel.org

Getting technical help

Contact your switch support supplier for hardware, firmware, and software support, including product repairs and part ordering. To expedite your call, have the following information available:

1. General Information

•

Switch model

•

Switch operating system version

•

Error numbers and messages received

•

supportSave command output

•

Detailed description of the problem, including the switch or fabric behavior immediately following the problem, and specific questions

•

Description of any troubleshooting steps already performed and the results

•

Serial console and Telnet session logs

Fabric OS Troubleshooting and Diagnostics Guide xv 53-1003141-01

Page 16

•

'"!&'

FT00X0054E9

syslog message logs

2. Switch Serial Number

The switch serial number and corresponding bar code are provided on the serial number label, as illustrated below.

The serial number label is located as follows:

•

Brocade 5424 — On the bottom of the switch module.

•

Brocade 300, 5100, and 5300 — On the switch ID pull-out tab located on the bottom of the port side of the switch.

•

Brocade 6505, 6510, and 6520— On the switch ID pull-out tab located inside the chassis on the port side on the left.

•

Brocade 7800 — On the bottom of the chassis.

•

Brocade DCX Backbone — On the bottom right on the port side of the chassis.

•

Brocade DCX-4S Backbone — On the bottom right on the port side of the chassis.

•

Brocade DCX 8510-4 — On the nonport side of the chassis, on the left just below the left power supply.

•

Brocade DCX 8510-8 — On the bottom right on the port side of the chassis and directly above the cable management comb.

3. World Wide Name (WWN)

Use the licenseIdShow command to display the chassis WWN.

If you cannot use the licenseIdShow command because the switch is inoperable, you can get the WWN from the same place as the serial number, except for the Brocade DCX. For the Brocade DCX, access the numbers on the WWN cards by removing the Brocade logo plate at the top of the nonport side of the chassis.

Document feedback

Quality is our first concern at Brocade and we have made every effort to ensure the accuracy and completeness of this document. However, if you find an error or an omission, or you think that a topic needs further development, we want to hear from you. Forward your feedback to:

documentation@brocade.com

Provide the title and version number of the document and as much detail as possible about your comment, including the topic heading and page number and your suggestions for improvement.

xvi Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 17

Chapter

Introduction

In this chapter

•

Troubleshooting overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

•

Most common problem areas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

•

Questions for common symptoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

•

Gathering information for your switch support provider . . . . . . . . . . . . . . . . . 5

•

Building a case for your switch support provider . . . . . . . . . . . . . . . . . . . . . . 7

Troubleshooting overview

This book is a companion guide to be used in conjunction with the Fabric OS Administrator’s Guide. Although it provides a lot of common troubleshooting tips and techniques, it does not teach troubleshooting methodology.

Troubleshooting should begin at the center of the SAN — the fabric. Because switches are located between the hosts and storage devices and have visibility into both sides of the storage network, starting with them can help narrow the search path. After eliminating the possibility of a fault within the fabric, see if the problem is on the storage side or the host side, and continue a more detailed diagnosis from there. Using this approach can quickly pinpoint and isolate problems.

For example, if a host cannot detect a storage device, run the switchShow command to determine if the storage device is logically connected to the switch. If not, focus first on the switch directly connecting to storage. Use your vendor-supplied storage diagnostic tools to better understand why it is not visible to the switch. If the storage can be detected by the switch, and the host still cannot detect the storage device, then there is still a problem between the host and the switch.

Network Time Protocol

One of the most frustrating parts of troubleshooting is trying to synchronize a switch’s message logs and portlogs with other switches in the fabric. If you do not have Network Time Protocol (NTP) set up on your switches, then trying to synchronize log files to track a problem is more difficult.

Fabric OS Troubleshooting and Diagnostics Guide 1 53-1003141-01

Page 18

Most common problem areas

Tab le 1 identifies the most common problem areas that arise within SANs and identifies tools to

use to resolve them.

TABLE 1

Problem area Investigate Tools

Fabric

Common troubleshooting problems and tools

•

Missing devices

•

Marginal links (unstable connections)

•

Incorrect zoning configurations

•

Incorrect switch configurations

•

Storage Devices

Hosts

Storage Managemen t Applications

Physical issues between switch and devices

•

Incorrect storage software configurations

•

Physical issues between switch and devices

•

Downgrade HBA firmware

•

Incorrect device driver installation

•

Incorrect device driver configuration

•

Incorrect installation and configuration of the storage devices that the software references.

For example, if using a volume-management application, check for:

Incorrect volume installation

Incorrect volume configuration

•

Also, make sure you use the latest HBA firmware recommended by the switch supplier or on the HBA supplier's website

•

Switch LEDs Switch commands (for example,

switchShow or nsAllShow) for diagnostics Web or GUI-based monitoring and management software tools

Device LEDs Storage diagnostic tools Switch commands (for example, switchShow or nsAllShow) for diagnostics

Device LEDs Host operating system diagnostic tools Device driver diagnostic tools Switch commands (for example, switchShow or nsAllShow) for

diagnostics

Application-specific tools and resources

Questions for common symptoms

You first must determine what the problem is. Some symptoms are obvious, such as the switch rebooted without any user intervention, or more obscure, such as your storage is having intermittent connectivity to a particular host. Whatever the symptom is, you must gather information from the devices that are directly involved in the symptom.

Tab le 2 lists common symptoms and possible areas to check. You may notice that an intermittent

connectivity problem has lots of variables to look into, such as the type of connection between the two devices, how the connection is behaving, and the port type involved.

2 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 19

Questions for common symptoms

TABLE 2

Common symptoms

Symptom Areas to check Chapter or Document

Blade is faulty Firmware or application download

Hardware connections

Blade is stuck in the “LOADING” state Firmware or application download Chapter 5, “Firmware Download Errors”

Configuration upload or download fails FTP or SCP server or USB availability Chapter 4, “Configuration”

E_Port failed to come online Correct licensing

Fabric parameters Zoning

EX_Port does not form Links Chapter 3, “Connectivity”

Fabric merge fails Fabric segmentation Chapter 2, “General Troubleshooting”

Fabric segments Licensing

Zoning Virtual Fabrics Fabric parameters

FCIP tunnel bounces FCIP tunnel, including the network between FCIP

tunnel endpoints

FCIP tunnel does not come online FCIP tunnel, including the network between FCIP

tunnel endpoints

FCIP tunnel does not form Licensing

Fabric parameters

FCIP tunnel is sluggish FCIP tunnel, including the network between FCIP

tunnel endpoints

Feature is not working Licensing Chapter 2, “General Troubleshooting”

FCR is slowing down FCR LSAN tags Chapter 2, “General Troubleshooting”

FICON switch does not talk to hosts FICON settings FICON Administrator’s Guide

Firmware download fails FTP or SCP server or USB availability

Firmware version compatibility Unsupported features enabled Firmware versions on switch

Host application times out FCR LSAN tags

Marginal links

Intermittent connectivity Links

Tru nki ng Buffer credits FCIP tunnel

LEDs are flashing Links Chapter 3, “Connectivity”

LEDs are steady Links Chapter 3, “Connectivity”

Chapter 2, “General Troubleshooting” Chapter 5, “Firmware Download Errors” Chapter 7, “Virtual Fabrics”

Chapter 2, “General Troubleshooting” Chapter 3, “Connectivity” Chapter 7, “Virtual Fabrics” Chapter 9, “Zoning”

Chapter 7, “Virtual Fabrics”

Chapter 3, “Connectivity” Chapter 7, “Virtual Fabrics” Chapter 9, “Zoning”

Chapter 2, “General Troubleshooting” Chapter 3, “Connectivity” Chapter 7, “Virtual Fabrics” Chapter 9, “Zoning”

Fibre Channel over IP Administrator’s Guide

Chapter 2, “General Troubleshooting”

Fibre Channel over IP Administrator’s Guide

Chapter 5, “Firmware Download Errors” Chapter 7, “Virtual Fabrics”

Chapter 2, “General Troubleshooting” Chapter 3, “Connectivity”

Chapter 3, “Connectivity” Chapter 8, “ISL Trunking”

Fibre Channel over IP Administrator’s Guide

Fabric OS Troubleshooting and Diagnostics Guide 3 53-1003141-01

Page 20

Questions for common symptoms

TABLE 2

Common symptoms (Continued)

Symptom Areas to check Chapter or Document

License issues Licensing Chapter 2, “General Troubleshooting”

LSAN is slow or times-out LSAN tagging Chapter 2, “General Troubleshooting”

Marginal link Links Chapter 3, “Connectivity”

No connectivity between host and storage Cables

SCSI timeout errors SCSI retry errors Zoning

No connectivity between switches Licensing

Fabric parameters Segmentation Virtual Fabrics Zoning, if applicable

No light on LEDs Links Chapter 3, “Connectivity”

Performance problems Links

FCR LSAN tags FCIP tunnels

Port cannot be moved Virtual Fabrics Chapter 7, “Virtual Fabrics”

SCSI retry errors Buffer credits

FCIP tunnel bandwidth

SCSI timeout errors Links

HBA Buffer credits FCIP tunnel bandwidth

Switch constantly reboots Rolling Reboot Detection

FIPS

Switch is unable to join fabric Security policies

Zoning Fabric parameters

Switch reboots during configup/download Configuration file discrepancy Chapter 4, “Configuration”

Syslog messages Hardware

SNMP management station

Trunk bounces Cables are on same port group

SFPs Tru nked ports

Trunk failed to form Licensing

Cables are on same port group SFPs Tru nked ports Zoning E_Port QoS configuration mismatch

User forgot password Password recovery Chapter 6, “Security”

Chapter 3, “Connectivity” Chapter 8, “ISL Trunking” Chapter 9, “Zoning”

Fibre Channel over IP Administrator’s Guide

Chapter 2, “General Troubleshooting” Chapter 3, “Connectivity”

Chapter 7, “Virtual Fabrics” Chapter 9, “Zoning”

Chapter 3, “Connectivity” Chapter 2, “General Troubleshooting”

Fibre Channel over IP Administrator’s Guide

Chapter 3, “Connectivity” Chapter 8, “ISL Trunking”

Fibre Channel over IP Administrator’s Guide

Chapter 6, “Security”

Chapter 3, “Connectivity” Chapter 7, “Virtual Fabrics” Chapter 9, “Zoning”

Chapter 2, “General Troubleshooting” Chapter 6, “Security”

Chapter 8, “ISL Trunking”

Chapter 2, “General Troubleshooting” Chapter 3, “Connectivity”

Chapter 8, “ISL Trunking” Chapter 9, “Zoning”

4 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 21

Gathering information for your switch support provider

NOTE

TABLE 2

Symptom Areas to check Chapter or Document

User is unable to change switch settings RBAC settings

Virtual Fabric does not form FIDs Chapter 7, “Virtual Fabrics”

Zone configuration mismatch Effective configuration Chapter 9, “Zoning”

Zone content mismatch Effective configuration Chapter 9, “Zoning”

Zone type mismatch Effective configuration Chapter 9, “Zoning”

Common symptoms (Continued)

Chapter 6, “Security”

Account settings

Gathering information for your switch support provider

If you are troubleshooting a production system, you must gather data quickly. As soon as a problem is observed, perform the following tasks. For more information about these commands and their operands, refer to the Fabric OS Command Reference.

1. Enter the supportSave command to save RASlog, TRACE, supportShow, core file, FFDC data, and other support information from the switch, chassis, blades, and logical switches.

2. Gather console output and logs.

To execute the supportSave command on the chassis, you must log in to the switch on an account with the admin role that has the chassis role permission.

Setting up your switch for FTP

1. Connect to the switch and log in using an account with admin permissions.

2. Enter the supportFtp command and respond to the prompts.

Example of supportFTP command

switch:admin> supportftp -s Host IP Addr[1080::8:800:200C:417A]: User Name[njoe]: userFoo Password[********]: <hidden> Remote Dir[support]: supportftp: parameters changed

Refer to “Automatic trace dump transfers” on page 100 for more information on setting up for automatic transfer of diagnostic files as part of standard switch configuration.

Using the supportSave command

The supportSave command uses the default switch name to replace the chassis name regardless of whether the chassis name has been changed to a non-factory setting. If Virtual Fabrics is enabled, the supportSave command uses the default switch name for each logical fabric.

Fabric OS Troubleshooting and Diagnostics Guide 5 53-1003141-01

Page 22

Gathering information for your switch support provider

1. Connect to the switch and log in using an account with admin permissions.

2. Enter the appropriate supportSave command based on your needs:

•

If you are saving to an FTP or SCP server, use the following syntax:

supportSave [-n] [-c]

When invoked without operands, this command goes into interactive mode. The following operands are optional:

-n—Does not prompt for confirmation. This operand is optional; if omitted, you are prompted for confirmation.

-c—Uses the FTP parameters saved by the supportFtp command. This operand is optional; if omitted, specify the FTP parameters through command line options or interactively. To display the current FTP parameters, run supportFtp (on a dual-CP system, run supportFtp on the active CP).

•

On platforms that support USB devices, you can use your Brocade USB device to save the support files. To use your USB device, use the following syntax:

supportsave [-U -d remote_dir]

-U—Saves support data to an attached USB device. When using this option, a target directory must be specified with the -d option.

-d remote_dir—Specifies the remote directory to which the file is to be transferred. When saving to a USB device, the remote directory is created in the /support directory of the USB device by default.

Changing the supportSave timeout value

While running the supportSave command, you may encounter a timeout. A timeout occurs if the system is in a busy state due to the CPU or I/O bound from a lot of port traffic or file access. A timeout can also occur on very large machine configurations or when the machine is under heavy usage. If this occurs, an SS-1004 message is generated to both the console and the RASlog to report the error. You must rerun the supportSave command with the -t option.

Example of SS-1004 message:

SS-1004: “One or more modules timed out during supportsave. Please retry supportsave with -t option to collect all logs.”

Use this feature when you observe that supportSave has timed out.

1. Connect to the switch and log in using an account with admin permissions.

2. Enter the supportSave command with the -t operand, and specify a value from 1 through 5.

The following example increases the supportSave modules timeout to two times of the original timeout setting.

switch:admin> supportSave –t 2

6 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 23

Building a case for your switch support provider

Capturing output from a console

Some information, such as boot information is only outputted directly to the console. To capture this information, you must connect directly to the switch through its management interface, either a serial cable or an RJ-45 connector that is specifically used for Ethernet connection to the management network.

1. Connect directly to the switch using hyperterminal.

2. Log in to the switch using an account with admin permissions.

3. Set the utility to capture output from the screen.

Some utilities require this step to be performed prior to opening up a session. Check with your utility vendor for instructions.

4. Enter the command or start the process to capture the required data on the console.

Capturing command output

1. Connect to the switch through a Telnet or SSH utility.

2. Log in using an account with admin permissions.

3. Set the Telnet or SSH utility to capture output from the screen.

Some Telnet or SSH utilities require this step to be performed prior to opening up a session. Check with your Telnet or SSH utility vendor for instructions.

4. Enter the command or start the process to capture the required data on the console.

Building a case for your switch support provider

The questions listed in “Basic information” should be printed out and answered in their entirety and be ready to send to your switch support provider when you contact them. Having this information immediately available expedites the information-gathering process that is necessary to begin determining the problem and finding a solution.

Basic information

1. What is the switch’s current Fabric OS level?

To determine the switch’s Fabric OS level, enter the firmwareShow command and write down the information.

2. What is the switch model?

To determine the switch model, enter the switchShow command and write down the value in the switchType field. Cross-reference this value with the chart located in Appendix A, “Switch

Type and Blade ID”.

3. Is the switch operational? Yes or no.

4. Impact assessment and urgency:

•

Is the switch down? Yes or no.

•

Is it a standalone switch? Yes or no.

Fabric OS Troubleshooting and Diagnostics Guide 7 53-1003141-01

Page 24

Building a case for your switch support provider

•

Are there VE, VEX, or EX ports connected to the chassis? Yes or no.

Use the switchShow command to determine the answer.

•

How large is the fabric?

Use the nsAllShow command to determine the answer.

•

Do you have encryption blades or switches installed in the fabric? Yes or no.

•

Do you have Virtual Fabrics enabled in the fabric? Yes or no.

Use the switchShow command to determine the answer.

•

Do you have IPsec installed on the switch’s Ethernet interface? Yes or no.

Use the ipsecConfig

•

Do you have In-band Management installed on the switch’s Gigabit Ethernet ports? Yes or no.

Use the portShow iproute geX command to determine the answer.

•

Are you using NPIV? Yes or no.

Use the switchShow command to determine the answer.

•

Are there security policies turned on in the fabric? If so, what are they? Gather the output from the following commands:



secPolicyShow



fddCfg --showall



ipFilter --show



authUtil --show



secAuthSecret --show



fipsCfg --showall

5. Is the fabric redundant? If yes, what is the MPIO software? (List vendor and version.)

6. If you have a redundant fabric, did a failover occur? To verify, view the RASlogs on both CPs and look for messages related to haFailover, for example, HAM-1004.

7. Was POST enabled on the switch? Use the diagPost command to verify if POST is enabled or not.

show command to determine the answer.

8. Which CP blade was active? (Only applicable to Brocade DCX, DCX-4S, and DCX 8510 family enterprise-class platforms.) Use the haShow command in conjunction with the RASlogs to determine which is the active and standby CP. They will reverse roles in a failover and their logs are separate.

Detailed problem information

Obtain as much of the following informationas possible prior to contacting the SAN technical support vendor.

Document the sequence of events by answering the following questions:

•

When did the problem occur?

•

Is this a new installation?

•

How long has the problem been occurring?

8 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 25

Building a case for your switch support provider

•

Are specific devices affected?

If so, what are their World Wide Number Names?

•

What happened prior to the problem?

•

Is the problem reproducible?

If so, what are the steps to reproduce the problem?

•

What configuration was in place when the problem occurred?

•

A description of the problem with the switch or the fault with the fabric.

•

The last actions or changes made to the system environment:

Settings

supportSave output

•

Host information:

OS version and patch level

HBA type

HBA firmware version

HBA driver version

Configuration settings

•

Storage information:

Disk/tape type

Disk/tape firmware level

Controller type

Controller firmware level

Configuration settings

Storage software (such as EMC Control Center, Veritas SPC, and so on.)

•

If this is a Brocade DCX, DCX-4S, or DCX 8510 family enterprise-class platform,, are the CPs in-sync? Yes or no.

Use the haShow command to determine the answer.

•

List out when and what were the last actions or changes made to the switch, the fabric, and the SAN or metaSAN.

•

In Tabl e 3, list the environmental changes added to the network.

TABLE 3

Type of Change Date when change occurred

Fabric OS Troubleshooting and Diagnostics Guide 9 53-1003141-01

Environmental changes

Page 26

Building a case for your switch support provider

Gathering additional information

The following features that require you to gather additional information. The additional information is necessary in order for your switch support provider to effectively and efficiently troubleshoot your issue. Refer to the chapter or document specified for the commands used for the data you must capture:

•

Configurations, refer to Chapter 3, “Connectivity”.

•

Firmware download, refer to Chapter 5, “Firmware Download Errors”.

•

Tru nking , refer to Chapter 8, “ISL Trunking”.

•

Zoning, refer to Chapter 9, “Zoning”.

•

FCIP tunnels, refer to the Fibre Channel over IP Administrator’s Guide.

•

FICON, refer to the FICON Administrator’s Guide.

10 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 27

Chapter

General Troubleshooting

In this chapter

•

Licenses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

•

Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

•

Frame Viewer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

•

Switch message logs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

•

Switch boot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

•

FC-FC routing connectivity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Licenses

Some features require licenses in order to work properly. To view a list of features and their associated licenses, refer to the Fabric OS Administrator’s Guide. Licenses are created using a switch’s License Identifier so you cannot apply one license to different switches. Before calling your switch support provider, verify that you have the correct licenses installed by using the licenseShow command.

Time

Symptom A feature is not working.

Probable cause and recommended action

Refer to the Fabric OS Administrator’s Guide to determine if the appropriate licenses are installed on the local switch and any connecting switches.

Determining installed licenses

1. Connect to the switch and log in using an account with admin permissions.

2. Enter the licenseShow command.

A list of the currently installed licenses on the switch is displayed.

Symptom Time is not in-sync.

Probable cause and recommended action

NTP is not set up on the switches in your fabric. Set up NTP on your switches in all fabrics in your SAN and metaSAN.

For more information on setting up NTP, refer to the Fabric OS Administrator’s Guide.

Fabric OS Troubleshooting and Diagnostics Guide 11 53-1003141-01

Page 28

Frame Viewer

Symptom Frames are being dropped.

When a frame is unable to reach its destination due to timeout, unroutable, or destination unreachable, it is discarded. You can use Frame Viewer to find out the flows that contained the dropped frames, which can help you determine which applications might be impacted. Using Frame Viewer, you can see exactly what time the frames were dropped. (Timestamps are accurate to within one second.) Additionally, this assists in the debug process.

You can view and filter up to 20 discarded frames per chip per second for 1200 seconds using a number of fields with the framelog command.

Probable cause and recommended action

Frames are timing out.

Viewing frames

1. Connect to the switch and log in using an account with admin permissions.

2. Enter the framelog --show command.

Switch message logs

Switch message logs (RAS logs) contain information on events that happen on the switch or in the fabric. This is an effective tool in understanding what is going on in your fabric or on your switch. RAS logs are independent on director class switches. Weekly review of the RAS logs is necessary to prevent minor problems from becoming larger issues, or in catching problems at an early stage. There are two sets of logs. The ipAddrShow command provides the IP addresses of the CP0 and CP1 control processor blades and associated RAS logs.

The following common problems can occur with or in your system message log.

Symptom Inaccurate information in the system message log.

Probable cause and recommended action

In rare instances, events gathered by the Track Change feature can report inaccurate information to the system message log.

For example, a user enters a correct user name and password, but the login was rejected because the maximum number of users had been reached. However, when looking at the system message log, the login was reported as successful.

If the maximum number of switch users has been reached, the switch still performs correctly, in that it rejects the login of additional users, even if they enter the correct user name and password information.

12 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 29

However, in this limited example, the Track Change feature reports this event inaccurately to the system message log; it appears that the login was successful. This scenario only occurs when the maximum number of users has been reached; otherwise, the login information displayed in the system message log reflects reality.

Refer to the Fabric OS Administrator’s Guide for information regarding enabling and disabling Track Changes (TC).

Symptom MQ errors are appearing in the switch log.

Probable cause and recommended action

An MQ error is a message queue error. Identify an MQ error message by looking for the two letters MQ followed by a number in the error message:

2004/08/24-10:04:42, [MQ-1004], 218,, ERROR, ras007, mqRead, queue = raslog-test- string0123456-raslog, queue I D = 1, type = 2

MQ errors can result in devices dropping from the switch’s Name Server or can prevent a switch from joining the fabric. MQ errors are rare and difficult to troubleshoot; resolve them by working with the switch supplier. When encountering an MQ error, issue the supportSave command to capture debug information about the switch; then, forward the supportSave data to the switch supplier for further investigation.

Switch message logs

Symptom I

C bus errors are appearing in the switch log.

Probable cause and recommended action

C bus errors generally indicate defective hardware or poorly seated devices or blades; the specific item is listed in the error message. Refer to the Fabric OS Message Reference for information specific to the error that was received. Some Chip-Port (CPT) and Environmental Monitor (EM) messages contain I

If the I

C message does not indicate the specific hardware that may be failing, begin debugging the

C-related information.

hardware, as this is the most likely cause.

Symptom Core file or FFDC warning messages appear on the serial console or in the system log.

Probable cause and recommended action

Issue the supportSave command. The messages can be dismissed by issuing the supportSave -R command after all data is confirmed to be collected properly.

Error example:

*** CORE FILES WARNING (10/22/08 - 05:00:01 ) *** 3416 KBytes in 1 file(s) use "supportsave" command to upload

Fabric OS Troubleshooting and Diagnostics Guide 13 53-1003141-01

Page 30

Switch boot

ATTENTION

Switch boot

Symptom The enterprise-class platform model rebooted again after an initial bootup.

Probable cause and recommended action

This issue can occur during an enterprise-class platform bootup with two CPs. If any failure occurs on the active CP, before the standby CP is fully functional and has obtained HA sync, the standby CP may not be able to take on the active role to perform failover successfully.

In this case, both CPs reboot to recover from the failure.

Rolling Reboot Detection

A rolling reboot occurs when a switch or enterprise-class platform has continuously experienced unexpected reboots. This behavior is continuous until the rolling reboot is detected by the system. Once the Rolling Reboot Detection (RRD) occurs, the switch is put into a stable state so that only minimal supportSave output need be collected and sent to your service support provider for analysis. USB is also supported in RRD mode. The USB device can be enabled by entering usbstorage -e and the results ccollected by entering supportsave -U -d MySupportSave.Not every reboot activates the Rolling Reboot Detection feature.

If a rolling reboot is caused by a Linux kernel panic, then the RRD feature is not activated.

Reboot classification

There are two types of reboots that occur on a switch and enterprise-class platform: expected and unexpected. Expected reboots occur when the reboots are initialized by commands; these types of reboots are ignored by the Rolling Reboot Detection (RRD) feature. They include the following commands:

•

reboot

•

haFailover

•

fastBoot

•

firmwareDownload

The RRD feature is activated and halts rebooting when an unexpected reboot reason is shown continuously in the reboot history within a certain period of time. The period of time depends on the switch. The following reboots are considered unexpected reboots:

•

Reset

A reset reboot may be caused by one of the following:

Power-cycle of the switch or CP

Linux reboot command

Hardware watchdog timeout

Heartbeat loss-related reboot

14 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 31

Switch boot

•

Software Fault: Kernel Panic

If the system detects an internal fatal error from which it cannot safely recover, it outputs an error message to the console, dumps a stack trace for debugging, and then performs an automatic reboot.

After a kernel panic, the system may not have enough time to write the reboot reason, causing the reboot reason to be empty. This is treated as a reset case.

•

Software fault

Software Watchdog

ASSERT

•

Software recovery failure

This is an HA bootup-related issue and happens when a switch is unable to recover to a stable state. The HASM log contains more details and specific information on this type of failure, such as one of the following:

Failover recovery failed: Occurs when failover recovery fails and the CP must reboot.

Failover when standby CP unready: Occurs when the active CP must fail over, but the standby CP is not ready to take over mastership.

Failover when LS trans incomplete: Takes place when a logical switch transaction is incomplete.

•

Software bootup failure

This is an HA bootup-related issue and happens when a switch is unable to load the firmware to a usable state. The HASM log contains more details and specific information on this type of failure, such as one of the following:

System bring up timed out: The CP failed to come up within the time allotted.

LS configuration timed out and failed: The logical switch configuration failed and timed out.

After RRD is activated, the admin level permission is required to log in. Enter the supportShow or supportSave command to collect a limited amount of data to resolve the issue.

Restrictions

The following restrictions are applicable on the RRD feature:

•

RRD works only on CFOS-based systems and is not available on AP blades.

•

If FIPS mode is enabled, then the RRD feature works in record-only mode.

•

RRD relies on the bootprom and the Linux kernel working properly.

•

RRD only works during the 30 minutes immediately after the switch boots. If the switch does not reboot for 30 minutes, then RRD is deactivated.

Collecting limited supportSave output on the Rolling Reboot Detection

1. Log in to the switch on the admin account.

A user account with admin privileges is not able to collect limited supportSave output.

2. After you see the message in the following example, press Enter.

Fabric OS Troubleshooting and Diagnostics Guide 15 53-1003141-01

Page 32

FC-FC routing connectivity

ATTENTION

3. Enter the supportSave command to go into interactive mode. Alternatively, if using a USB

device, enter usbstorage -e.

4. Respond to the prompts.

5. Once the supportSave command had completed, contact your service support provider to

provide them with the data.

The following message is an example of the screen on a Brocade DCX.

Fabos Version 7.1.0_main_bld23 switch login: admin Password: ************************************************************** * * * Fabric OS has detected frequent switch reboot condition. * * Following actions can be taken to recover the switch: * * - take off or replace the bad blades. * * - use supportsave to collect supportsave data. * * * * **************************************************************

Please change passwords for switch default accounts now. Use Control-C to exit or press 'Enter' key to proceed.

FC-FC routing connectivity

This section describes tools you can use to troubleshoot Fibre Channel routing connectivity and performance.

Generating and routing an ECHO

The FC-FC Routing Service enables you to route the ECHO generated when an fcPing command is issued on a switch, providing fcPing capability between two devices in different fabrics across the FC router.

The fcPing command sends a Fibre Channel ELS ECHO request to a pair of ports. It performs a zone check between the source and destination. In addition, two Fibre Channel Extended Link Service (ELS) requests are generated. The first ELS request is from the domain controller to the source port identifier. The second ELS request is from the domain controller to the destination port identifiers. The ELS ECHO request elicits an ELS ECHO response from a port identifier in the fabric and validates link connectivity.

Use the fcPing command to validate link connectivity to a single device or between a pair of devices.

There are some devices that do not support the ELS ECHO request. In these cases, the device either does not respond to the request or send an ELS reject. When a device does not respond to the ELS request, further debugging is required; however, do not assume that the device is not connected.

16 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 33

FC-FC routing connectivity

On the edge Fabric OS switch, make sure that the source and destination devices are properly configured in the LSAN zone before entering the fcPing command. This command performs the following functions:

•

Checks the zoning configuration for the two ports specified.

•

Generates an ELS ECHO request to the source port specified and validates the response.

•

Generates an ELS ECHO request to the destination port specified and validates the response.

switch:admin> fcping 0x020800 22:00:00:04:cf:75:63:85 Source: 0x020800 Destination: 22:00:00:04:cf:75:63:85 Zone Check: Zoned Pinging 0x020800 with 12 bytes of data: received reply from 0x020800: 12 bytes time:1159 usec received reply from 0x020800: 12 bytes time:1006 usec received reply from 0x020800: 12 bytes time:1008 usec received reply from 0x020800: 12 bytes time:1038 usec received reply from 0x020800: 12 bytes time:1010 usec 5 frames sent, 5 frames received, 0 frames rejected,0 frames timeout Round-trip min/avg/max = 1006/1044/1159 usec

Regardless of the device’s zoning configuration, the fcPing command sends the ELS frame to the destination port. A destination device can take any one of the following actions:

•

Send an ELS Accept to the ELS request.

•

Send an ELS Reject to the ELS request.

•

Ignore the ELS request.

For details about the fcPing command, refer to the Fabric OS Command Reference.

Example of one device that accepts the request and another device that rejects the request

switch:admin> fcping 10:00:00:00:c9:29:0e:c4 21:00:00:20:37:25:ad:05 Source: 10:00:00:00:c9:29:0e:c4 Destination: 21:00:00:20:37:25:ad:05 Zone Check: Not Zoned Pinging 10:00:00:00:c9:29:0e:c4 [0x20800] with 12 bytes of data: received reply from 10:00:00:00:c9:29:0e:c4: 12 bytes time:1162 usec received reply from 10:00:00:00:c9:29:0e:c4: 12 bytes time:1013 usec received reply from 10:00:00:00:c9:29:0e:c4: 12 bytes time:1442 usec received reply from 10:00:00:00:c9:29:0e:c4: 12 bytes time:1052 usec received reply from 10:00:00:00:c9:29:0e:c4: 12 bytes time:1012 usec 5 frames sent, 5 frames received, 0 frames rejected, 0 frames timeout Round-trip min/avg/max = 1012/1136/1442 usec Pinging 21:00:00:20:37:25:ad:05 [0x211e8] with 12 bytes of data: Request rejected by 21:00:00:20:37:25:ad:05: Command not supported: time: 1159 usec Request rejected by 21:00:00:20:37:25:ad:05: Command not supported: time: 1006 usec Request rejected by 21:00:00:20:37:25:ad:05: Command not supported: time: 1008 usec Request rejected by 21:00:00:20:37:25:ad:05: Command not supported: time: 1038 usec Request rejected by 21:00:00:20:37:25:ad:05: Command not supported: time: 1010 usec 5 frames sent, 0 frames received, 5 frames rejected, 0 frames timeout Round-trip min/avg/max = 0/0/0 usec

Fabric OS Troubleshooting and Diagnostics Guide 17 53-1003141-01

Page 34

FC-FC routing connectivity

Example using fcPing with a single destination (in this example, the destination is a device node WWN)

switch:admin> fcping 20:00:00:00:c9:3f:7c:b8 Destination: 20:00:00:00:c9:3f:7c:b8 Pinging 20:00:00:00:c9:3f:7c:b8 [0x370501] with 12 bytes of data: received reply from 20:00:00:00:c9:3f:7c:b8: 12 bytes time:825 usec received reply from 20:00:00:00:c9:3f:7c:b8: 12 bytes time:713 usec received reply from 20:00:00:00:c9:3f:7c:b8: 12 bytes time:714 usec received reply from 20:00:00:00:c9:3f:7c:b8: 12 bytes time:741 usec received reply from 20:00:00:00:c9:3f:7c:b8: 12 bytes time:880 usec 5 frames sent, 5 frames received, 0 frames rejected, 0 frames timeout Round-trip min/avg/max = 713/774/880 usec

Superping

Superping refers to the fcPing --allpaths command, which is a diagnostic tool used to test all least-cost ISLs between a source and destination switch. When you run the command, you are provided with a list of all available least-cost paths from a source domain to a destination device. Superping isolates links with potential failures so that you can investigate these ISLs to determine the exact links.

Superping works by sending ECHO frames to a destination device and outputting the status of each ISL it traverses whether or not the response from the destination device is received. Each ECHO frame can choose any path from multiple available paths in the fabric to reach the destination device. This utility allows you to do the following:

•

Run a sanity test that exercises all the ISLs and internal links in different paths that route to the destination device.

•

Determines the least-cost path to aid in designing fabric redundancy.

•

Determines the specific ISLs and internal links with failures.

•

Exercises all ISL links in the base fabric for a logical fabric configuration.

The number of actual paths covered when using the superping feature depends on two other parameters that you can specify optionally. When you issue the fcPing --allpaths command without any other options, Superping covers all ISLs in the routes between source to destination, as shown in Figure 1.

18 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 35

FC-FC routing connectivity

NOTE

FIGURE 1

Superping and fcPing paths

In the following example, Superping is invoked using the fcPing --allpaths command to destination domain 165. The following example displays each hop in (Domain1/Index1-> Domain2/Index2) format. To reach destination domain 165 from source domain 3 there are two unique end-to-end paths. In the first path, the frame traverses from egress port index 205 on source domain 3 to ingress port index 25 on domain 207. On domain 207, the frame traverses from egress port index 42 to ingress port index 3 in domain 101. On domain 101, the frame goes from egress port index 16 to ingress port index 99 on domain 165.

ECP80:FID128:admin> fcping -allpaths 165 Pinging(size:12 bytes) destination domain 165 through all paths

PATH SWITCH1--> SWITCH2--> SWITCH3 SWITCH4 STATUS

---------------------------------------------------------------------------------

1. (3/EMB, 3/205)[128] (207/25,207/42)[128] (101/3,101/16)[128] (165/99,165/0)[128] SUCCESS

2. (3/EMB, 3/204)[128] (207/27,207/42)[128] (101/3,101/16)[128] (165/99,165/0)[128] SUCCESS

Superping can isolate links with failures so that you can further investigate these ISLs to determine the exact links giving the errors.

Superping provides an indication if all ISLs are covered. If all the ISLs are not covered, you can increase the coverage count and maximum retries to transmit, so that complete coverage o f all ISLs is achieved.

Fabric OS Troubleshooting and Diagnostics Guide 19 53-1003141-01

Page 36

FC-FC routing connectivity

Consider the following example in which a few errors are recorded on ISLs 3/205-->2/25, 3/204-->2/27, 2/42-->101/3, and 2/1-->101/8. The potential faulty link is internal port 0/284 on domain 2 with the maximum of 100 percent failure.

ISL COVERAGE

-------------

SNO ISL STATUS

---------------------------------------------------------------------------- 1 3/123[128]--> 165/96[128] SUCCESS(5/5) 2 3/205[128]--> 2/25[128] FAILURE(7/50) 3 3/204[128--> 2/27[128] FAILURE(11/50) 4 165/99[128]--> 101/16[128] SUCCESS(5/5) 6 2/42[128]--> 101/3[128] FAILURE(10/67) 7 2/1[128]--> 101/8[128] FAILURE(8/33)

INTERNAL PORT COVERAGE

-----------------------

SNO DOMAIN INTRNL_PORT STATUS

---------------------------------------------------------------------------- 1 2[128] 0/272 SUCCESS(40/40) 2 2[128] 0/276 SUCCESS(44/44) 3 2[128] 0/280 SUCCESS(30/30) 4 2[128] 0/284 FAILURE(20/20) <== 100% failure

When an ECHO frame is dropped, all the ISLs in the path are marked as failed. It is not possible to determine the exact ISL link that dropped the frame. Because of this, all the ISLs in the path record some failures. The ISL with the actual error has the maximum percentage of failures, as this ISL, when selected in any possible path, causes the ECHO frame to be dropped and accumulates a higher failure percentage.

Restrictions

•

Fabric reconfiguration cannot occur while using the Superping feature. It is assumed that the fabric is stable before the fcPing --allpaths command is executed.

•

The control path for interswitch communication should be available, even if the data path for device-to-device communication may have resource starvation.

•

When executed in a fabric with trunk ports, only the trunk master index is output to the user (for example, individual coverage statistics for each trunk-member are not available).

•

All switches must have Fabric OS v6.3.0 or later.

•

Superping requires that the FC ECHO ELS frame is supported by end-devices.

•

In TI Zones, when failover is disabled and Superping is executed on a destination device included in the TI Zone, then Superping displays failures on all ISLs that are not part of the TI Zone. Also, when Superping is executed on a device that is not present in a TI Zone, failures are shown on all ISLs that are part of any TI Zone.

•

This feature is not supported in interopMode 2 or 3.

•

In frame redirection configurations, where there is a physical host, physical target, virtual initiator, and virtual target; Superping only identifies the path from the physical host to the physical target regardless of whether the data path consists of the path from the physical target to the virtual target through the virtual initiator.

20 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 37

FC-FC routing connectivity

ATTENTION

Routing and statistical information

The pathInfo command displays routing and statistical information from a source port index on the local switch to a destination port index on another switch. This routing information describes the full path that a data stream travels between these ports, including all intermediate switches.

Using the pathInfo command when exchange-based routing is turned on can provide different paths with each attempt.

The routing and statistics information are provided by every switch along the path, based on the current routing-table information and statistics calculated continuously in real time. Each switch represents one hop.

Use the pathInfo command to display routing information from a source port on the local switch to a destination port on another switch. The command output describes the exact data path between these ports, including all intermediate switches.

When using the pathInfo command in Fabric OS v6.3.0 across fabrics connected through an FC router, the command represents backbone information as a single hop. The command captures details about the FC router to which ingress and egress EX_Ports are connected, but it hides the details about the path the frame traverses from the ingress EX_Ports to the egress EX_Ports in the backbone.

To u se pathInfo across remote fabrics, you must specify both the fabric ID (FID) and the domain ID of the remote switch. Optionally, you can specify the source PID and destination PID. You cannot use the pathInfo command to obtain source port information across remote FCR fabrics. When obtaining path information across remote fabrics, the destination switch must be identified by its domain ID. Identifying the switch by name or WWN is not accepted.

Use the pathInfo command to display basic path information to a specific domain in command line mode:

switch:admin> pathinfo 5 Hop In Port Domain ID (Name) Out Port BW Cost

----------------------------------------------------------------------------0 2 1 (sw0) 6 4G 500 1 23 2 (sw0) 8 4G 500 2 4 3 (sw0) 3 4G 500 3 12 4 (sw0) 18 4G 10000 4 4 7 (switch_3) 0 4G 500 5 26 5 (switch_3) E - -

Use the pathInfo command to display basic and extended statistics in interactive mode:

switch:admin> pathinfo Max hops: (1..127) [25] Fabric Id: (1..128) [-1] Domain|Wwn|Name: [] 8 Source port: (0..15) [-1] Destination port: (0..255) [-1] Source pid: (0x0..0xefff00) [ffffffff] 0x061600 Desination pid: (0x0..0xefff00) [0] 0x01f001 Basic stats (yes, y, no, n): [no] y Extended stats (yes, y, no, n): [no] y Trace reverse path (yes, y, no, n): [no] Source route (yes, y, no, n): [no] Timeout: (1000..30000) [10000]

Fabric OS Troubleshooting and Diagnostics Guide 21 53-1003141-01

Page 38

FC-FC routing connectivity

Target port is Embedded Hop In Port Domain ID (Name) Out Port BW Cost

----------------------------------------------------------------------------0 2 1 (sw0) 6 4G 500 1 23 2 (sw0) 8 4G 500 2 4 3 (sw0) 3 4G 500 3 2 4 (sw0) 24 4G 10000 4 3 7 (switch_3) 2 4G 500 5 27 5 (switch_3) 24 - -

Reverse path

6 24 5 (switch_3) 27 4G 500 7 2 7 (switch_3) 3 4G 500 8 24 4 (sw0) 2 4G 500 9 3 3 (sw0) 4 4G 10000 10 8 2 (sw0) 23 4G 500 11 6 1 (sw0) 2 - (output truncated)

For details about the pathInfo command, refer to the Fabric OS Command Reference.

Performance issues

Symptom General slow-down in FCR performance and scalability.

Probable cause and recommended action

As LSAN zone databases get bigger, it takes more switch resources to process them. Use the enforce tag feature to prevent a backbone switch from accepting unwanted LSAN zone databases into its local database.

Symptom Host application times out.

Probable cause and recommended action

The FCR tends to take a long time (more than 5 seconds) to present and set up paths for the proxy devices. Certain hosts are able to do discovery much faster and, as a result, they end up timing out. Use the speed tag feature to always present the target proxy to the host. This helps sensitive hosts to do a quick discovery without timing out or causing an application failure.

22 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 39

Chapter

Connectivity

In this chapter

•

Port initialization and FCP auto-discovery process . . . . . . . . . . . . . . . . . . . . 23

•

Link issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

•

Connection problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

•

Link failures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

•

Marginal links . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

•

Device login issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

•

Media-related issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

•

Segmented fabrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

•

Port mirroring. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

Port initialization and FCP auto-discovery process

The steps in the port initialization process represent a protocol used to discover the type of connected device and establish the port type and port speed. The possible port types are as follows:

•

U_Port—Universal FC port. The base Fibre Channel port type and all unidentified, or uninitiated ports are listed as U_Ports.

•

L_Port/FL_Port—Fabric Loop port. Connects public loop devices.

•

G_Port—Generic port. Acts as a transition port for non-loop fabric-capable devices.

•

E_Port—Expansion port. Assigned to ISL links.

•

F_Port—Fabric port. Assigned to fabric-capable devices.

•

EX_Port—A type of E_Port. It connects a Fibre Channel router to an edge fabric. From the point of view of a switch in an edge fabric, an EX_Port appears as a normal E_Port. It follows applicable Fibre Channel standards as other E_Ports. However, the router terminates EX_Ports rather than allowing different fabrics to merge as would happen on a switch with regular E_Ports.

•

M_Port—A mirror port. A mirror port allows you to configure a switch port to connect to a port to mirror a specific source port and destination port traffic passing though any switch port. This is only supported between F_Ports.

•

VE_Port—A virtual E_Port. A Gigabit Ethernet switch port configured for an FCIP tunnel is called a VE_Port (virtual E_Port). However, with a VEX_Port at the other end, it does not propagate fabric services or routing topology information from one edge fabric to another.

Fabric OS Troubleshooting and Diagnostics Guide 23 53-1003141-01

Page 40

Port initialization and FCP auto-discovery process

•

VEX_Port—A virtual EX_Port. It connects a Fibre Channel router to an edge fabric. From the point of view of a switch in an edge fabric, a VEX_Port appears as a normal VE_Port. It follows the same Fibre Channel protocol as other VE_Ports. However, the router terminates VEX_Ports rather than allowing different fabrics to merge as would happen on a switch with regular VE_Ports.

Figure 2 shows the process behind port initialization. Understanding this process can help you

determine where a problem resides. For example, if your switch cannot form an E_Port, you understand that the process never got to that point or does not recognize the switch as an E_Port. Possible solutions would be to look at licensing and port configuration. Verify that the correct licensing is installed or that the port is not configured as a loop port, a G_Port, or the port speed is not set.

FIGURE 2

The FCP auto-discovery process enables private storage devices that accept the process login (PRLI) to communicate in a fabric.

If device probing is enabled, the embedded port commences with port login (PLOGI) and attempts a PRLI into the device to retrieve information to enter into the Name Server. This enables private devices that do not perform a fabric login (FLOGI), but accept PRLI, to be entered in the Name Server and receive full fabric citizenship.

A fabric-capable device registers information with the Name Server during a FLOGI. These devices typically register information with the Name Server before querying for a device list. The embedded port will still conduct PLOGI and attempt PRLI with these devices.

To display the contents of a switch’s Name Server, use the nsShow or nsAllShow command. For more information about these Name Server commands, refer to the Fabric OS Command Reference.

24 Fabric OS Troubleshooting and Diagnostics Guide

Simple port initialization process

53-1003141-01

Page 41

Link issues

Symptom Port LEDs are flashing.

Symptom Port LEDs are steady.

Symptom No light from the port LEDs.

Link issues

Probable cause and recommended action

Depending on the rate of the flash and the color of the port LED, this could mean several things. To determine what is happening on either your port status LED or power status LED, refer to that switch’s model hardware reference manual. There is a table that describes the purpose of the LEDs and explains the current behavior as well as suggested resolutions.

Probable cause and recommended action

The color of the port LED is important in this instance. To determine what is happening on either your port status LED or power status LED, refer to that switch’s model hardware reference manual. There is a table that describes the purpose of the LEDs and explains the current behavior as well as suggested resolutions.

Probable cause and recommended action

If there is no light coming from the port LED, then no signal is being detected. Check your cable and SFP to determine the physical fault.

Connection problems

Determine if the problem is the target or the host, then continue to divide the suspected problem-path in half until you can pinpoint the problem. One of the most common solutions is zoning. Verify that the host and target are in the same zone. For more information on zoning, refer to Chapter 9, “Zoning”.

Checking the physical connection

•

Check the cables running to and from the host and storage to the switch.

This path includes the patch panel. Verify that none of the cables are damaged, including indentations or bent cable.

•

Check the SFP on the HBAs and switches.

Verify that they are known to be in good working condition. You can do this by swapping the current SFP with a known working SFP.

•

Clean the optics.

There are many kits on the market for cleaning fiber optics. You want to find a product that does not leave residue either from a lint-free wipe or from the solvent.

Fabric OS Troubleshooting and Diagnostics Guide 25 53-1003141-01

Page 42

Connection problems

Checking the logical connection

1. Enter the switchShow command.

2. Review the output from the command and determine if the device successfully logged in to the

switch.

•

A device that is logically connected to the switch is registered as an F_Port, L_Port, E_Port, EX_Port, VE_Port, VEX_Port, or N_Port.

•

A device that is not logically connected to the switch is registered as a G_Port or U_Port, if NPIV is not on the switch.

3. Enter the slotShow -m command to verify that all blades are enabled and not faulty, disabled,

or in some other non-available state.

4. Perform the appropriate actions based on how your missing device is connected:

•

If the missing device is logically connected, proceed to the next troubleshooting procedure (“Checking the Name Server” on page 27).

•

If the missing device is not logically connected, check the device and everything on that side of the data path. Also refer to “Link failures” on page 28 for additional information.

Checking the path includes verifying the following for the host:



The host OS is configured correctly.



The third-party vendor multi-pathing input/output (MPIO) software, if it is being used, is configured correctly.



The HBA and storage device and the driver and firmware are compatible with the switch based on the compatibility matrix.



The driver settings and binaries are up-to-date.



The device Basic Input Output System (BIOS) settings are correct.



The HBA configuration is correct according to manufacturer’s specifications.



The SFPs in the HBA are compatible with the host’s HBA.



The SFP on the switch is compatible with the switch.



The switch settings related to the host are configured correctly.

Checking the path includes the following for the target:



The driver settings and binaries are up-to-date.



The device Basic Input Output System (BIOS) settings are correct.



The HBA configuration is correct according to the manufacturer’s specifications.



The SFPs in the HBA are compatible with the target HBA.



The switch settings related to the target are configured correctly.

Refer to “Checking for a loop initialization failure” on page 29 as the next potential trouble spot.

26 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 43

Connection problems

Checking the Name Server

1. Enter the nsShow command on the switch to determine if the device is attached:

switch:admin> nsshow The Local Name Server has 9 entries {

Type Pid COS PortName NodeName TTL(sec)

*N 021a00; 2,3;20:00:00:e0:69:f0:07:c6;10:00:00:e0:69:f0:07:c6; 895 Fabric Port Name: 20:0a:00:60:69:10:8d:fd NL 051edc; 3;21:00:00:20:37:d9:77:96;20:00:00:20:37:d9:77:96; na FC4s: FCP [SEAGATE ST318304FC 0005]

Fabric Port Name: 20:0e:00:60:69:10:9b:5b NL 051ee0; 3;21:00:00:20:37:d9:73:0f;20:00:00:20:37:d9:73:0f; na FC4s: FCP [SEAGATE ST318304FC 0005]

Fabric Port Name: 20:0e:00:60:69:10:9b:5b NL 051ee1; 3;21:00:00:20:37:d9:76:b3;20:00:00:20:37:d9:76:b3; na FC4s: FCP [SEAGATE ST318304FC 0005]

Fabric Port Name: 20:0e:00:60:69:10:9b:5b NL 051ee2; 3;21:00:00:20:37:d9:77:5a;20:00:00:20:37:d9:77:5a; na FC4s: FCP [SEAGATE ST318304FC 0005]

Fabric Port Name: 20:0e:00:60:69:10:9b:5b NL 051ee4; 3;21:00:00:20:37:d9:74:d7;20:00:00:20:37:d9:74:d7; na FC4s: FCP [SEAGATE ST318304FC 0005]

Fabric Port Name: 20:0e:00:60:69:10:9b:5b NL 051ee8; 3;21:00:00:20:37:d9:6f:eb;20:00:00:20:37:d9:6f:eb; na FC4s: FCP [SEAGATE ST318304FC 0005] Fabric Port Name: 20:0e:00:60:69:10:9b:5b NL 051eef; 3;21:00:00:20:37:d9:77:45;20:00:00:20:37:d9:77:45; na FC4s: FCP [SEAGATE ST318304FC 0005]

Fabric Port Name: 20:0e:00:60:69:10:9b:5b N 051f00; 2,3;50:06:04:82:bc:01:9a:0c;50:06:04:82:bc:01:9a:0c; na FC4s: FCP [EMC SYMMETRIX 5267]

Fabric Port Name: 20:0f:00:60:69:10:9b:5b

2. Look for the device in the Name Server (NS) list, which lists the nodes connected to that

switch. This allows you to determine if a particular node is accessible on the network.

•

If the device is not present in the NS list, the problem is between the device and the switch. There may be a time-out communication problem between edge devices and the name server, or there may be a login issue. First check the edge device documentation to determine if there is a time-out setting or parameter that can be reconfigured. Also, check the port log for NS registration information and FCP probing failures (using the fcpProbeShow command). If these queries do not help solve the problem, contact the support organization for the product that appears to be inaccessible.

•

If the device is listed in the NS, the problem is between the storage device and the host. There may be a zoning mismatch or a host/storage issue. Proceed to Chapter 9, “Zoning”.

3. Enter the portLoginShow command to check the port login status.

Fabric OS Troubleshooting and Diagnostics Guide 27 53-1003141-01

Page 44

Link failures

NOTE

Link failures

4. Enter the fcpProbeShow command to display the FCP probing information for the devices

attached to the specified F_Port or L_Port. This information includes the number of successful logins and SCSI INQUIRY commands sent over this port and a list of the attached devices.

5. Check the port log to determine whether or not the device sent the FLOGI frame to the switch,

and the switch probed the device.

A link failure occurs when a server, storage, or switch device is connected to a switch, but the link between the devices does not come up. This prevents the devices from communicating to or through the switch.

If the switchShow command or LEDs indicate that the link has not come up properly, use one or more of the following procedures.

The port negotiates the link speed with the opposite side. The negotiation usually completes in one or two seconds; however, sometimes the speed negotiation fails.

Determining a successful speed negotiation

Skip this procedure if the port speed is set to a static speed through the portCfgSpeed command.

1. Enter the portCfgShow command to display the port speed settings of all the ports.

2. Enter the switchShow command to determine if the port has module light.

3. Enter the portCfgSpeed command to change the port speed to 1, 2, 4 or 8 Gbps, depending on

what speed can be used by both devices. This should correct the negotiation by setting to one speed.

4. Enter the portLogShow or portLogDump command.

5. Check the events area of the output:

time task event port cmd args

----------------------------------------------------------------14:38:51.976 SPEE sn <Port#> NC 00000001,00000000,00000001 14:39:39.227 SPEE sn <Port#> NC 00000002,00000000,00000001

•

In the event column, sn indicates a speed negotiation.

•

In the cmd column, NC indicates the negotiation has completed.

If these fields do not appear, proceed to the step 6.

6. Correct the negotiation by entering the portCfgSpeed [slotnumber/]portnumber, speed_level

command if the fields in step 5 do not appear.

switch:admin> portcfgspeed Usage: portCfgSpeed PortNumber Speed_Level Speed_Level: 0 - Auto Negotiate 1 - 1Gbps 2 - 2Gbps 4 - 4Gbps 8 - 8Gbps ax - Auto Negotiate + enhanced retries

28 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 45

Link failures

Checking for a loop initialization failure

1. Verify the port is an L_Port.

a. Enter the switchShow command.

b. Check the last field of the output to verify that the switch port indicates an L_Port. If a loop

device is connected to the switch, the switch port must be initialized as an L_Port.

c. Check to ensure that the state is online; otherwise, check for link failures.

Example of an online L_Port

Area Port Media Speed State Proto =====================================

(output truncated)

66 66 -- N8 No_Module 67 67 id AN No_Sync

68 68 id N2 Online L-Port 13 public

2. Verify that loop initialization occurred if the port to which the loop device is attached does not

negotiate as an L_Port.

a. Enter the portLogShow or portLogDump command to display the port log for all ports on

the switch; or if you are looking for a specific port, enter the portLogDumpPort command.

b. Check argument number four for the loop initialization soft assigned (LISA) frame

0x11050100.

switch:admin> portlogdumpport 4 time task event port cmd args

----------------------------------------------------------------11:40:02.078 PORT Rx3 23 20 22000000,00000000,ffffffff, Received LISA frame

11050100

The LISA frame indicates that the loop initialization is complete.

3. Skip point-to-point initialization by using the portCfgLport command.

The switch changes to point-to-point initialization after the LISA phase of the loop initialization. This behavior sometimes causes trouble with old HBAs.

Checking for a point-to-point initialization failure

1. Enter the switchShow command to confirm that the port is active and has a module that is

synchronized.

If a fabric device or another switch is connected to the switch, the switch port must be online.

2. Enter the portLogShow or portLogDump commands.

3. Verify the event area for the port state entry is pstate. The command entry AC indicates that

the port has completed point-to-point initialization.

switch:admin> portlogdumpport 4 time task event port cmd args

------------------------------------------------11:38:21.726 INTR pstate 4 AC

Fabric OS Troubleshooting and Diagnostics Guide 29 53-1003141-01

Page 46

NOTE

Marginal links

4. Skip over the loop initialization phase.

After becoming an active port, the port becomes an F_Port or an E_Port depending on the device on the opposite side. If the opposite device is a host or target device, the port becomes an F_Port. If the opposite device is another switch, the port becomes an E_Port.

If there is a problem with the host or target device, enter portCfgGPort to force the port to try to come up as point-to-point only.

Correcting a port that has come up in the wrong mode

1. Enter the switchShow command.

2. Check the output from the switchShow command and follow the suggested actions in Table 4.

TABLE 4

Output Suggested action

Disabled If the port is disabled because persistent disable or security reasons, attempt to resolve

Bypassed The port may be testing.

Loopback The port may be testing.

E_Port If the opposite side is not another switch, the link has come up in a wrong mode. Check the

F_Port If the opposite side of the link is a private loop device or a switch, the link has come up in a

G_Port The port has not come up as an E_Port or F_Port. Check the output from portLogShow or

L_Port If the opposite side is not a loop device, the link has come up in a wrong mode. Check the

SwitchShow output and suggested action

the issue and then enter the portEnable or, if persistently disabled, portCfgPersistentEnable command.

output from the portLogShow or PortLogDump commands and identify the link initialization stage where the initialization procedure went wrong.

wrong mode. Check the output from portLogShow or PortLogDump commands.

PortLogDump commands and identify the link initialization stage where the initialization procedure went wrong.

output from portLogShow or PortLogDump commands and identify the link initialization stage where the initialization procedure went wrong.

If you are unable to read a portlog dump, contact your switch support provider for assistance.

Marginal links

A marginal link involves the connection between the switch and the edge device. Isolating the exact cause of a marginal link involves analyzing and testing many of the components that make up the link (including the switch port, switch SFP, cable, edge device, and edge device SFP). Troubleshooting a marginal link can involve inspecting the error counters described in

“Troubleshooting a marginal link,” or running diagnostics on a link, a port, or an end-to-end path.

The portLoopbackTest command is used to verify the functional operation of a path on a switch. This test sends frames from a given port’s transmitter and loops them back into the same port’s receiver. The loopback is done at the parallel loopback path. The path traversed in this test does not include the media or the fiber cable.

30 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 47

Marginal links

NOTE

Only one frame is transmitted and received at any given time. An external cable is not required to run this test. The port LEDs flicker green rapidly while the test is running.

Tab le 5 shows the different loopback modes you can use when using portLoopbackTest to test a

marginal link.

TABLE 5

Loopback mode Description

1 Port Loopback (loopback plugs)

2 External Serializer/Deserializer (SerDes) loopback

5 Internal (parallel) loopback (indicates no external equipment)

7 Back-end bypass and port loopback

8 Back-end bypass and SerDes loopback

9 Back-end bypass and internal loopback

Loopback modes

Troubleshooting a marginal link

1. Enter the portErrShow command.

2. Determine whether there is a relatively high number of errors (such as CRC errors or ENC_OUT

errors), or if there are a steadily increasing number of errors to confirm a marginal link. Sample the data every 5 minutes until you see the counters increment.

•

The frames tx and rx are the number of frames being transmitted and received.

•

The crc_err counter are frames with CRC errors. If this counter goes up, then the physical path should be inspected. Check the cables to and from the switch, patch panel, and other devices. Check the SFP by swapping it with a known good working SFP.

If you see this issue on an 8 Gbps blade, use the portCfgFillWord command to reduce EMI.

•

The crc_g_eof counter are frames with CRC errors and a good EOF. The first port detecting a CRC error marks the frame with a bad EOF and passes the frame on to its destination. Subsequent ports in the path also detect the CRC error and the crc_err counter increments on these ports. However, because the first port marked the frame with a bad EOF, the good EOF counter on the subsequent ports does not increment. The marginal link associated with the port with an increasing good EOF counter is the marginal link and the source of the errors.

•

The enc_out are errors that occur outside the frame and usually indicating a bad primitive. To determine if you are having a cable problem, take snapshots of the port errors by using the portErrShow command in increments of 5 to 10 minutes. If you notice the crc_err counter go up, you have a bad or damaged cable, or a bad or damaged device in the path.

ICLs see enc_out errors when ports on one side of the link are disabled.

•

The disc_c3 errors are discarded class 3 errors, which means that the switch is holding onto the frame longer than the hold time allows. One problem this could be related to is ISL oversubscription.

Fabric OS Troubleshooting and Diagnostics Guide 31 53-1003141-01

Page 48

Marginal links

switch:admin> porterrshow

frames enc crc crc too too bad enc disc link loss loss frjt fbsy

tx rx in err g_eof shrt long eof out c3 fail sync sig

============================================================================

0: 665k 7.0k 0 0 0 0 0 0 6 0 0 1 2 0 0 1: 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 2: 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 3: 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 4: 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 5: 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 6: 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 7: 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 8: 78 60 0 0 0 0 0 0 7 0 0 3 6 0 0 9: 12 4 0 0 0 0 0 0 3 0 0 1 2 0 0 10: 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 11: 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 12: 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 13: 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 14: 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 15: 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 16: 665k 7.4k 0 0 0 0 0 0 6 0 0 1 2 0 0

(output truncated)

3. If you suspect a marginal link, isolate the areas by moving the suspected marginal port cable to

a different port on the switch. Reseating of SFPs may also cure marginal port problems.

If the problem stops or goes away, the switch port or the SFP is marginal (proceed to step 6).

If the problem does not stop or go away, refer to step 7.

4. Run portLoopbackTest on the marginal port. You need an adapter to run the loopback test for

the SFP. Otherwise, run the test on the marginal port using the loopback mode lb=5. Use the different modes shown in Table 5 to test the port. Refer to the Fabric OS Command Reference for additional information on this command.

5. Check the results of the loopback test and proceed as follows:

•

If the loopback test failed, the port is bad. Replace the port blade or switch.

•

If the loopback test did not fail, the SFP was bad.

6. Replace the SFP on the marginal port.

7. Perform the following steps to rule out cabling issues:

a. Insert a new cable in the suspected marginal port.

b. Enter the portErrShow command to determine if a problem still exists.

•

If the portErrShow output displays a normal number of generated errors, the issue is solved.

•

If the portErrShow output still displays a high number of generated errors, follow the troubleshooting procedures for the Host or Storage device in the following section,

“Device login issues”.

32 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 49

Device login issues

A correct login is when the port type matches the device type that is plugged in. In the following example, it shows that the device connected to Port 1 is a fabric point-to-point device and it is correctly logged in an F_Port.

switch:admin> switchshow switchName:brcd5300 switchType:64.3 switchState:Online switchMode:Native switchRole:Subordinate switchDomain:1 switchId:fffc01 switchWwn:10:00:00:05:1e:40:ff:c4 zoning:OFF switchBeacon:OFF FC Router:OFF FC Router BB Fabric ID:1

Area Port Media Speed State Proto ===================================== 0 0 -- N8 No_Module 1 1 -- N4 Online FC F-Port 10:00:00:05:1e:8f:c1:31 2 2 -- N8 No_Module 3 3 -- N8 No_Module

(output truncated)

61 61 -- N8 No_Module 62 62 -- N8 No_Module 63 63 -- N8 No_Module 64 64 id N2 Online E-Port 10:00:00:05:1e:34:d0:05 "1_d1" (Trunk master) 65 65 -- N8 No_Module 66 66 -- N8 No_Module 67 67 id AN No_Sync 68 68 id N2 Online L-Port 13 public 69 69 -- N8 No_Module 70 70 -- N8 No_Module 71 71 id N2 Online L-Port 13 public 72 72 -- N8 No_Module 73 73 -- N8 No_Module 74 74 -- N8 No_Module 75 75 -- N8 No_Module 76 76 id N2 Online E-Port 10:00:00:05:1e:34:d0:05 "1_d1" (upstream)(Trunk master) 77 77 id N4 Online F-Port 10:00:00:06:2b:0f:6c:1f 78 78 -- N8 No_Module 79 79 id N2 Online E-Port 10:00:00:05:1e:34:d0:05 "1_d1" (Trunk master)

Device login issues

Pinpointing problems with device logins

1. Log in to the switch as admin.

2. Enter the switchShow command; then, check for state logins.

Fabric OS Troubleshooting and Diagnostics Guide 33 53-1003141-01

Page 50

Device login issues

3. Enter the portCfgShow command to see if the port is configured correctly.

In some cases, you may find that the port has been locked as an L_Port and the device attached is a fabric point-to-point device such as a host or switch. This would be an incorrect configuration for the device and therefore the device cannot log in to the switch.

To correct this type of problem, remove the Lock L_Port configuration using the portCfgDefault command.

switch:admin> portcfgshow Ports of Slot 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

-----------------+--+--+--+--+----+--+--+--+----+--+--+--+----+--+--+-Speed AN AN AN AN AN AN AN AN AN AN AN AN AN AN AN AN Trunk Port ON ON ON ON ON ON ON ON ON ON ON ON ON ON ON ON

Long Distance .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..

VC Link Init .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..

Locked L_Port .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..

Locked G_Port .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..

Disabled E_Port .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..

ISL R_RDY Mode .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..

RSCN Suppressed .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..

Persistent Disable.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..

NPIV capability ON ON ON ON ON ON ON ON ON ON ON ON ON ON ON ON

where AN:AutoNegotiate, ..:OFF, ??:INVALID, SN:Software controlled AutoNegotiation.

4. Enter the portErrShow command; then, check for errors that can cause login problems. A

steadily increasing number of errors can indicate a problem. Track errors by sampling the port errors every five or ten minutes until you see the problem occur again.

5. Enter the portFlagsShow command; then, check to see how a port has logged in and where a

switch:admin> portflagsshow Port SNMP Physical Flags

------------------------------ 0 Offline In_Sync PRESENT U_PORT LED 1 Online In_Sync PRESENT ACTIVE F_PORT G_PORT U_PORT LOGICAL_ONLINE LOGIN NOELP LED ACCEPT 2 Offline No_Light PRESENT U_PORT LED 3 Offline No_Module PRESENT U_PORT LED 4 Offline No_Module PRESENT U_PORT LED 5 Offline No_Light PRESENT U_PORT LED 6 Offline No_Module PRESENT U_PORT LED 7 Offline No_Module PRESENT U_PORT LED 8 Offline No_Light PRESENT U_PORT LED 9 Offline No_Light PRESENT U_PORT LED 10 Offline No_Module PRESENT U_PORT LED 11 Offline No_Module PRESENT U_PORT LED 12 Offline No_Module PRESENT U_PORT LED 13 Offline No_Module PRESENT U_PORT LED 14 Online In_Sync PRESENT ACTIVE F_PORT G_PORT U_PORT LOGICAL_ONLINE LOGIN NOELP LED ACCEPT 15 Online In_Sync PRESENT ACTIVE E_PORT G_PORT U_PORT SEGMENTED LOGIN LED

34 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 51

Media-related issues

NOTE

6. Enter the portLogDumpPort portid command where the port ID is the port number; then, view

the device-to-switch communication.

switch:admin> portlogdumpport 8 | more time task event port cmd args

------------------------------------------------Thu Nov 6 16:52:39 2008 16:52:39.066 PORT scn 8 1 00010004,4302000f,02000000 16:52:39.066 PORT scn 8 2 ce3dfab0,d9672800,00000002 16:52:39.066 PORT scn 8 2 ce3dfab0,d9672800,00000080 16:52:39.066 PORT scn 8 5 00000000,00000000,00000002 16:52:39.066 PORT scn 8 1 00010004,4302000f,00000002 16:52:39.066 PORT scn 8 1 00010004,4302000f,02000000 16:52:39.071 PORT ioctl 88010004 1,0 * 4 16:52:42.311 SPEE sn 8 WS 00000000,00000000,00000000 16:52:42.558 SPEE sn 8 NM 00000000,00000000,00000000 16:52:42.558 SPEE sn 8 NF 00000000,00000000,00000000 16:52:42.558 SPEE sn 8 NC 00000001,00000000,00000000 16:52:42.559 LOOP loopscn 8 LIP 8002 16:52:42.559 LOOP loopscn 8 LIP f7f7 16:52:42.572 LOOP loopscn 8 LIM 0 16:52:42.572 PORT Tx3 8 12 22000000,00000000,ffffffff,11010000 16:52:42.572 PORT Rx3 8 12 22000000,00000000,ffffffff,11010000 16:52:42.572 PORT Tx3 8 20 22000000,00000000,ffffffff,11020000 16:52:42.572 PORT Rx3 8 20 22000000,00000000,ffffffff,11020000 16:52:42.572 PORT Tx3 8 20 22000000,00000000,ffffffff,11030000 16:52:42.572 PORT Rx3 8 20 22000000,00000000,ffffffff,11030000

Refer to “Port log” on page 96 for overview information about portLogDump.

Media-related issues

This section provides procedures that help pinpoint any media-related issues, such as bad cables and SFPs, in the fabric. The tests listed in Tab le 6 are a combination of structural and functional tests that can be used to provide an overview of the hardware components and help identify media-related issues.

•

Structural tests perform basic testing of the switch circuit. If a structural test fails, replace the main board or port blade.

•

Functional tests verify the intended operational behavior of the switch by running frames through ports or bypass circuitry.

TABLE 6

Test name Operands Checks

portTest [-ports itemlist] [-iteration count]

spinFab [-nmegs count] [-ports itemlist] [-setfail mode] Tests switch-to-switch ISL cabling and trunk group

Component test descriptions

[-userdelay time] [-timeout time] [-pattern pattern] [-patsize size] [-seed seed] [-listtype porttype]

Used to isolate problems to a single replaceable element and isolate problems to near-end terminal equipment, far-end terminal equipment, or transmission line. Diagnostics can be executed every day or on demand.

operations.

Fabric OS Troubleshooting and Diagnostics Guide 35 53-1003141-01

Page 52

Media-related issues

The following procedures are for checking switch-specific components.

Testing the external transmit and receive path of a port

1. Connect to the switch and log in as admin.

2. Connect the port you want to test to any other switch port with the cable you want to test.

3. Enter the portLoopbackTest -lb_mode 2 command.

Testing the internal components of a switch

1. Connect to the switch and log in as admin.

2. Connect the port you want to test to any other switch port with the cable you want to test.

3. Enter the portLoopbackTest -lb_mode 5 command where 5 is the operand that causes the test

to run on the internal switch components (this is a partial list—refer to the Fabric OS Command Reference for additional command information):

[-nframes count]—Specify the number of frames to send.

[-lb_mode mode]—Select the loopback point for the test.

[-spd_mode mode]—Select the speed mode for the test.

[-ports itemlist]—Specify a list of user ports to test.

Testing components to and from the HBA

1. Connect to the switch and log in as admin.

2. Enter the portTest command (refer to the Fabric OS Command Reference for information on

the command options).

Refer to Table 7 on page 36 for a list of additional tests that can be used to determine the switch components that are not functioning properly. Refer to the Fabric OS Command Reference for additional command information.

The HBA bcu fcDiag more information on using this command, refer to the Brocade Adapters Administrator’s Guide.

TABLE 7

Test Function

portBeacon Sets port beaconing mode.

portLoopbackTest Performs a functional test of port N to N path. Verifies the functional components of the

turboRamTest Verifies that the on chip SRAM located in the 4 and 8 Gbps ASIC is using the Turbo-Ram BIST

linkbeacon command can be used to beacon a target port on the switch. For

Switch component tests

switch.

circuitry. This allows the BIST controller to perform the SRAM write and read operations at a much faster rate.

36 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 53

Segmented fabrics

Fabric segmentation is generally caused by one of the following conditions:

•

Incompatible fabric parameters (refer to “Reconciling fabric parameters individually” on page 37)

•

Incompatible zoning configuration (refer to Chapter 9, “Zoning”)

•

Domain ID conflict (refer to “Reconciling fabric parameters individually” on page 37)

•

Fabric ID conflict (refer to Chapter 7, “Virtual Fabrics”)

•

Incompatible security policies

•

Incorrect fabric mode

•

Incorrect policy distribution

•

Incompatible software features

There are a number of settings that control the overall behavior and operation of the fabric. Some of these values, such as the domain ID, are assigned automatically by the fabric and can differ from one switch to another in the fabric. Other parameters, such as the BB credit, can be changed for specific applications or operating environments, but must be the same among all switches to allow the formation of a fabric.

Segmented fabrics

The following fabric parameters must be identical on each switch for a fabric to merge:

•

R_A_TOV

•

E_D_TOV

•

Data field size

•

Sequence level switching

•

Suppress class F traffic

•

Per-frame route priority

•

Long-distance fabric (not necessary on Brocade DCX, DCX-4S, 6505, 6510, 6520, and the Brocade DCX 8510 Backbone families; for more information regarding these product types, refer to Appendix A, “Switch Type and Blade ID”.)

Reconciling fabric parameters individually

1. Log in to one of the segmented switches as admin.

2. Enter the configShow -pattern “fabric.ops” command.

3. Log in to another switch in the same fabric as admin.

4. Enter the configShow -pattern “fabric.ops” command.

5. Compare the two switch configurations line by line and look for differences. Do this by

comparing the two Telnet windows or by printing the configShow -pattern “fabric.ops” output. Also, verify that the fabric parameter settings (refer to the list of fabric parameters in

“Segmented fabrics”) are the same for both switches.

6. Connect to the segmented switch after the discrepancy is identified.

7. Disable the switch by entering the switchDisable command.

Fabric OS Troubleshooting and Diagnostics Guide 37 53-1003141-01

Page 54

Segmented fabrics

8. Enter the configure command to edit the appropriate fabric parameters for the segmented

switch.

9. Enable the switch by entering the switchEnable command.

Alternatively, you can reconcile fabric parameters by entering the configUpload command for each switch and upload a known-good configuration file. If you do this option, the two switches must be the same model.

Downloading a correct configuration

You can restore a segmented fabric by downloading a previously saved correct backup configuration to the switch. Downloading in this manner reconciles any discrepancy in the fabric parameters and allows the segmented switch to rejoin the main fabric. For details on uploading and downloading configurations, refer to the Fabric OS Administrator’s Guide.

Reconciling a domain ID conflict

If a domain ID conflict appears, the conflict is only reported at the point where the two fabrics are physically connected. However, there may be several conflicting domain IDs, which appear as soon as the initial conflict is resolved.

Typically, the fabric automatically resolves domain conflicts during fabric merges or builds unless Insistent Domain ID (IDID) is configured. If IDID is enabled, switches that cannot be programmed with a unique domain ID are segmented out. Check each switch that has IDID configured and make sure their domain IDs are unique within the configuration.

Repeat the following procedure until all domain ID conflicts are resolved.

1. Enter the fabricShow command on a switch from one of the fabrics.

2. In a separate Telnet window, enter the fabricShow command on a switch from the second

fabric.

3. Compare the fabricShow output from the two fabrics. Note the number of domain ID conflicts;

there may be several duplicate domain IDs that must be changed. Determine which switches have domain overlap and change the domain IDs for each of those switches.

4. Choose the fabric on which to change the duplicate domain ID; connect to the conflicting

switch in that fabric.

5. Enter the switchDisable command.

6. Enter the configure command.

7. When the Fabric Parameters prompt displays, type y.

8. When the Domain prompt displays, type in the new number.

9. Press Enter on all prompts to accept their default settings.

10. Enter the switchEnable command.

This enables the joining switch to obtain a new domain ID as part of the process of coming online. The fabric principal switch allocates the next available domain ID to the new switch during this process.

11. Repeat step 4 through step 10 if additional switches have conflicting domain IDs.

38 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 55

Example of setting the domain ID

switch_89:FID89:admin> switchdisable switch_89:FID89:admin> configure

Configure...

Fabric parameters (yes, y, no, n): [no] y

Domain: (1..239) [1] 89 WWN Based persistent PID (yes, y, no, n): [no] Allow XISL Use (yes, y, no, n): [yes] R_A_TOV: (4000..120000) [10000] E_D_TOV: (1000..5000) [2000] WAN_TOV: (0..30000) [0] MAX_HOPS: (7..19) [7] Data field size: (256..2112) [2112] Sequence Level Switching: (0..1) [0] Disable Device Probing: (0..1) [0] Suppress Class F Traffic: (0..1) [0] Per-frame Route Priority: (0..1) [0] Long Distance Fabric: (0..1) [0] BB credit: (1..27) [16] Disable FID Check (yes, y, no, n): [no] Insistent Domain ID Mode (yes, y, no, n): [no] Virtual Channel parameters (yes, y, no, n): [no] F-Port login parameters (yes, y, no, n): [no] Zoning Operation parameters (yes, y, no, n): [no] RSCN Transmission Mode (yes, y, no, n): [no] Arbitrated Loop parameters (yes, y, no, n): [no] System services (yes, y, no, n): [no] Portlog events enable (yes, y, no, n): [no] ssl attributes (yes, y, no, n): [no] rpcd attributes (yes, y, no, n): [no] webtools attributes (yes, y, no, n): [no]

Segmented fabrics

WARNING: The domain ID will be changed. The port level zoning may be affected

Reconciling incompatible software features

Earlier releases of software may not be supported in new versions of Fabric OS code. This may be due to a software feature changing or new services being supported. If you suspect that you are trying to introduce a switch into a fabric that has an older version of code, check the release notes to verify that any features on that switch are supported in the fabric with the newer code.

When the Management Server (MS) Platform services are enabled on a switch running Fabric OS v7.0.0 and later and you try to merge this switch into a fabric that does not have this feature enabled, the switch does not merge and a segmentation occurs. To resolve this, either turn the MS Platform services off or enable them on every switch in the fabric.

In Fabric OS v7.0.0 and later, an ESC frame is used to exchange fabric parameters to detect Enhanced TI Zones, interoperability mode, and Virtual Fabric FID conflicts. If at any point during the ESC frame exchange, a link with incompatible parameters is detected, the switch running Fabric OS v7.0.0 and later does not join into the existing fabric. To fix this issue, refer to the Fabric OS Administrator’s Guide for more information on that specific software feature.

Fabric OS Troubleshooting and Diagnostics Guide 39 53-1003141-01

Page 56

Port mirroring

With port mirroring, you can configure a switch port to mirror the traffic between a specific source and destination port. This is only supported between F_Ports. This is a useful way to troubleshoot a problem port without bringing down the host and destination links to insert an inline analyzer.

Port mirroring captures traffic between two devices. It mirrors only the frames containing the SID/DID to the mirror port. Because of the way it handles mirroring, a single mirror port can mirror multiple mirror connections. This also means that the port cannot exceed the maximum bandwidth of the mirror port. Attempts to mirror more traffic than what available bandwidth allows results in the port mirror throttling the SID/DID traffic so that traffic does not exceed the maximum available bandwidth.

The bandwidth of the mirror port is unidirectional. In general, a host (SID) talks to multiple storage devices (DIDs). Thus, a host does not send full line rate to a single target. A mirror port configured at 4 Gbps can only support up to 4 Gbps of traffic. A normal 4 Gbps F_Port is bidirectional and can support up to 8 Gbps (4 Gbps transmit and 4 Gbps receive) of traf fic. If the mirror por t bandwidth is exceeded, no credits are returned to the receiver port and thus those devices involved in the mirror connection see a degraded level of performance.

Use port mirroring to detect missing frames, which may occur with zoning issues or hold timeouts, capture protocol errors, and capture ULP traffic (SCSI/FICON). This feature cannot be used on embedded switch traffic.

In-Order Delivery

If In-Order Delivery (IOD) is enabled, adding or deleting a port mirror connection causes a frame drop. Port mirroring basically reroutes a given connection to the mirror port. The mirror traffic takes an extra route to the mirror port. When the extra route is removed, the frames between the two ports go directly to the destination port. The frames at the mirror port could be queued at the destination port behind those frames that went directly to the destination port. To prevent this IOD issue, port mirroring drops those frames from the mirror port when a connection is disabled. If IOD has been disabled, port mirroring does not drop any frames, but does have an IOD error.

Port mirroring considerations

Before creating port mirror connections, consider the following limitations:

•

A mirror port can be any port on the same switch as the source identifier port.

•

If FCR is enabled, do not enable port mirroring.

•

Only one domain can be mirrored. After a domain is defined, only mirror ports on the defined domain can be used. The first connection defines the restriction on the domain, which can be either the local domain or a remote domain.

•

The mirror port must be attached to one port of the analyzer port pair. The second analyzer port must be set for parallel loopback. Do not attach a N_Port, a NL_Port, or an E_Port to the analyzer port, as the mirror port cannot negotiate linkup with any other port.

•

A switch that is capable of port mirroring can support a minimum of one and a maximum of three mirror connections. Refer to Table 9 on page 42 to determine the number of mirror connections your switch or blade can support.

•

Mirror port bandwidth limits mirror connections.

40 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 57

Port mirroring

•

Deleting a port mirroring connection with IOD enabled causes frame drop between two endpoints.

•

Using the firmware download procedure to downgrade to previous Fabric OS releases that do not support port mirroring requires that you remove all the port mirroring connections. If you downgrade to a previous version of Fabric OS, you cannot proceed until the mirroring connections are removed.

•

Port mirroring is supported with Virtual Fabrics with the limitation that you cannot have FCR enabled within the same 8-port group.

•

If you have NPIV or 10-bit address mode enabled on a Brocade 300, 5300, 5410, 5450, 5460, 5470, 5480, 7800, or the M5424 platforms, all devices from the same NPIV port or 10-bit addressing mode hit the same mirror connection, regardless of different AL_PAs, because the validation is done only for the first 16 bits of the SID/DID.

•

Port mirroring is not supported for the shared area ports of 48-port blades in the default switch. However, when the ports are assigned to a 10-bit address logical switch, port mirroring is supported.

•

For the 48-port and 64-port blades, port mirroring is always supported on the non-shared area ports.

For the shared area ports, port mirroring is supported in Virtual Fabrics mode if the ports are assigned to a 10-bit address logical switch. Port mirroring is not supported on the shared area ports in non-Virtual Fabrics mode.

If a Fabric Vision license is installed and if port mirror connections are defined or enabled on 16 Gbps-capable Fibre Channel ports, flow creation for Flow Mirror will fail.

Similarly, if a Fabric Vision license is installed and a flow is created or activated for Flow Mirror, if you attempt to create a port mirror connection using 16Gbps-capable Fibre Channel ports, the port mirror commands will fail.

Supported platforms

Port mirroring is supported only in FC ports. In general, a platform or blade supporting port mirroring supports both the mirror ports and the mirror connections. Exceptions are listed in

Tab le 8.

TABLE 8

Brocade Model Fabric OS v6.1.1_enc Fabric OS v6.3.0 Fabric OS v6.4.0 Fabric OS v7.1.0

Brocade Encryption Switch

FS8-18

DCX-4S

FCOE10-24 Port mirroring is not

Port mirroring platform supportability

Not supported Not supported Not supported Not supported

Port mirroring is

supported.

supported on the Brocade FCOE10-24.

Port mirroring is supported on the Brocade FCOE10-24.

Fabric OS Troubleshooting and Diagnostics Guide 41 53-1003141-01

Page 58

Port mirroring

TABLE 8

Port mirroring platform supportability (Continued)

Brocade Model Fabric OS v6.1.1_enc Fabric OS v6.3.0 Fabric OS v6.4.0 Fabric OS v7.1.0

FX8-24 Port mirroring is

supported only on the FC ports. It is not supported over GbE ports/FCIP VE tunnels.

Brocade 7800 Port mirroring is

supported only on the FC ports. It is not supported over GbE ports/FCIP VE tunnels.

Port mirroring is supported over GbE ports/FCIP VE tunnels.

Maximum mirror connections

Tab le 9 shows the maximum number of mirror connections you can add to a mirror port.

TABLE 9

Model Maximum number of mirror connections (chassis-wide)

Brocade 300 1

Brocade 5100 3

Brocade 5300 1

Brocade 5410 1

Brocade 5431 1

Brocade 5450 1

Brocade 5460 1

Brocade 5470 1

Brocade 5480 1

Brocade 7800 1

Brocade 6505 3

Brocade 6510 3

Brocade 6520 3

Brocade 6547 3

Brocade DCX 3

Brocade DCX-4S 3

Brocade 8510-4 3

Brocade 8510-8 3

Brocade Encryption Switch 3

Brocade M5424 1

Brocade M6505 3

FC8-16/32 3

Maximum number of mirror connections

42 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 59

Port mirroring

NOTE

TABLE 9

Model Maximum number of mirror connections (chassis-wide)

FS8-18 3

FX8-24 3

Maximum number of mirror connections (Continued)

Configuring a port to be a mirror port

1. Connect to the switch and log in using an account with admin permissions.

2. Enter the portCfg mirrorport [slot number/] port number

The enable command enables the port as a mirror port. The disable command disables the mirror port configuration.

enable command.

Adding a port mirror connection

1. Connect to the switch and log in using an account with admin permissions.

2. Enter the portMirror

The configuration database keeps information about the number of port mirror connections configured on a switch, the number of chunks of port mirroring data that are stored, and the chunk number.

add slotnumber/portnumber SourceID DestID command.

Deleting a port mirror connection

1. Connect to the switch and log in using an account with admin permissions.

2. Enter the portMirror

For example, to delete the port mirror connection on mirror port 2, you could enter the following command:

switch:admin> portMirror --del 0x011400 0x240400

When removing a mirror connection, always use this method to ensure that the data is cleared. Deleting a connection removes the information from the database.

del SourceID DestID command.

Fabric OS Troubleshooting and Diagnostics Guide 43 53-1003141-01

Page 60

Port mirroring

Displaying port mirror connections

1. Connect to the switch and log in using an account with admin permissions.

2. Enter the portMirror

show command.

You should see output similar to the following:

switch:admin> portmirror --show

Number of mirror connection(s) configured: 4

Mirror_Port SID DID State

---------------------------------------18 0x070400 0x0718e2 Enabled 18 0x070400 0x0718e3 Enabled 18 0x070400 0x0718ef Enabled 18 0x070400 0x0718e0 Enabled

44 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 61

Chapter

NOTE

Configuration

In this chapter

•

Configuration upload and download issues . . . . . . . . . . . . . . . . . . . . . . . . . 45

•

Brocade configuration form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

Configuration upload and download issues

It is important to maintain consistent configuration settings on all switches in the same fabric because inconsistent parameters (such as inconsistent PID formats) can cause fabric segmentation. As part of standard configuration maintenance procedures, it is recommended that you back up all important configuration data for every switch on a host computer server for emergency reference.

For information about AD-enabled switches using Fabric OS v5.2.0 or later, refer to the Fabric OS Administrator’s Guide.

For information about Virtual Fabrics using Fabric OS v6.3.0 or later, refer to the Fabric OS Administrator’s Guide.

Symptom The configuration upload fails.

Probable cause and recommended action

If the configuration upload fails, It may be because of one or more of the following reasons:

•

The FTP or SCP server’s host name is not known to the switch.

Verify with your network administrator that the switch has access to the FTP server.

•

The USB path is not correct.

If your platform supports a USB memory device, verify that it is connected and running. Verify that the path name is correct by using the usbStorage -l command.

Example of usbStorage -l command

switch:admin> usbstorage -l firmwarekey\ 0B 2007 Aug 15 15:13 support\ 106MB 2007 Aug 24 05:36

support1034\ 105MB 2007 Aug 23 06:11 config\ 0B 2007 Aug 15 15:13 firmware\ 380MB 2007 Aug 15 15:13

FW_v6.0.0\ 380MB 2007 Aug 15 15:13 Available space on usbstorage 74%

Fabric OS Troubleshooting and Diagnostics Guide 45 53-1003141-01

Page 62

Configuration upload and download issues

•

The FTP or SCP server’s IP address cannot be contacted.

Verify that you can connect to the FTP server. Use your local PC to connect to the FTP server or ping the FTP server.

Example of a successful ping

C:\> ping 192.168.163.50 Pinging 192.168.163.50 with 32 bytes of data: Reply from 192.168.163.50: bytes=32 time=5ms TTL=61 Ping statistics for 192.168.163.50: Packets: Sent = 4, Received = 4, Lost = 0 (0%loss), Approximate round trip times in milli-seconds: Minimum = 4ms, Maximum = 5ms, Average = 4ms

If your ping is successful from your computer, but you cannot reach it from inside your data center, there could be a block on the firewall to not allow FTP connections from inside the data center. Contact your network administrator to determine if this is the cause and to resolve it by opening the port up on both inbound and outbound UDP and TCP traffic.

Example of a failed ping

C:\> ping 192.168.163.50 Pinging 192.168.163.50 with 32 bytes of data: Request timed out. Request timed out. Request timed out. Request timed out. Ping statistics for 192.168.163.50: Packets: Sent = 4, Received = 0, Lost = 4 (100% loss),

If your ping has failed, then you should verify the following:

The ports are open on the firewall.

The FTP server is up and running.

•

You do not have configuration upload permission on the switch.

There may be some restrictions if you are using Admin Domains or Role-Based Access Control. For more information on these types of restrictions, refer to the Fabric OS Administrator’s Guide.

•

You do not have permission to write to the directory on the FTP or SCP server.

Example of a failed login to the FTP server

The output should be similar to the following on an unsuccessful login:

C:\> ftp 192.168.163.50 Connected to 192.168.163.50 220 Welcome to Services FTP service. User (10.10.252.50:(none)): userFoo 331 Please specify the password. Password: 530 Login incorrect. Login failed.

If your login to the FTP or SCP server has failed, verify the user name and password are correct.

46 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 63

•

NOTE

On a Virtual Fabrics-enabled switch, you do not have the chassis role permission set on your user account.

Implement one change at a time, then issue the command again. By implementing one change at a time, you are able to determine what works and what does not work. Knowing which change corrected the problems help you to avoid this problem in future endeavors.

Symptom The configuration download fails.

Probable cause and recommended action

If the configuration download fails, It may be because of one or more of the following reasons:

•

The FTP or SCP server’s host name is not known to the switch.

Verify with your network administrator that the switch has access to the FTP server.

•

The USB path is incorrect.

If your platform supports a USB memory device, verify that it is connected and running. Verify that the path name is correct. It should be the relative path from /usb/usbstorage/brocade/configdownload or use absolute path.

Configuration upload and download issues

Root access is required to see the mentioned path.

•

The FTP or SCP server’s IP address cannot be contacted.

Verify that you can connect to the FTP server. Use your local PC to connect to the FTP server or ping the FTP server.

•

There was a reason to disable the switch.

Note, however, that you must disable the switch for some configuration downloads. For more information on how to perform a configuration download without disabling a switch, refer to the Fabric OS Administrator’s Guide.

•

You do not have permission on the host to perform configuration download.

There may be some restrictions if you are using Admin Domains or Role-Based Access Control. For more information on these types of restrictions, refer to the Fabric OS Administrator’s Guide.

•

The configuration file you are trying to download does not exist on the host.

•

The configuration file you are trying to download is not a switch configuration file.

•

If you selected the (default) FTP protocol, the FTP server is not running on the host.

•

The configuration file that you are trying to download uses incorrect syntax.

•

The user name and password are incorrect.

Symptom The switch reboots during the configuration download.

Probable cause and recommended action

If you are issuing the command with the -vf option, the rebooting is normal. You can continue with the instructions. Otherwise, issue the command again as follows.

•

Enter the configDownload -vf command to download Virtual Fabrics-related data. This causes both CPs to reboot.

Fabric OS Troubleshooting and Diagnostics Guide 47 53-1003141-01

Page 64

Brocade configuration form

•

Enter the configDownload command, without the -vf operand, to download the regular configuration data. This step does not cause a reboot. However, the configDownload command issues a notice that "A switch reboot is required for the changes to take effect." There is no reliable mechanism to determine, which parameters may or may not require a reboot. To ensure that all configuration changes are applied correctly, Brocade strongly recommends that you always reboot the switch after a configuration download.

Symptom Configuration did not seem to change after the configuration download process finished.

Probable cause and recommended action

Verify that the switch was rebooted by checking the system log. If you are doing this on a enterprise-class platform, verify that both CPs rebooted by checking the system log.

If any error occurs during the download, such as an error about a particular key, it is important to issue the configDefault command and attempt to repeat the configDownload command.

Gathering additional information

Be sure to capture the output from the commands you are issuing both from the switch and from your computer when you are analyzing the problem.

Send this and all logs to your switch support provider.

Messages captured in the logs

Configuration download generates both RASlog and Audit log messages resulting from execution of the configDownload command.

The following messages are written to the logs:

•

configDownload completed successfully … (RASlog and Audit log)

•

configUpload completed successfully … (RASlog)

•

configDownload not permitted … (Audit log)

•

configUpload not permitted … (RASlog)

•

(Warning) Downloading configuration without disabling the switch was unsuccessful. (Audit log)

Brocade configuration form

Use this form in Table 10as a hard copy reference for your configuration information.

In the hardware reference manuals for the Brocade DCX, and DCX-4S modular switches, there is a guide for FC port setting tables. Print out Table 10 and use it to record configuration information for the various blades.

TABLE 10

Brocade configuration settings Value

IP address

Gateway address

48 Fabric OS Troubleshooting and Diagnostics Guide

Brocade configuration and connection

53-1003141-01

Page 65

Brocade configuration form

TABLE 10

Brocade configuration settings Value

Chassis configuration option

Management connections

Serial cable tag

Ethernet cable tag

Configuration information

Domain ID

Switch name

Ethernet IP address

Ethernet subnet mask

Total number of local devices (nsShow)

Total number of devices in fabric (nsAllShow)

Total number of switches in the fabric (fabricShow)

Brocade configuration and connection (Continued)

Fabric OS Troubleshooting and Diagnostics Guide 49 53-1003141-01

Page 66

Brocade configuration form

50 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 67

Chapter

Firmware Download Errors

In this chapter

•

Blade troubleshooting tips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

•

Firmware download issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

•

Troubleshooting with the firmwareDownload command . . . . . . . . . . . . . . . 55

•

USB error handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

•

Considerations for downgrading firmware. . . . . . . . . . . . . . . . . . . . . . . . . . . 56

Blade troubleshooting tips

This chapter refers to the following specific types of blades inserted into the Brocade DCX, DCX-4S, and DCX 8510 family enterprise-class platforms:

•

FC blades or port blades contain only Fibre Channel ports: Brocade FC8-16/32/32E/48/48E/64 and FC16-32/48/64.

•

AP blades contain extra processors, some with specialized ports, and have the Fabric OS code downloaded to them automatically in the chassis: FX8-24, FS8-18, and FCOE10-24.

•

CP blades have a control processor (CP) used to control the entire switch; they can be inserted only into slots 6 and 7 on the Brocade DCX and DCX 8510-8, and slots 4 and 5 on the Brocade DCX-4S and DCX 8510-4.

•

CR8 and CR4S-8 core blades provide ICL functionality between two Brocade DCX Backbones. CR8 blades can be inserted only into slots 5 and 8 on the Brocade DCX. CR4S-8 blades can be inserted only into slots 3 and 6 on the Brocade DCX-4S.

•

CR16-8 and CR16-4 core blades provide ICL functionality between two Brocade DCX 8510 Backbones. CR16-8 blades can be inserted only into slots 5 and 8 on the Brocade DCX 8510-8. CR16-4 blades can be inserted only into slots 3 and 6 on the Brocade DCX 8510-4.

Typically, issues detected during firmware download to AP blades do not require recovery actions on your part.

If you experience frequent failovers between CPs that have different versions of firmware, then you may notice multiple blade firmware downloads and a longer startup time.

Symptom Relocation of internal image times out on CP8 control processor blade.

Probable cause and recommended action

This can be caused by issues in the co-CPU. If this happens, the firmware download process synchronizes the partitions in the main-CPU and co-CPU by starting a firmware commit operation. Wait at least 15 minutes for the commit operation to complete, issue the firmwareShow command to verify the partitions are synchronized, and reissue the firmwareDownload command. If the problem persists, you must contact the switch service provider.

Fabric OS Troubleshooting and Diagnostics Guide 51 53-1003141-01

Page 68

Firmware download issues

NOTE

CAUTION

This message only appears with Fabric OS v7.1.0 and earlier versions.

Symptom The blade is faulty (issue slotShow to confirm).

Probable cause and recommended action

If the port or application blade is faulty, enter the slotPowerOff and slotPowerOn commands for the port or application blade. If the port or application blade still appears to be faulty, remove it and re-insert it into the chassis.

Symptom The AP blade is stuck in the “LOADING” state (issue slotShow to confirm).

Probable cause and recommended action

If the blade remains in the loading state for a significant period of time, the firmware download times out. Remove the blade and re-insert it. When it boots up, autoleveling is triggered and the firmware download is attempted again.

Firmware download issues

After you start the firmware download process, do not enter any disruptive commands (such as reboot) that interrupt the process. The entire firmware download and commit process can take up to 30 minutes.

If there is a problem, wait for the time-out (30 minutes for network problems) before issuing the firmwareDownload command again. Disrupting the process can render the switch inoperable and require you to seek help from your switch service provider.

Do not disconnect the switch from power during the process because the switch could become inoperable when rebooted.

The following symptoms describe common firmware download issues and their recommended actions.

Symptom Firmware download times out.

Probable cause and recommended action

This can be caused by an excessively slow network. If it takes more than 30 minutes to download firmware on a switch, or on each CP in a director, the firmware download process times out. If a timeout occurs on a switch, the firmware download process synchronizes the two partitions on the switch by starting a firmware commit operation. If a timeout occurs in a director, the firmware download process synchronizes the firmware on the two partitions on the CP blades by starting a firmware commit operation on each CP.

52 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 69

Wait at least 15 minutes for the commit operation to complete, then use the firmwareShow

NOTE

command to verify the partitions are synchronized. In some older versions of firmware, the firmware commit operation may not be started automatically on the switch (or on the standby CP in the director). In this case, you can enter the firmwareCommit command manually on the switch (or on the standby CP in the director) to synchronize the partitions. After the firmware commit operation completes, re-issue the firmwareDownload command to upgrade the system.

Symptom Cannot upgrade directly to v7.2.0.

Probable cause and recommended action

If the switch is running Fabric OS v7.0.0 or earlier, you are not allowed to upgrade directly to v7.2.0 because of the “one-version” rule. Upgrade your switch to Fabric OS version v7.1.0 before upgrading to v7.2.0. The “one-version” rule also applies to downgrading.

Symptom Server is inaccessible or firmware path is invalid.

The entire firmware directory must be loaded onto the USB device under the firmware directory prior to using the USB device for firmware download.

Firmware download issues

Probable cause and recommended action

•

The FTP or SCP server’s host name is not known to the switch.

Verify with your network administrator that the switch has access to the FTP server.

Verify the path to the FTP or SCP server is accessible from the switch. For more information on checking your FTP or SCP server, refer to Chapter 4, “Configuration”.

•

The USB path is not correct.

If your platform supports a USB memory device, verify that it is connected and running. Verify that the path name is correct by using the usbStorage -l command.

Example of usbStorage -l command

switch:admin> usbstorage -l firmware\ 1585MB 2012 Oct 22 17:46 v7.2.0_main_bld14\ 1585MB 2013 Apr 22 14:16 config\ 0B 2012 Oct 22 17:46 support\ 0B 2012 Oct 22 17:46 firmwarekey\ 0B 2012 Oct 22 17:46 Available space on USB storage 16%

Example of error message

switch:admin> firmwaredownload Download from USB [No]: y Firmware filename: v7.2.0_main_bld24 Server IP: 127.1.1.7, Protocol IPv4 Checking system settings for firmwaredownload... Failed to access usb://127.1.1.7//usb/usbstorage/brocade/firmware/v7.2.0_main_bld24/release.pl ist Cannot access the firmware on USB device. Please check the firmware path.

Fabric OS Troubleshooting and Diagnostics Guide 53 53-1003141-01

Page 70

Firmware download issues

Symptom Cannot download the requested firmware.

Probable cause and recommended action

The firmware you are trying to download on the switch is incompatible. Check the firmware version against the switch type. If the firmware is incompatible, retrieve the correct firmware version and try again.

Example of error message

SW3900:admin> firmwaredownload Server Name or IP Address: 192.168.126.115 User Name: userFoo File Name: /users/home/userFoo/firmware/v7.1.0 Network Protocol(1-auto-select, 2-FTP, 3-SCP) [1]: 2 Password: Server IP: 192.168.168.115, Protocol IPv4 Checking system settings for firmwaredownload...

Cannot download the requested firmware because the firmware doesn't support this platform. Please enter another firmware path.

Symptom Cannot download on a switch with Interop mode turned on.

Probable cause and recommended action

On single CP system, the Interop fabric does not support Coordinated HotCode Load.

Perform a firmwareDownload -o command. The operand bypasses the checking of Coordinated HotCode Load (HCL). On single CP systems in interop fabrics, the HCL protocol is used to ensure data traffic is not disrupted during firmware upgrades. This option allows a firmware download to continue even if HCL is not supported in the fabric or the protocol fails. Using this option may cause traffic disruption for some switches in the fabric.

Symptom You receive a “firmwaredownload is already in progress” message.

Probable cause and recommended action

The firmware download process has already been started and it is in progress. Wait until it completes. You can use the firmwareDownloadStatus and firmwareShow commands to monitor its progress. If the problem persists, contact your switch support provider.

Example of a firmwaredownload already in progress

switch:admin> firmwaredownload

Server Name or IP Address: 192.168.168.115 User Name: userFoo File Name: /users/home/userFoo/firmware/v7.1.0 Network Protocol(1-auto-select, 2-FTP, 3-SCP) [1]: 2 Password: Server IP: 192.168.168.115, Protocol IPv4 Checking system settings for firmwaredownload...

Sanity check failed because firmwaredownload is already in progress.

54 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 71

Troubleshooting with the firmwareDownload command

ATTENTION

NOTE

Troubleshooting with the firmwareDownload command

A network diagnostic script and preinstallation check is a part of the firmware download procedure. The script and preinstallation check performs troubleshooting and automatically checks for any blocking conditions. If the firmware download fails, refer to the Fabric OS Message Reference for details about error messages. Also refer to, “Considerations for downgrading firmware” on page 56.

Do not run mixed firmware versions on CPs.

If a firmware download fails in a director, the firmwareDownload command synchronizes the firmware on the two partitions of each CP by starting a firmware commit operation. Wait at least 15 minutes for this commit operation to complete before attempting another firmware download.

If the firmware download fails in a director or enterprise-class platform, the CPs may end up with different versions of firmware and are unable to achieve HA synchronization between the two blades. In such cases, issue the firmwareDownload -s command on the standby CP; the single mode (-s) option allows you to upgrade the firmware on the standby CP to match the firmware version running on the active CP. This situation can also occur during a CP replacement in which the replacement CP has a lower firmware version. HA should be disabled and the firmware should be updated on the replacement CP before any other operations are considered.

Then reissue the firmwareDownload command to download the desired firmware version to both CPs. For example, if CP0 is running Fabric OS v7.1.0 on the primary and secondary partitions, and CP1 is running Fabric OS v7.2.0 on the primary and secondary partitions, then synchronize them by issuing the firmwareDownload command.

Some of the messages include error codes (as shown in the following example). These error codes are for internal use only and you can disregard them.

Port configuration with EX ports enabled along with trunking for port(s) 63, use the portCfgEXPort, portCfgVEXPort, and portCfgTrunkPort commands to remedy this. Verify blade is ENABLED. (error 3)

Gathering additional information

You should follow these best practices for firmware download before you start the procedure:

•

Keep all session logs.

•

Enter the supportSave or the supportShow command before and after entering the firmwareDownload command.

•

If a problem persists, package together all of the information (the Telnet session logs and serial console logs, output from the supportSave command) for your switch support provider. Make sure you identify what information was gathered before and after issuing the firmwareDownload command.

Fabric OS Troubleshooting and Diagnostics Guide 55 53-1003141-01

Page 72

USB error handling

NOTE

USB error handling

Tab le 11 outlines how the USB device handles errors under specific scenarios and details what

actions you should take after the error occurs.

TABLE 11

Scenario under which download fails Error handling Action

An access error occurs during firmware download because the removal of the USB device, or USB device hardware failure, and so on.

USB device is not enabled. Firmware download fails with an error

USB error handling

Firmware download times out and commit is started to repair the partitions of the CPUs that are affected.

message

Considerations for downgrading firmware

The pre-installation check of the firmwareDownload command detects all of the blocking conditions that can prevent a successful downgrade, and warns you about all these conditions. The error messages displayed by the firmwareDownload command states the blocking conditions and the corresponding commands to correct them. You must address all of these blocking conditions before proceeding. Refer to the Fabric OS Administrator’s Guide for more information regarding individual features and commands.

To avoid failure of a firmware downgrade, verify the firmware you are downgrading to supports all the blades in the chassis, and that the switch, blades, or chassis supports all the features you are currently using. If not, you must disable or remove those features that are not supported.

None.

Enable the USB device using the usbStorage -e command and retry firmware download.

Preinstallation messages

The system messages in this section are displayed if an exception is encountered during firmware download. The following examples show feature-related messages that you may see if you were upgrading from Fabric OS v7.1.0 to v7.2.0:

The system messages in this section are for illustration purposes only. They do not represent the entire range of possible error messages appropriate to a wide variety of installation scenarios.

Upgrade to 7.2 will set default QoS port configuration from AE to OFF.

Cannot upgrade directly to 7.2. Please upgrade to 7.0 first and then upgrade to 7.2.

Upgrade to 7.2 is not allowed because FC Fastwrite is not supported on this version. Please use "fastwritecfg" to deconfigure FC Fastwrite for all slots and try again.

Upgrade to 7.2 is not allowed due to the presence of ioddelay configuration. Please reset the feature with "ioddelayreset" before upgrading to v7.1.

56 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 73

Considerations for downgrading firmware

Upgrade to 7.2 is not allowed since base switch has R_RDY enabled ports. Please disable the R_RDY enabled ports in base switch using portcfgislmode command.

Firmware upgrade to Fabric OS 7.2.0 or higher is not allowed when there are more than 4 chassis connected through Inter-Chassis Links (ICLs) and the Enterprise ICL (EICL) license is not installed in the system. Note that even with an EICL license installed, only 9 chassis are allowed to connect through ICLs. You can either install an EICL license, or you must disable the additional ICL links before performing a firmware upgrade.

The following example shows hardware-related messages for a downgrade:

ecp:admin> firmwaredownload Type of Firmware (FOS, SAS, or any application) [FOS]: Server Name or IP Address: 10.1.2.3 User Name: userfoo File Name: /home/userfoo/v6.3.0 Network Protocol (1-auto-select, 2-FTP, 3-SCP) [1]: Password: <hidden>

Checking System Settings... Version compatibility check passed.

Checking version compatibility... Version compatibility check passed..

The following items must be addressed before downloading the specified firmware:

FC8-32E and FC8-48E are not supported by the targeted firmware. Please use slotshow to determine which of these are installed and remove them before continuing.

The following messages are displayed if an exception case is encountered during firmware downgrade. The following example shows feature-related messages that you may see if you were downgrading from Fabric OS v7.2.0 to v7.1.0:

Downgrade is not allowed as QoS mode is enabled (ON) on some of the ports ( other than E_Port(s) and F_Port with AoQ feature ) and there is no Adaptive Networking license installed on the switch. To proceed further with downgrade either install Adaptive Networking license, or set QoS mode to OFF/AE. Please use "portcfgqos --disable [slot/]port" to set QoS mode on these ports to OFF or use "portcfgqos --default [slot/]port" to set QoS mode on these ports to AE.

Downgrade is not allowed as some of the E_Port(s) are using QoS feature and there is no Adaptive Networking license installed on the switch.To proceed further with downgrade either install Adaptive Networking license, or set QoS mode on these ports to OFF. Please use "islshow" to view list of E_Port with QoS feature enabled and use "portcfgqos --disable [slot/]port" to set QoS mode on these ports to OFF.

Downgrade to selected version is not allowed. Please remove E-Port Credit configuration using "portcfgdefault" or "portcfgeportcredits --disable" CLI on these ports.

Downgrade is not allowed because Flow Vision is configured. Please delete all the flows using "flow" CLI.

Fabric OS Troubleshooting and Diagnostics Guide 57 53-1003141-01

Page 74

Considerations for downgrading firmware

Downgrade is not allowed because there are few ports other than the first octet of the blade/switch configured for 10G/N10 speeds.Please change the octet speed combo of those ports to 1 using "portcfgoctetspeedcombo" CLI and change their speeds using "portcfgspeed" CLI.

Downgrade is not allowed because Monitoring and Alerting Policy Suite (MAPS) is enabled. Please disable MAPS using "mapsconfig --disable" before proceeding.

Downgrade to pre-7.2 Fabric OS versions is not allowed. Please configure TACACS+ keys of length in the range (8-40) characters using aaaconfig CLI {Usage: aaaconfig --change <server> -conf tacacs+ [-s <secret>], before performing a firmware downgrade.

Downgrade is not allowed because ICL ports are configured as EX-ports.Please use "portcfgshow" to view the ICL EX ports and use "portcfgexport" to disable the EX feature on those ports before downgrading.

The following example shows feature-related messages that you may see if you were downgrading from Fabric OS v7.1.0 to v7.0.x:

Downgrade to selected version is not allowed because few ports are configured with Longdistance -buffers option. Please remove the configuration using "portcfglongdistance L0" CLI or change the configuration with -distance option.

Downgrade is not allowed because Location ID is configured. Please use "configure" command to clear Location ID.

Downgrade is not allowed because switch is in AG mode and D-Ports are configured. Please use "switchshow" to view the D-port list and use "portcfgdport --disable <port_no>" to disable it before downgrading.

Downgrade is not allowed because R-RDY flow control ports are configured as D-Ports. Please use "portdporttest --show all" to view the port list and "portcfgdport --disable" to disable before downgrading.

Downgrade is not allowed because D-Port is configured with DWDM mode. Please use "portcfgshow" to view the port list and "portcfgdport --enable" to reset DWDM mode before downgrading.

Downgrade is not allowed because ICL ports are configured as D-Ports. Please use "switchshow" to view the D-port list and use "portcfgdport --disable <port_no>" to disable it before downgrading.

Downgrade is not allowed because one or more ports have credit recovery enabled. Please use "portcfgcreditrecovery --disable" command to disabled credit recovery.

Downgrade is not allowed because one or more ports have FEC enabled.Please use "portcfgfec --disable" command to disabled FEC.

58 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 75

Considerations for downgrading firmware

ATTENTION

Blade types

Where blades are incompatible with a firmware download, they must be removed or powered off before a firmware download begins, as noted in the following message.

Message The FC10-6 (type 39) blade is not supported by the target firmware. Please use slotshow to find out

which slot it is in and remove it first.

Probable cause and recommended action

The firmware download operation was attempting to upgrade a system to Fabric OS v7.2.0 with one or more of the Brocade FC10-6 blades (blade ID 39) in the system. The Brocade FC10-6 blades are not supported on firmware v7.2.0, so the firmware download operation failed.

Use the slotShow command to display which slots the Brocade FC10-6 blades occupy. Physically remove the blades from the chassis, or use the micro-switch to turn the blade off. Retry the firmware download operation.

Firmware versions

The system messages in this section refer to differences between the current firmware and the firmware you are applying to the switch.

Brocade does not support upgrades from more than one previous release. For example, upgrading from Fabric OS v7.1.0 to v7.2.0 is supported, but upgrading from Fabric OS v7.0.0 or a previous release directly to v7.2.0 is not. In other words, upgrading a switch from Fabric OS v7.0.0 to v7.2.0 is a two-step process: first upgrade to v7.1.0, and then upgrade to v7.2.0. If you are running a pre-Fabric OS v6.2.0 version, first you must upgrade to v6.2.0, then to v6.3.0, then to v6.4.0, then to v7.0.0, then to v7.1.0, and finally to v7.2.0.

Message Cannot upgrade directly to v7.2.0. Upgrade your switch to v7.1.0 first before upgrading to the

requested version.

Probable cause and recommended action

If the switch is running v7.0.0 or earlier, you are not allowed to upgrade directly to v7.2.0 because of the “two-version” rule.

Upgrade your switch to Fabric OS version v7.1.0 before upgrading to v7.2.0

Message Non-disruptive firmwaredownload is not supported when downgrading to 7.0. Please use

firmwaredownload -s to download the 7.0 firmware.

Probable cause and recommended action

If the switch is running v7.2.0, you are not allowed to downgrade directly to v7.0.x without causing disruption to your fabric.

Downgrade using the firmwareDownload -s command.

Fabric OS Troubleshooting and Diagnostics Guide 59 53-1003141-01

Page 76

Considerations for downgrading firmware

Platform

The following system message pertains to switch features or fabric-wide settings that must be removed or disabled before downgrading the firmware.

Message Downgrade is not allowed because VF is enabled. Please run "lscfg - -config" and "lscfg - -delete"

commands to remove the non-default LS first, then run "fosconfig - -disable vf" to disable VF before proceeding.

Probable cause and recommended action

You cannot downgrade because Virtual Fabrics is enabled. Delete the logical switches, delete the base switch, and disable Virtual Fabrics prior to downgrading the firmware.

Routing

The following system message refers to any route settings that must be changed prior to downgrading the switch’s firmware.

Message Please use "configure" to clear the location ID.

Probable cause and recommended action

If the switch is running v7.1.0 or later, and Location-Oriented Link Aggregation is configured, you cannot downgrade the switch to v7.0.0 or earlier.

Use the configure command to clear the location ID.

60 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 77

Chapter

Security

In this chapter

Passwords

The following section describes various ways to recover forgotten passwords.

Symptom User forgot password.

Probable cause and recommended action

•

Passwords . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

•

Device authentication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

•

Protocol and certificate management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

•

SNMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

•

FIPS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

If you know the root password, you can use this procedure to recover the password for the default accounts of user, admin, and factory. If you do not know the root password, you must contact your service support provider to recover admin passwords.

Recovering passwords

1. Open a CLI connection (serial or Telnet) to the switch.

2. Log in as root.

3. Enter the command for the type of password that was lost:

passwd user passwd admin passwd factory

4. Enter the requested information at the prompts.

Symptom Unable to log in as root password.

Probable cause and recommended action

To recover your root password, contact your switch service provider.

Symptom Unable to log into the boot PROM.

Probable cause and recommended action

To recover a lost boot PROM password, contact your switch service provider. You must have previously set a recovery string to recover the boot PROM password.

This does not work on lost or forgotten passwords in the account database.

Fabric OS Troubleshooting and Diagnostics Guide 61 53-1003141-01

Page 78

Device authentication

Password recovery options

Tab le 12 describes the options available when one or more types of passwords are lost.

TABLE 12

Topic Solution

If all the passwords are forgotten, what is the password recovery mechanism? Are these procedures non-disruptive recovery procedures?

If a user has only the root password, what is the password recovery mechanism?

How to recover boot PROM password? Contact your switch service provider and provide the

How to recover a user, admin, or factory password? Refer to “Passwords” on page 61 for more information on

Password recovery options

Symptom User is unable to modify switch settings.

Probable cause and recommended action

The most common error when managing user accounts is not setting up the default Admin Domain and access control list or Role-Based Access Control (RBAC).

Errors such as a user not being able to run a command or modify switch settings are usually related to what role the user has been assigned.

Contact your switch service provider. A non-disruptive procedure is available.

Use the passwd command to set other passwords. Use the passwdDefault command to set all passwords to

default.

recovery string.

recovering these passwords.

Device authentication

Symptom Switch is unable to authenticate device.

Probable cause and recommended action

When the device authentication policy is set to ON, the switch expects a FLOGI with the FC-SP bit set. If this bit is not set, the switch rejects the FLOGI with reason LS_LOGICAL_ERROR (0x03), in the switch log with the explanation of “Authentication Required”(0x48), and disables the port. Set the device authentication policy mode on the switch to ON.

Symptom Switch is unable to form an F_Port.

Probable cause and recommended action

Regardless of the device authentication policy mode on the switch, the F_Port is disabled if the DH-CHAP protocol fails to authenticate. If the HBA sets the FC-SP bit during FLOGI and the switch sends a FLOGI accept with FC-SP bit set, then the switch expects the HBA to start the AUTH_NEGOTIATE. From this point on until the AUTH_NEGOTIATE is completed, all ELS and CT frames, except the AUTH_NEGOTIATE ELS frame, are blocked by the switch. During this time, the Fibre Channel driver rejects all other ELS frames. The F_Port does not form until the AUTH_NEGOTIATE is completed. It is the HBA's responsibility to send an Authentication Negotiation ELS frame after receiving the FLOGI accept frame with the FC-SP bit set.

62 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 79

Protocol and certificate management

This section provides information and procedures for troubleshooting standard Fabric OS security features such as protocol and certificate management.

Symptom Troubleshooting certificates.

Probable cause and recommended action

If you receive messages in the browser or in a pop-up window when logging in to the target switch using HTTPS, refer to Tab le 13 for recommended actions you can take to correct the problem.

TABLE 13

Message Action

The page cannot be displayed The SSL certificate is not installed correctly or HTTPS is not

The security certificate was issued by a company you have not chosen to trust.

The security certificate has expired or is not yet valid

The name on the security certificate is invalid or does not match the name of the site file

This page contains both secure and nonsecure items. Do you want to display the nonsecure items?

SSL messages and actions

Protocol and certificate management

enabled correctly. Make sure that the certificate has not expired, that HTTPS is enabled, and that certificate file names are configured correctly.

The certificate is not installed in the browser. Install it as described in the Fabric OS Administrator’s Guide.

Either the certificate file is corrupted or it needs to be updated. Click View Certificate to verify the certificate content. If it is corrupted or out of date, obtain and install a new certificate.

The certificate is not installed correctly in the Java Plug-in. Install it as described in the Fabric OS Administrator’s Guide.

Click No in this pop-up window. The session opens with a closed lock icon on the lower-right corner of the browser, indicating an encrypted connection.

Gathering additional information

For security-related issues, use the following guidelines to gather additional data for your switch support provider.

•

Use the supportSave -n command.

•

If not sure about the problem area, use the supportSave -n to collect data from all switches in the fabric.

•

If you think it may be related to E_Port authentication, use the supportSave -n command to collect data from both switches of the affected E_Port.

•

If you think this is a policy-related issue, FCS switch, or other security server-related issue, then use supportSave -n to collect data from the Primary FCS switch and all affected switches.

•

If login-related, then also include the following information:

Does login problem appear on a Serial, CP IP, or Switch IP address connection?

Is it CP0 or CP1?

Is the CP in active or standby?

Is it the first time login after firmwareDownload and reboot?

Fabric OS Troubleshooting and Diagnostics Guide 63 53-1003141-01

Page 80

SNMP

Symptom SNMP management station server is unable to receive traps from fabric.

This section describes symptoms with associated causes and recommended actions for SNMP-related issues.

Probable cause and recommended action

There are several causes related to this generic issue. You must verify the following:

•

There are no port filters in the firewalls between the fabric and the SNMP management station.

•

If your SNMP management station is a dual-homed server, check that the routing tables are set up correctly for your network.

If you continue to have problems, collect the data in the next section and contact your switch support provider.

Gathering additional information

In addition to supportSave, gather the MIB browser snapshot with the problem (like Adventnet screen snapshot) for an MIB variable.

FIPS

This section describes symptoms with associated causes and recommended actions for problems related to FIPS.

Symptom When FIPS is turned on, the switch constantly reboots.

Probable cause and recommended action

When FIPS is turned on, the switch runs conditional tests each time it is rebooted. These tests run random number generators and are executed to verify the randomness of the random number generator. The conditional tests are executed each time prior to using the random number provided by the random number generator.

The results of all self-tests, for both power-up and conditional, are recorded in the system log or are output to the local console. This includes logging both passing and failing results. If the tests fail on your switch, it constantly reboots. Because boot PROM access is disabled, you are not able to exit out of the reboot. You must send the switch back to your switch service provider for repair.

64 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 81

Chapter

Virtual Fabrics

In this chapter

•

General Virtual Fabrics troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

•

Fabric identification issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

•

Logical Fabric issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

•

Base switch issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

•

Logical switch issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

•

Switch configuration blade compatibility. . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

•

Gathering additional information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

General Virtual Fabrics troubleshooting

All of the following constraints apply when the Virtual Fabrics feature is enabled:

•

The base fabric works only in Brocade native mode, not in an interoperable mode.

•

The base switch does not have any devices. The base fabric can have devices in remote Layer 2 switches; traffic between those devices is supported.

•

A non-base switch in a Virtual Fabrics-capable chassis must not be part of a fabric that serves as a base fabric for some other logical fabric traffic. Although software does not detect or prevent users from deploying such a configuration, such a configuration is not supported.

•

ICL ports can only be in the base or default switch. If the XISL is turned off, you can connect ICLs to other logical switches.

•

A default switch can be configured as a base switch in the fixed-port switches, but not in a Brocade Backbone. Fabric IDs of default switches cannot be manually changed.

•

The default switch is able to participate in a logical fabric using extended ISLs (XISLs). In the

Brocade Backbones, the default switch does not participate in a logical fabric and is a purely Layer 2 logical switch.

•

EX_Ports and VEX_Ports are supported in the base switch. EX_Ports cannot be part of any other switch other than the base switch.

•

EX_Ports and VEX_Ports cannot connect to a fabric that has a logical switch with the Allow XISL use mode on. The port is disabled with the reason “Conflict: XISL capability domain”.

•

External device sharing is supported through EX_Ports. Internal device sharing (sharing a device in a logical fabric with other fabrics, without having an EX_Port) is not supported.

•

A logical fabric cannot have EX_Ports using XISLs and cannot serve as a backbone to any EX_Port traffic. Similarly, the default switch cannot be part of a fabric that serves as a backbone to any EX_Port traffic.

Fabric OS Troubleshooting and Diagnostics Guide 65 53-1003141-01

Page 82

Fabric identification issues

NOTE

•

Admin Domains are mutually exclusive with Virtual Fabrics. When Virtual Fabrics is enabled, all access control is based on the Virtual Fabrics context.

•

Traffic Isolation zones with no failover option are not supported in logical fabrics. TI zones defined in the base fabric for logical fabric traffic must allow failover.

The configure command has a “Disable FID check” fabric parameter option, which can be used to disable the FID check for FICON logical switches.

Fabric identification issues

Symptom E_Ports directly connecting two logical switches do not form, or are disabled.

Probable cause and recommended action

The FIDs on each of the logical switches must be the same.

Use the lsCfg

change FID -newfid FID command to change the FID.

Symptom Invalid FID.

Probable cause and recommended action

FIDs for switches may be from 1 through 128 as long as they are not already in use, except for EX_Ports, which are only assigned FIDs from 1 through 127.

Use the lsCfg

Symptom The FID is currently in use.

Probable cause and recommended action

You may not create two (2) or more logical switches with the same FID.

Use the lsCfg

Logical Fabric issues

Symptom Logical port <port_number> disabled.

Probable cause and recommended action

show command to view the current FIDs on the chassis and then the lsCfg

show command to verify whether the FID is in use. If it is, use another FID.

show and fcrFabricShow commands to view FIDs in use.

This message indicates an LISL was disabled because of some protocol conflict or security or policy violation. This can result in possible traffic issues. You should resolve the cause of the conflict and re-enable the LISL by using the lfCfg

66 Fabric OS Troubleshooting and Diagnostics Guide

lislenable command.

53-1003141-01

Page 83

Symptom The switch with domain <domain> with firmware version <fw version> has joined the FID <fid>

fabric and may not be compatible with XISL use.

Probable cause and recommended action

This message indicates the specified switch in the logical fabric that is using XISLs is running an incompatible firmware version and must be upgraded to Fabric OS v6.2.0 or later.

Base switch issues

All logical switches in a fabric should have the same base switch attribute. If a base switch is connected to a non-base switch, then you must take the appropriate action to resolve the conflict.

Symptom EX_Port is disabled with reason “Export in non base switch”.

Probable cause and recommended action

An EX_Port must be in the base switch.

Base switch issues

Use the lsCfg FID -slot [slot | slot_range] -port [ port | port_range] [-force] command and move the port to the base switch. If the port is not intended to be used as an EX_Port, use the portCfgDefault command to reset the port to its default configuration.

Symptom An EX_ or VEX_Port is disabled with reason “Conflict: XISL capability domain”.

Probable cause and recommended action

Use the configure command to set the value on the Allow XISL use to OFF on all logical switches of the connecting edge fabric.

Symptom E_Ports connecting two logical switches are disabled.

Probable cause and recommended action

If a base switch is directly connected to a non-base switch, all E_Ports to that logical switch are disabled.

Symptom Fabric ID and base switch are conflicted.

Probable cause and recommended action

If there is a Fabric ID conflict and a base switch conflict that exists between two switches, the Fabric ID conflict is detected first.

Use the lsCfg

create FID -b base command to create a base switch. Then use the lsCfg --config

change FID -newfid FID command to change the FID.

Symptom A base switch already exists on this system.

Probable cause and recommended action

Only one base switch is allowed on a platform. Use the lsCfg

create FID -b base command to remove the current base switch, and then create a new one.

Fabric OS Troubleshooting and Diagnostics Guide 67 53-1003141-01

delete FID command, then the lsCfg

Page 84

Logical switch issues

CAUTION

Logical switch issues

When a logical switch is created, all configuration for the logical switch is set to factory defaults. When a logical switch is deleted, all configuration for the logical switch is deleted permanently and is not recoverable.

Symptom The indicated slot is empty.

Probable cause and recommended action

You used the lsCfg command and an empty slot was specified.

Re-issue the command with the appropriate slot number.

Symptom Validation of switch configuration changes is not supported on this platform.

Probable cause and recommended action

This platform is unknown to the logical switch subsystem.

Symptom Given slot number is not valid on this platform.

Probable cause and recommended action

You are specifying a slot number that is not valid on the platform, for example, slot 0 on a Brocade DCX or slot 12 on a Brocade DCX-4S.

Symptom Slot must be enabled to configure ports.

Probable cause and recommended action

You may only attempt to configure ports on enabled blades (blades may be faulted).

Symptom Unable to determine slot type.

Probable cause and recommended action

The slot type is not known to the logical switch. Verify the slot and try again.

Symptom There are no ports on this slot.

Probable cause and recommended action

There are no configurable ports on the slot indicated by the lsCfg command. Verify the ports and try again.

Symptom Unable to remove ports from their current switch.

Probable cause and recommended action

When moving ports to a switch, you must first remove them from the switch in which they reside. This error message is displayed if this step fails.

68 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 85

Switch configuration blade compatibility

Symptom A non-GE blade is within the slot range.

Probable cause and recommended action

You are attempting to configure a GE port on a slot that does not contain GE ports.

Symptom A port or ports is already in the current switch.

Probable cause and recommended action

You may not move a port to the same switch.

Symptom The maximum number of switches for this platform has been reached.

Probable cause and recommended action

Each platform that supports Virtual Fabrics has a maximum number of logical switches that may be supported. The platform has reached this limit.

Symptom Unable to create the switch.

Probable cause and recommended action

There was an error while creating the switch.

Symptom A port or ports cannot be moved to the requested switch because it would exceed the 256 area

limit for this switch.

Probable cause and recommended action

The area limit would be exceeded if the lsCfg command were allowed.

Symptom A port or ports cannot be moved to the requested switch because it may only exist in a base or

default switch.

Probable cause and recommended action

You are attempting to move ports on a core blade into a non-default or non-base switch.

Switch configuration blade compatibility

Symptom A slot in the chassis displays a FAULTY(91) in the output of the slotShow command.

Probable cause and recommended action

When an enterprise-class platform is coming up or when a blade is inserted, the switch configuration is checked based on the blade type. If the configuration does not match with the blade type, the blade is faulted. This is displayed as FAULTY(91) in the output of the slotShow command.

Fabric OS Troubleshooting and Diagnostics Guide 69 53-1003141-01

Page 86

Gathering additional information

Use the lsCfg –restore_slot_to_default command to correct the problem. Once the configuration discrepancy has been fixed, you may use slotPowerOff followed by slotPowerOn to recover.

>lscfg –restore_slot_to_default 1 >slotpoweroff 1 >slotpoweron 1 Slot Blade Type ID Model Name Status

------------------------------------------------- 1 SW BLADE 77 FC8-64 ENABLED 2 SW BLADE 97 FC16-32 ENABLED 3 SW BLADE 96 FC16-48 ENABLED 4 SW BLADE 96 FC16-48 ENABLED 5 CORE BLADE 98 CR16-8 ENABLED 6 CP BLADE 50 CP8 ENABLED 7 CP BLADE 50 CP8 ENABLED 8 CORE BLADE 98 CR16-8 ENABLED 9 SW BLADE 55 FC8-32 ENABLED 10 SW BLADE 96 FC16-48 ENABLED 11 AP BLADE 75 FX8-24 ENABLED 12 SW BLADE 97 FC16-32 ENABLED

Gathering additional information

For Virtual Fabrics-related issues, use the following guidelines to gather additional data for your switch support provider:

•

Use the supportSave command.

•

If not sure about the problem area, use the supportSave command on all chassis and logical switches in the fabric.

•

If you think it may be related to E_Port authentication, then use the supportSave -n command on both switches or logical switches of the affected E_Port.

70 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 87

Chapter

ISL Trunking

In this chapter

Link issues

This section describes trunking link issues that can come up and recommended actions to take to correct the problems.

Symptom A link that is part of an ISL trunk failed.

Probable cause and recommended action

Use the trunkDebug port1 port2 command to troubleshoot the problem, as shown in the following procedure.

•

Link issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

•

Buffer credit issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

1. Connect to the switch and log in as admin.

2. Enter the trunkDebug port1 port2 command:

In the trunkDebug command, the following operands are used::

port1 Specify the area number or index of port 1. Use the switchShow command to

view the area or index numbers for a port. This operand is required.

port2 Specify the area number or index of port 2. Use the switchShow command to

view the area or index numbers for a port. This operand is required.

Example of an unformed E_Port

This example shows that port 3 is not configured as an E_Port:

ecp:admin> trunkdebug 126, 127 port 126 is not E/EX port port 127 is not E/EX port

Example of a formed E_Port

ecp:admin> trunkdebug 100, 101 port 100 and 101 connect to the switch 10:00:00:05:1e:34:02:45

The trunkDebug command displays the possible reason that two ports cannot be trunked. Possible reasons are:

•

The switch does not support trunking.

•

A trunking license is required.

Fabric OS Troubleshooting and Diagnostics Guide 71 53-1003141-01

Page 88

Buffer credit issues

•

Trunking is not supported in switch interoperability mode.

•

Port trunking is disabled.

•

The port is not an E_Port.

•

The port is not 2 Gbps, 4 Gbps, or 8 Gbps.

•

The port connects to a switch other than the one you want.

To correct this issue, connect additional ISLs to the switch with which you want to communicate.

•

The ports are not the same speed or they are not set to an invalid speed.

Manually set port speeds to a speed supported on both sides of the trunk.

•

The ports are not set to the same long distance mode.

Set the long distance mode to the same setting on all ports on both sides of the trunk.

•

Local or remote ports are not in the same port group.

Move all ISLs to the same port group. The port groups begin at port 0 and are in groups of 4 or 8, depending on the switch model. Until this is done, the ISLs do not trunk.

•

The difference in the cable length among trunked links is greater than the allowed difference.

Buffer credit issues

This section describes a trunk going online and offline or hosts not being able to talk to a storage device.

Symptom Trunk goes offline and online (bounces).

Probable cause and recommended action

A port disabled at one end because of buffer underallocation causes all the disabled ports at the other end to become enabled. Some of these enabled ports become disabled because of a lack of buffers, which in turn triggers ports to be enabled once again at the other end.

While the system is stabilizing the buffer allocation, it warns that ports are disabled because of a lack of buffers, but it does not send a message to the console when buffers are enabled. The system requires a few passes to stabilize the buffer allocation. Ultimately, the number of ports for which buffers are available come up and stabilize. You should wait for stabilization, and then proceed with correcting the buffer allocation situation.

Getting out of buffer-limited mode

Occurs on LD_Ports.

1. Change the LD port speed to a lower speed (of non-buffer limited ports).

2. Change the LD port’s estimated distance to a shorter distance (of non-buffer limited ports).

3. Change LD back to L0 (of non-buffer limited ports).

4. If you are in buffer-limited mode on the LD port, then increase the estimated distance.

5. Enable any of these changes on the buffer-limited port or switch by issuing the commands portDisable and portEnable.

72 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 89

Chapter

Zoning

In this chapter

•

Overview of corrective action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

•

Segmented fabrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

•

Zone conflicts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

•

Gathering additional information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

Overview of corrective action

The following overview provides a basic starting point for you to troubleshoot your zoning problem.

1. Verify that you have a zone problem.

2. Determine the nature of the zone conflict.

3. Take the appropriate steps to correct zone conflict.

To correct a merge conflict without disrupting the fabric, first verify that it was a fabric merge problem, then edit zone configuration members, and then reorder the zone member list if necessary.

The newly changed zone configuration are not effective until you issue the cfgEnable command. This should be done during a maintenance window because this may cause disruption in large fabrics.

Verifying a fabric merge problem

1. Enter the switchShow command to validate that the segmentation is because of a zone issue.

2. Review “Segmented fabrics,” to view the different types of zone discrepancies and determine what might be causing the conflict.

Verifying a TI zone problem

Use the zone --show command to display information about TI zones. This command displays the following information for each zone:

•

Zone name

•

E_Port members

•

N_Port members

•

Configured status (the latest status, which may or may not have been activated by cfgEnable)

•

Enabled status (the status that has been activated by cfgEnable)

Fabric OS Troubleshooting and Diagnostics Guide 73 53-1003141-01

Page 90

Segmented fabrics

If you enter the cfgShow command to display information about all zones, the TI zones appear in the defined zone configuration only and do not appear in the effective zone configuration.

1. Connect to the switch and log in as admin.

2. Enter the zone

show command where .name is the name of the zone to be displayed. If the name is omitted, the command displays information about all TI zones in the defined configuration.

zone --show [name]

To display information about the TI zone purplezone, use the following command:

switch:admin> zone --show purplezone Defined TI zone configuration:

TI Zone Name: redzone: Port List: 1,2; 1,3; 3,3; 4,5

Configured Status: Activated / Failover-Enabled Enabled Status: Activated / Failover-Enabled

To display information about all TI zones in the defined configuration, use the following command:

switch:admin> zone --show Defined TI zone configuration:

TI Zone Name: greenzone: Port List: 2,2; 3,3; 5,3; 4,11;

Configured Status: Activated / Failover-Enabled Enabled Status: Activated / Failover-Enabled

TI Zone Name: purplezone: Port List: 1,2; 1,3; 3,3; 4,5;

Configured Status: Activated / Failover-Enabled Enabled Status: Deactivated / Failover-Enabled

TI Zone Name: bluezone: Port List: 9,2; 9,3; 8,3; 8,5;

Configured Status: Deactivated / Failover-Disabled Enabled Status: Activated / Failover-Enabled

Segmented fabrics

This section discusses fabric segmentation. Fabric segmentation occurs when two or more switches are joined together by ISLs and do not communicate to each other. Each switch appears as a separate fabric when you use the fabricShow command.

74 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 91

Symptom Zone conflict appears in logs and fabric is segmented.

ATTENTION

Probable cause and recommended action

This issue is usually caused by incompatible zoning configurations. Verify the following is true:

•

The effective cfg (zone set) on each end of the segmented ISL is identical.

•

Any zone object with the same name has the same entries in the same sequence.

Symptom Fabric segmentation is caused by a “configuration mismatch”.

Probable cause and recommended action

Occurs when zoning is enabled in both fabrics and the zone configurations are different in each fabric.

Symptom Fabric segmentation is caused by a “type mismatch”.

Probable cause and recommended action

Occurs when the name of a zone object in one fabric is also used for a different type of zone object in the other fabric. A zone object is any device in a zone

Symptom Fabric segmentation is caused by a “content mismatch”.

Zone conflicts

Probable cause and recommended action

Occurs when the definition in one fabric is different from the definition of a zone object with the same name in the other fabric.

Zone conflicts

Zone conflicts can be resolved by saving a configuration file with the configUpload command, examining the zoning information in the file, and performing a cut and paste operation so that the configuration information matches in the fabrics being merged.

After examining the configuration file, you can choose to resolve zone conflicts by using the cfgDisable command followed by the cfgClear command on the incorrectly configured segmented fabric, then enter the cfgSave command followed by the portDisable and portEnable commands on one of the ISL ports that connects the fabrics. This causes a merge, making the fabric consistent with the correct configuration.

Be careful using the cfgClear command because it deletes the defined configuration.

Tab le 14 summarizes commands that are useful for debugging zoning issues.

TABLE 14

Commands for debugging zoning

Command Function

aliCreate Use to create a zone alias.

aliDelete Use to delete a zone alias.

cfgCreate Use to create a zone configuration.

cfgShow Displays zoning configuration.

Fabric OS Troubleshooting and Diagnostics Guide 75 53-1003141-01

Page 92

ATTENTION

Zone conflicts

TABLE 14

Command Function

cfgDisable Disables the active (effective) configuration

cfgEnable Use to enable and activate (make effective) the specified configuration.

cfgSave Use to save changes to the zone configuration database.

cfgTransAbort Use to abort the current zoning transaction without committing it.

cfgTransShow Use to display the ID of the current zoning transaction

defZone Sets the default zone access mode to No Access, initializes a zoning transaction (if one is not

licenseShow Displays current license keys and associated (licensed) products.

switchShow Displays currently enabled configuration and any E_Port segmentations resulting from zone

zoneAdd Use to add a member to an existing zone.

zoneCreate Use to create a zone. Before a zone becomes active, the cfgSave and cfgEnable commands

zoneHelp Displays help information for zone commands.

zoneShow Displays zone information.

Commands for debugging zoning (Continued)

already in progress), and creates the reserved zoning objects.

conflicts.

must be used.

For more information about setting up zoning on your switch, refer to the Fabric OS Administrator’s Guide.

You can correct zone conflicts by using the cfgClear command to clear the zoning database.

The cfgClear command is a disruptive procedure.

Resolving zoning conflicts

When merging two fabrics, mulitple zoning CLI sessions can be launched on the same switch, or on different switches. This section describes these situations and how they are automatically resolved.

•

Dual-CLI sessions from the same switch: If you start a zone transaction from CLI-Session1 and then try to perform a zone modification from CLI-Session2, the CLI-Session2 zone transaction is not allowed due to CLI-Session2 not being the owner of the open transaction. If CLI-Session1 logs out, this ends the open transaction and aborts any current zone modifications. CLI-Session2 is then able to perform zone modifications. Therefore, the zone transaction locking mechanism works on a single switch from the CLI perspective and there is no dangling transaction.

•

Dual-CLI sessions from different switches: If you start a CLI zone transaction on Switch1 and start another CLI zone transaction on Switch2, when committing the zone transaction from Switch1, the open zone transaction from Switch2 is aborted by Switch1. The following message is posted on Switch2 at the time of zone commit from Switch1:

2012/03/09-21:45:26, [ZONE-1027], 3285, FID 128, INFO, sw0, Zoning transaction aborted Zone Config update Received

76 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 93

Zone conflicts

ATTENTION

Correcting a fabric merge problem quickly

1. Determine which switches have the incorrect zoning configuration; then, log in to the switches as admin.

2. Enter the switchDisable command on all problem switches.

3. Enter the cfgDisable command on each switch.

4. Enter the cfgClear command on each switch.

5. Enter the cfgSave command on each switch to commit the change.

The cfgClear command clears the zoning database on the switch where the command is run.

6. Enter the switchEnable command on each switch once the zoning configuration has been cleared.

This forces the zones to merge and populates the switches with the correct zoning database. The fabrics then merge.

Changing the default zone access

A switch is not allowed to merge with another switch that has an active effective configuration if the default zone is set to “no access”. Before the switch can join, the default zone setting must be set to “all access”. When the default zone no access option is enabled and the active configuration is disabled by using the cfgDisable command, a special hidden configuration with no members is activated. This configuration does not allow the switch to merge with switches that have an active effective configuration.

1. Connect to the switch and log in using an account with admin permissions.

2. Display the current setting with the defZone –show command.

3. If your default zone is set to “no access”, use the defZone –allaccess command to change the default zone.

4. Enter the cfgSave command to save the current configuration.

Editing zone configuration members

1. Log in to one of the switches in a segmented fabric as admin.

2. Enter the cfgShow command and print the output.

3. Start another Telnet session and connect to the next fabric as an admin.

4. Enter the cfgShow command and print the output.

5. Compare the two fabric zone configurations line by line and look for an incompatible configuration.

6. Connect to one of the fabrics.

7. Run zone configure edit commands to edit the fabric zone configuration for the segmented switch (refer to Ta ble 14 on page 75 for specific commands).

Fabric OS Troubleshooting and Diagnostics Guide 77 53-1003141-01

Page 94

Zone conflicts

If the zoneset members between two switches are not listed in the same order in both configurations, the configurations are considered a mismatch; this results in the switches being segmented in the fabric.

For example:

[cfg1 = z1; z2]

configuration are the same.

One simple approach to making sure that the zoneset members are in the same order is to keep the members in alphabetical order.

is different from

[cfg1 = z2; z1],

even though the members of the

Reordering the zone member list

1. Obtain the output from the cfgShow command for both switches.

2. Compare the order in which the zone members are listed. Members must be listed in the same order.

3. Rearrange zone members so the configuration for both switches is the same. Arrange zone members in alphabetical order, if possible.

Checking for Fibre Channel connectivity problems

Enter the fcPing command to:

•

Generate an Extended Link Service (ELS) frame ECHO request to the source port specified and validate the response.

•

Generate an ELS ECHO request to the destination port specified and validate the response.

Regardless of the device’s zoning, the fcPing command sends the ELS frame to the destination port. A device can take any of the following actions:

•

Send an ELS Accept to the ELS request.

•

Send an ELS Reject to the ELS request.

•

Ignore the ELS request.

The following example is sample output from the fcPing command in which one device accepts the request and another device rejects the request:

switch:admin> fcping 10:00:00:00:c9:29:0e:c4 21:00:00:20:37:25:ad:05 Source: 10:00:00:00:c9:29:0e:c4 Destination: 21:00:00:20:37:25:ad:05 Zone Check: Not Zoned

Pinging 10:00:00:00:c9:29:0e:c4 [0x20800] with 12 bytes of date: received reply from 10:00:00:00:c9:29:0e:c4: 12 bytes time:1162 usec received reply from 10:00:00:00:c9:29:0e:c4: 12 bytes time:1013 usec received reply from 10:00:00:00:c9:29:0e:c4: 12 bytes time:1442 usec received reply from 10:00:00:00:c9:29:0e:c4: 12 bytes time:1052 usec received reply from 10:00:00:00:c9:29:0e:c4: 12 bytes time:1012 usec 5 frames sent, 5 frames received, 0 frames rejected, 0 frames timeout

78 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 95

Zone conflicts

Round-trip min/avg/max = 1012/1136/1442 usec

Pinging 21:00:00:20:37:25:ad:05 [0x211e8] with 12 bytes of data: Request rejected Request rejected Request rejected Request rejected Request rejected 5 frames sent, 0 frames received, 5 frames rejected, 0 frames timeout Round-trip min/avg/max = 0/0/0 usec switch:admin>

The following example is sample output from the fcPing command in which one device accepts the request and another device does not respond to the request:

switch:admin> fcping 0x020800 22:00:00:04:cf:75:63:85 Source: 0x20800 Destination: 22:00:00:04:cf:75:63:85 Zone Check: Zoned

Pinging 0x020800 with 12 bytes of data: received reply from 0x020800: 12 bytes time:1159 usec received reply from 0x020800: 12 bytes time:1006 usec received reply from 0x020800: 12 bytes time:1008 usec received reply from 0x020800: 12 bytes time:1038 usec received reply from 0x020800: 12 bytes time:1010 usec 5 frames sent, 5 frames received, 0 frames rejected, 0 frames timeout Round-trip min/avg/max = 1006/1044/1159 usec

Pinging 22:00:00:04:cf:75:63:85 [0x217d9] with 12 bytes of data: Request timed out Request timed out Request timed out Request timed out Request timed out 5 frames sent, 0 frames received, 0 frames rejected, 5 frames timeout Round-trip min/avg/max = 0/0/0 usec switch:admin>

Checking for zoning problems

1. Enter the cfgActvShow command to determine if zoning is enabled.

•

If zoning is enabled, it is possible that the problem is being caused by zoning enforcement (for example, two devices in different zones cannot detect each other).

•

If zoning is disabled, check the default zone mode by entering the defZone --show command. If it is no access, change it to all access. To modify default zone mode from no access to all access, enter the defZone

2. Confirm that the specific edge devices that must communicate with each other are in the same zone.

•

If they are not in the same zone and zoning is enabled, proceed to step 3.

all command, and then the cfgSave command.

Fabric OS Troubleshooting and Diagnostics Guide 79 53-1003141-01

Page 96

Gathering additional information

•

If they are in the same zone, perform the following tasks:



Enter the portCamShow command on the host port to verify that the target is present.



Enter the portCamShow command on the target.



Enter the nsZoneMember command with the port ID for the zoned devices on the host and target to determine whether the name server is aware that these devices are zoned together.

3. Resolve zoning conflicts by putting the devices into the same zoning configuration.

4. Verify that no configuration is active by using the cfgActvShow command. Enter the defZone

show command to display the current state of the zone access mode and the access level.

The defZone command sets the default zone access mode to No Access

switch:admin> defzone --show Default Zone Access Mode committed - No Access transaction - No Transaction

Refer to “Zone conflicts” on page 75 for additional information.

Gathering additional information

Collect the data from a supportSave -n command. Then collect the data from the cfgTransShow command. For the port having the problem, collect the data from the filterPortShow port command.

80 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 97

Chapter

Diagnostic Features

In this chapter

•

Fabric OS diagnostics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

•

Diagnostic information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

•

Power-on self-test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

•

Switch status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

•

Using the spinFab and portTest commands . . . . . . . . . . . . . . . . . . . . . . . . . 87

•

Port information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

•

Equipment status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

•

System message log . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

•

Port log . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

•

Syslogd configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

•

Automatic trace dump transfers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

Fabric OS diagnostics

The purpose of the diagnostic subsystem is to evaluate the integrity of the system hardware.

Diagnostics are invoked in the one of the following ways:

•

Automatically during the power-on self-test (POST).

•

Automatically on an individual blade whenever it is installed into a director chassis.

•

Manually using Fabric OS CLI commands.

The error messages generated during these test activities are sent to the serial console and system message logs (output formats may differ slightly).

Use the diagHelp command to receive a list of all available diagnostic commands.

Diagnostic information

On the switch, you can enter the supportShow command to dump important diagnostic and status information to the session screen, where you can review it or capture its data. If you are using a Telnet client, you may have to set up the client to capture the data prior to opening the session.

Most information can be captured using the supportSave command and downloaded by FTP off the switch, but when you are collecting information from specialized commands, such as supportShow, this information must be captured using a Telnet client.

Fabric OS Troubleshooting and Diagnostics Guide 81 53-1003141-01

Page 98

NOTE

Power-on self-test

To save a set of files that customer support technicians can use to further diagnose the switch condition, enter the supportSave command. The command prompts for an FTP server, packages the following files, and sends them to the specified server:

•

Refer to“Automatic trace dump transfers” on page 100.

The output of the supportShow command.

Any core files, panic dumps or FFDC files that may have been generated.

System message (RAS) logs.

Other special feature logs.

Power-on self-test

By default, when you power on the system,.the boot loader automatically performs power-on self-tests and loads a Fabric OS kernel image. Likewise, if you issue the slotPower command, or insert a new blade, a power-on self test is run on that blade.

The POST tests provide a quick indication of hardware readiness when hardware is powered up. These tests do not require user input to function. They typically operate within several minutes, and support minimal validation because of the restriction on test duration. Their purpose is to give a basic health check before a new switch joins a fabric.

These tests are divided into two groups: POST1 and POST2. POST1 validates the hardware interconnect of the device, and POST2 validates the ability of the device to pass data frames between the ports. The specific set of diagnostic and test commands run during POST depends on the switch model.

When POST2 is running portloopbacktest mode 8 and if the peer blades are powered off while the test is running, the POST is expected to fail in the core blades as the mode 8 tests the peer links and the peer end is no longer active.

You can use the diagDisablePost command to disable both POST1 and POST2, and you can re-enable POST1 and POST2 using the diagEnablePost command.

The following example shows a typical boot sequence, including POST messages:

The system is coming up, please wait...

Read board ID of 0x80 from addr 0x23 Read extended model ID of 0x16 from addr 0x22 Matched board/model ID to platform index 4 PCI Bus scan at bus 0 ::: ::: Checking system RAM - press any key to stop test

Checking memory address: 00100000

System RAM test using Default POST RAM Test succeeded.

Press escape within 4 seconds to enter boot interface. Booting "Fabric Operating System" image.

Linux/PPC load:

82 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Page 99

Power-on self-test

BootROM command line: quiet Uncompressing Linux...done. Now booting the kernel Attempting to find a root file system on hda2... modprobe: modprobe: Can't open dependencies file /lib/modules/2.4.19/modules.dep (No such file or directory) INIT: version 2.78 booting INIT: Entering runlevel: 3 eth0: Link status change: Link Up. 100 Mbps Full duplex Auto (autonegotiation complete).

INITCP: CPLD Vers: 0x95 Image ID: 0x19 uptime: 2008; sysc_qid: 0 Fabric OS (Paulsa45) Paulsa45 console login: 2005/03/31-20:12:42, [TRCE-5000], 0,, INFO, ?, trace:, trace_buffer.c, line: 1170

2005/03/31-20:12:42, [LOG-5000], 0,, INFO, SW4100_P45, Previous message repeat 1 time(s), trace_ulib.c, line: 540 2005/03/31-20:12:43, [HAM-1004], 219,, INFO, SW4100_P45, Processor rebooted Unknown SNMP Research SNMP Agent Resident Module Version 15.3.1.4 Copyright 1989, 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001 SNMP Research, Inc. sysctrld: all services Standby FSSK 2: chassis0(0): state not synchronized FSSK 2: Services starting a COLD recovery 2005/03/31-20:12:48, [FSS-5002], 0,, INFO, SW4100_P45, chassis0(0): state not synchronized, svc.c, line: 318 2005/03/31-20:12:48, [FSS-5002], 0,, INFO, SW4100_P45, Services starting a COLD recovery, mdev.c, line: 638 2005/03/31-20:12:49, [MFIC-1002], 220,, INFO, Paulsa45, Chassis FRU header not programmed for switch NID, using defaults (applies only to FICON environments). sysctrld: all services Active 2005/03/31-20:12:50, [DGD-5001], 0,, INFO, SW4100_P45, Slot 0 has started POST., main.c, line: 1189 POST1: Started running Thu Mar 31 20:12:51 GMT 2005 POST1: Test #1 - Running turboramtest POST1: Test #2 - Running portregtest POST1: Script PASSED with exit status of 0 Thu Mar 31 20:12:54 GMT 2005 took (0:0:3) POST2: Started running Thu Mar 31 20:12:55 GMT 2005 POST2: Test #1 - Running portloopbacktest (SERDES) POST2: Test #2 - Running minicycle (SERDES) POST2: Running diagshow POST2: Script PASSED with exit status of 0 Thu Mar 31 20:13:12 GMT 2005 took (0:0:17) 2005/03/31-20:13:13, [BL-1000], 221,, INFO, Paulsa45, Initializing Ports... Enabling switch... 2005/03/31-20:13:13, [BL-1001], 222,, INFO, Paulsa45, Port Initialization Completed 2005/03/31-20:13:13, [EM-5012], 0,, INFO, SW4100_P45, EM: sent dumpready to ME., em.c, line: 2152 2005/03/31-20:13:13, [DGD-5002], 0,, INFO, SW4100_P45, Slot 0 has passed the POST tests., main.c, line: 936

If you choose to bypass POST or after POST completes, various system services are started and the boot process displays additional console status and progress messages.

Fabric OS Troubleshooting and Diagnostics Guide 83 53-1003141-01

Page 100

NOTE

Switch status

Disabling POST

A reboot is not required for diagDisablePost to take effect.

Disabling POST is not recommended and should only be done on the advice of your customer support technician.

1. Connect to the switch and log in with a user account that has admin privileges with the chassis-role permission.

2. Enter the diagDisablePost command.

This disables POST1 and POST2.

Enabling POST

A reboot is not required for diagEnablePost to take effect.

1. Connect to the switch and log in with a user account that has admin privileges with the chassis-role permission.

2. Enter the diagEnablePost command to enable POST without rebooting the switch.

switch:admin> diagenablepost Config update Succeeded Diagnostic POST is now enabled.

Switch status

Use the switchStatusShow command to display the overall status of the switch, including its power supplies, fans, and temperature. If the status of any one of these components is either marginal or down, the overall status of the switch is also displayed as marginal or down. If all components have a healthy status, the switch displays a healthy status.

To modify the rules used to classify the health of each component, use the switchStatusPolicySet command. To view the rules, use the switchStatusPolicyShow command.

In Fabric OS v7.2.0, the switchStatusShow command does work when Flow Vision is enabled. Refer to Fabric OS Command Reference and the Flow Vision Administrator’s Guide for more information.

84 Fabric OS Troubleshooting and Diagnostics Guide

53-1003141-01

Dell Brocade 6505, Brocade 5100, Brocade 5300 User Manual

Specifications and Main Features

Frequently Asked Questions

User Manual

About This Document

How this document is organized

Supported hardware and software

What’s new in this document

Document conventions

Text formatting

Command syntax conventions

Command examples

Notes, cautions, and warnings

Key terms

Additional information

Brocade resources

Other industry resources

Getting technical help

Document feedback

Introduction

Troubleshooting overview

Network Time Protocol

Most common problem areas

Questions for common symptoms

Gathering information for your switch support provider

Setting up your switch for FTP

Using the supportSave command

Building a case for your switch support provider

Capturing output from a console

Capturing command output

Basic information

Detailed problem information

Gathering additional information

General Troubleshooting

Licenses

Time

Frame Viewer

Switch message logs

Switch boot

Rolling Reboot Detection

FC-FC routing connectivity

Generating and routing an ECHO

Superping

Routing and statistical information

Performance issues

Connectivity

Port initialization and FCP auto-discovery process

Link issues

Connection problems

Checking the physical connection

Checking the logical connection

Checking the Name Server

Link failures

Determining a successful speed negotiation

Checking for a loop initialization failure

Checking for a point-to-point initialization failure

Marginal links

Correcting a port that has come up in the wrong mode

Troubleshooting a marginal link

Device login issues

Pinpointing problems with device logins

Media-related issues

Testing the external transmit and receive path of a port

Testing the internal components of a switch

Testing components to and from the HBA

Segmented fabrics

Reconciling fabric parameters individually

Downloading a correct configuration

Reconciling a domain ID conflict

Reconciling incompatible software features

Port mirroring

In-Order Delivery

Port mirroring considerations

Supported platforms

Maximum mirror connections

Configuring a port to be a mirror port

Adding a port mirror connection

Deleting a port mirror connection

Displaying port mirror connections

Configuration