Sun Microsystems Netra CT 810, Netra CT 410 Service Manual

Download

Page 1

Netra™CT Server Service Manual

For the Netra CT 810 Server and Netra CT 410 Server

Sun Microsystems, Inc. www.sun.com

Part No. 819-2741-10 February 2007, Revision A

Submit comments about this document at: http://www.sun.com/hwdocs/feedback

Page 2

Copyright 2007Sun Microsystems,Inc., 4150 Network Circle, SantaClara, California95054, U.S.A. Allrights reserved. Sun Microsystems, Inc.has intellectual property rights relating to technologythat isdescribed in thisdocument. Inparticular, and without

limitation, theseintellectual propertyrights may include one ormore ofthe U.S. patentslisted athttp://www.sun.com/patentsand one or more additional patents orpending patent applications in theU.S. and in other countries.

This documentand the product to which it pertainsare distributedunder licenses restricting theiruse, copying,distribution, and decompilation. Nopart of the product or of thisdocument may be reproduced in any formby any means without priorwritten authorizationof Sun andits licensors, if any.

Third-party software, including fonttechnology, iscopyrighted andlicensed from Sun suppliers. Parts ofthe productmay be derived from BerkeleyBSD systems,licensed from the University ofCalifornia. UNIX is a registered trademark in

the U.S.and in other countries, exclusivelylicensed throughX/Open Company,Ltd. Sun, SunMicrosystems, theSun logo, Java, docs.sun.com, Netra,Solstice DiskSuite, SunVTS, OPenBoot, andSolaris aretrademarks or

registered trademarks of SunMicrosystems, Inc.in the U.S. and inother countries. All SPARCtrademarks areused under license and are trademarksor registeredtrademarks ofSPARCInternational, Inc. inthe U.S.and in other

countries. Products bearingSPARCtrademarks arebased upon an architecture developed by SunMicrosystems, Inc. The OPENLOOK and Sun™ Graphical UserInterface was developed by SunMicrosystems, Inc.for its users and licensees. Sun acknowledges

the pioneeringefforts ofXerox inresearching anddeveloping the concept of visualor graphical user interfaces forthe computer industry.Sun holds anon-exclusive license from Xeroxto the Xerox GraphicalUser Interface,which license also covers Sun’slicensees who implement OPEN LOOK GUIsand otherwise comply with Sun’swritten license agreements.

U.S. GovernmentRights—Commercial use.Government users are subject tothe SunMicrosystems, Inc. standard license agreement and applicable provisions ofthe FAR and its supplements.

DOCUMENTATION IS PROVIDED "AS IS" AND ALL EXPRESS OR IMPLIED CONDITIONS, REPRESENTATIONS AND WARRANTIES, INCLUDING ANYIMPLIED WARRANTY OFMERCHANTABILITY,FITNESS FOR A PARTICULAR PURPOSE OR NON-INFRINGEMENT, ARE DISCLAIMED, EXCEPT TO THE EXTENT THAT SUCH DISCLAIMERS ARE HELD TO BE LEGALLY INVALID.

Copyright 2007Sun Microsystems,Inc., 4150 Network Circle, SantaClara, Californie95054, Etats-Unis. Tous droits réservés. Sun Microsystems, Inc.a les droits de propriété intellectuels relatants à latechnologie qui est décrit dansce document. En particulier,et sans la

limitation, cesdroits depropriété intellectuelspeuvent inclure un ou plusdes brevetsaméricains énumérés à http://www.sun.com/patents et un oules brevetsplus supplémentairesou les applicationsde breveten attente dans les Etats-Uniset dans les autres pays.

Ce produit oudocument est protégé par un copyright etdistribué avec des licences quien restreignentl’utilisation, la copie, la distribution,et la décompilation. Aucunepartie de ce produit ou document nepeut êtrereproduite sousaucune forme, par quelque moyenque ce soit, sans l’autorisation préalableet écrite de Sun etde ses bailleurs de licence,s’il yen a.

Le logicieldétenu par des tiers, etqui comprendla technologie relative aux policesde caractères,est protégépar un copyright et licenciépar des fournisseurs deSun.

Des partiesde ce produit pourrontêtre dérivées des systèmes BerkeleyBSD licenciés par l’Université deCalifornie. UNIX est une marque déposée auxEtats-Unis et dans d’autres pays et licenciéeexclusivement par X/Open Company, Ltd.

Sun, SunMicrosystems, lelogo Sun, Java, docs.sun.com, Netra,Solstice DiskSuite, SunVTS, OPenBoot, etSolaris sont des marques de fabrique ou desmarques déposéesde Sun Microsystems, Inc. auxEtats-Unis etdans d’autres pays.

Toutes les marquesSPARCsont utilisées sous licence etsont desmarques de fabrique ou desmarques déposéesde SPARC International, Inc. aux Etats-Uniset dans d’autres pays. Les produits portantles marquesSPARCsont basés sur une architecture développée par Sun Microsystems, Inc.

L’interfaced’utilisation graphiqueOPEN LOOK etSun™ aété développée par Sun Microsystems,Inc. pourses utilisateurs etlicenciés. Sun reconnaît les efforts depionniers deXerox pour la recherche et le développementdu conceptdes interfaces d’utilisationvisuelle ougraphique pour l’industriede l’informatique. Sun détient unelicense non exclusive de Xerox surl’interface d’utilisationgraphique Xerox, cette licence couvrant égalementles licenciées de Sun quimettent en place l’interface d’utilisation graphiqueOPEN LOOK etqui enoutre se conforment aux licencesécrites de Sun.

LA DOCUMENTATION EST FOURNIE "EN L’ÉTAT" ET TOUTES AUTRES CONDITIONS, DECLARATIONS ET GARANTIES EXPRESSES OU TACITES SONT FORMELLEMENT EXCLUES, DANSLA MESUREAUTORISEE PAR LA LOI APPLICABLE, YCOMPRIS NOTAMMENT TOUTE GARANTIE IMPLICITE RELATIVE A LA QUALITE MARCHANDE, A L’APTITUDE A UNE UTILISATION PARTICULIERE OU A L’ABSENCE DE CONTREFAÇON.

Please

Recycle

Page 3

Contents

Preface xxi

Part 1 Preparing for Service

1. Preparing for FRU Installation and Replacement 1–1

1.1 Tools Required 1–2

1.2 Attaching the Antistatic Wrist Strap 1–2

1.3 FRU Categories 1–3

1.3.1 Hot-Swappable FRUs 1–4

1.3.2 Cold-Swappable FRUs 1–4

1.4 Device Names 1–5

1.4.1 Device Names for I/O Board Slots in the Netra CT Servers 1–5

1.4.2 Device Names for hard drives 1–6

2. Powering the Server Off and On 2–1

2.1 Powering On the Server 2–2

2.2 Verifying Full Power-Up 2–6

2.3 Powering Off the Server 2–7

2.3.1 To Perform a Hardware Power-Down 2–7

2.3.2 To Perform a Software Power Down 2–8

iii

Page 4

3. Handling Boards and Assemblies 3–1

3.1 Handling Boards 3–2

3.2 Handling Assemblies 3–3

Part 2 Troubleshooting the System

4. Troubleshooting the System 4–1

4.1 Troubleshooting the System Using the System Status Panel 4–2

4.1.1 Locating and Understanding the System Status Panel 4–2

4.1.2 Using the System Status Panel LEDs to Troubleshoot the System 4–3

4.2 Troubleshooting the System Using prtdiag 4–9

4.3 Troubleshooting the System Using Diagnostic Software 4–14

4.4 Troubleshooting the System Using the Power-On Self Test (POST) 4–16

4.5 Troubleshooting the System Using the Alarm Card Software 4–18

4.6 Troubleshooting a Power Supply Using the Power Supply UnitLEDs 4–18

4.6.1 Troubleshooting the Power Supply Unit in the Netra CT 410 Server 4–18

4.6.2 Troubleshooting the Power Supply Units in the Netra CT 810 Server 4–19

4.7 Troubleshooting a Host CPU Board 4–20

4.7.1 General Troubleshooting Tips 4–20

4.7.2 Warning, Critical, and Shutdown Temperatures 4–22

4.7.3 General Troubleshooting Requirements 4–23

4.7.4 Mechanical Failures 4–23

4.7.5 Power-On Failures 4–23

4.7.6 Failures Subsequent to Power-On 4–24

4.7.7 Troubleshooting During POST/OpenBoot PROM and During Boot Process 4–24

iv Netra CT Server Service Manual • February 2007

Page 5

4.7.8 OpenBoot PROM On-Board Diagnostics 4–26

4.7.8.1 watch-clock 4–26

4.7.8.2 watch-net and watch-net-all 4–27

4.7.8.3 probe-scsi 4–28

4.7.8.4 test alias name, device path, -all 4–28

4.7.9 OpenBoot Diagnostics 4–29

4.7.9.1 PCI/PCIO 4–31

4.7.9.2 EBus DMA/TCR Registers 4–32

4.7.9.3 Ethernet 4–32

4.7.9.4 Parallel Port 4–33

4.7.9.5 Serial Port A 4–34

4.7.9.6 Serial Port B 4–34

4.7.9.7 NVRAM 4–35

4.7.9.8 All Above 4–36

Part 3 Replacing Hot-Swappable FRUs

5. Hot-Swap Software Commands 5–1

5.1 Understanding Hot-Swap 5–2

5.1.1 How High-Availability Hot-Swap Works 5–3

5.1.2 Hot-swap with Boards That Do Not Have Full Hot-swap 5–3

5.1.3 System Status Panel LED States and Meanings 5–3

5.2 Using the cfgadm Utility 5–5

5.2.1 Logging In to the Netra CT Server 5–5

5.2.2 Running the cfgadm Utility 5–5

5.2.3 Basic and Full Hot-Swap cfgadm Commands 5–6

5.2.3.1 Determining the Current Hot-Swap State 5–7

5.2.3.2 Enabling Basic Hot-Swap on I/O Slots 5–8

5.2.3.3 Enabling Full Hot-Swap on I/O Slots 5–9

Contents v

Page 6

6. Removing and Replacing Hot-Swappable Boards 6–1

6.1 Boards 6–2

6.1.1 Rules and Restrictions for Hot-Swapping Boards 6–5

6.1.1.1 General Rules and Restrictions for Hot-Swapping

Boards 6–5

6.1.1.2 Rules and Restrictions for Hot-Swapping the Host

Board 6–6

6.1.1.3 Rules and Restrictions for Hot-Swapping the Alarm

Card 6–9

6.1.2 Removing and Replacing Boards 6–9

6.1.2.1 Removing a Board 6–9

6.1.2.2 Inserting a Board 6–15

6.2 Rear Transition Modules 6–24

6.2.1 Host Rear Transition Modules 6–24

6.2.1.1 Removing a Host Rear Transition Module 6–27

6.2.1.2 Installing a Host Rear Transition Module 6–28

6.2.2 Satellite and I/O Rear Transition Modules 6–30

6.2.2.1 Satellite Rear Transition Module Sets 6–30

6.2.2.2 I/O Rear Transition Module Sets 6–32

6.2.2.3 Removing and Installing Satellite CPU or I/O Rear

Transition Module Sets 6–32

6.2.3 Alarm Rear Transition Module 6–35

6.2.3.1 Removing an Alarm Rear Transition Module 6–36

6.2.3.2 Installing an Alarm Rear Transition Module 6–38

7. Removing and Replacing Hard Drives and Removeable Media 7–1

7.1 Hard Drive 7–2

7.1.1 Removing a Hard Drive 7–3

7.1.2 Installing a Hard Drive 7–8

vi Netra CT Server Service Manual • February 2007

Page 7

7.2 DVD or DAT Drive 7–10

7.2.1 Removing a DVD or DAT Drive 7–10

7.2.2 Installing a DVD or DAT Drive 7–13

8. Removing and Replacing Hot-Swappable Subassemblies 8–1

8.1 System Status Panel 8–2

8.1.1 Removing the System Status Panel 8–2

8.1.2 Replacing the System Status Panel 8–4

8.2 System Controller Board 8–6

8.2.1 Removing the System Controller Board 8–6

8.2.2 Replacing the System Controller Board 8–8

8.3 Air Filters 8–10

8.3.1 Removing the Air Filters 8–10

8.3.2 Replacing the Air Filters 8–12

8.4 Power Supply Unit 8–13

8.4.1 Removing a Hot-Swappable Power Supply Unit 8–13

8.4.2 Replacing a Hot-Swappable Power Supply Unit 8–16

8.5 Fan Trays 8–17

8.5.1 Removing a Fan Tray 8–17

8.5.2 Replacing a Fan Tray 8–21

Part 4 Replacing Cold-Swappable FRUs

9. Removing and Replacing the Servers 9–1

9.1 Removing a Server 9–2

9.2 Installing a Server 9–9

Contents vii

Page 8

10. Removing and Replacing Cold-Swappable Subassemblies 10–1

10.1 Cold Swappable Hard dDrives 10–2

10.1.1 Removing a Cold-Swappable Hard Drive 10–2

10.1.2 Replacing a Cold-Swappable Hard Drive 10–4

10.2 Cold-Swappable Power Supply Unit 10–5

10.2.1 Removing a Cold-Swappable Power Supply Unit 10–5

10.2.2 Replacing a Cold-Swappable Power Supply Unit 10–8

Part 5 Illustrated Parts List

11. Illustrated Parts List 11–1

11.1 Chassis Components 11–2

11.2 Netra CT 810 Server 11–4

11.3 Netra CT 410 Server 11–6

Part 6 Appendixes, Glossary, and Index

A. Connector Pinouts A–1

A.1 Host CPU Boards A–2

A.1.1 Netra CP2140 Host CPU Board A–2

A.1.1.1 Ethernet A Port A–3

A.1.1.2 TTY A Port A–3

A.1.2 Netra CP2500 Host CPU Board A–5

A.1.2.1 PMC Connector A–5

A.1.2.2 PMC Connector Interfaces A–7

A.1.2.3 Front Panel Serial Connector A–11

A.1.2.4 Backplane Connectors A–12

A.1.2.5 CompactPCI J1/P1 Connector Pinouts A–13

A.1.2.6 CompactPCI J1/P1 Signal Descriptions A–15

A.1.2.7 CompactPCI J2/P2 Connector Pinouts A–16

A.1.2.8 CompactPCI J2/P2 Signal Descriptions A–17

viii Netra CT Server Service Manual • February 2007

Page 9

A.1.2.9 CompactPCI J3/P3 Connector Pinouts A–17

A.1.2.10 CompactPCI J3/P3 Signal Descriptions A–18

A.1.2.11 CompactPCI J5/P5 Connector Pinouts A–19

A.1.2.12 CompactPCI J5/P5 Signal Descriptions A–20

A.1.2.13 DIP Switch Settings A–20

A.2 Host Rear Transition Modules A–23

A.2.1 Netra CP2140 Host RTMs A–23

A.2.1.1 SCSI (VHDC) A–25

A.2.1.2 TTY A and B Ports A–26

A.2.1.3 Ethernet A and B Ports A–26

A.2.1.4 PIM Module Interface A–27

A.2.2 Netra CP2500 Host RTMs A–28

A.2.2.1 Serial Ports A–29

A.2.2.2 Ethernet Connectors A–30

A.2.2.3 On-Board Interfaces and Connectors on the Netra

CP2500 Transition Module A–33

A.2.2.4 Backplane Interfaces A–38

A.3 Alarm Card, 6U Single-Wide A–44

A.3.1 Alarm Port A–45

A.3.2 Ethernet Ports 1 and 2 A–46

A.3.3 COM Ports 1 and 2 A–47

A.4 Alarm Card, 3U Double-Wide A–48

A.4.1 Ethernet Ports 1 and 2 A–49

A.4.2 COM Ports 1 and 2 A–50

A.4.3 Alarm Port A–51

A.5 Alarm Rear Transition Module A–52

A.5.1 Ethernet Ports 1 and 2 A–53

A.5.2 COM Ports 1 and 2 A–54

A.5.3 Alarm Port A–55

Contents ix

Page 10

B. Connecting a Terminal Console to a Server B–1

B.1 To Use an ASCII Terminal B–2

B.2 To Use a Solaris Workstation B–4

B.3 To Use a PC Laptop B–5

C. Error Messages C–1

C.1 Generic Error Messages C–2

C.2 scsb Error Messages C–3

C.3 Anticipated Hardware Failure C–9

C.3.1 Transient Interrupts C–9

C.3.2 Soft Hang C–10

C.4 I2C Complaints C–11

C.5 Bus Busy Complaints C–12

D. System Specifications D–1

D.1 Physical Specifications D–1

D.2 Electrical Specifications D–2

D.3 Environmental Specifications D–3

Glossary Glossary–1

Index Index–1

x Netra CT Server Service Manual • February 2007

Page 11

Figures

FIGURE 1-1 Attaching the Antistatic Wrist Strap 1–3

FIGURE 2-1 Locating the Power Supply Locking Mechanism on the Netra CT 810 Server 2–3

FIGURE 2-2 System Status Panel Locations 2–4

FIGURE 2-3 System Power Button and System Power LED Locations (Netra CT 810 Server) 2–5

FIGURE 2-4 System Power Button and System Power LED Locations (Netra CT 410 Server) 2–5

FIGURE 4-1 System Status Panel Locations 4–2

FIGURE 4-2 System Status Panel (Netra CT 810 Server) 4–3

FIGURE 4-3 System Status Panel (Netra CT 410 Server) 4–3

FIGURE 4-4 Power and Okay to Remove LEDs 4–3

FIGURE 4-5 Power and Fault LEDs 4–4

FIGURE 4-6 Connectors on the Netra CP2140 Host Rear Transition Module 4–21

FIGURE 4-7 Connectors on the Netra CP2500 Rear Transition Module Host (RTM-H) Board 4–22

FIGURE 6-1 Boards Within a Netra CT 810 Server 6–3

FIGURE 6-2 Boards Within a Netra CT 410 Server 6–4

FIGURE 6-3 Location of SW0501 on SMC Module 6–7

FIGURE 6-4 Switch SW0501 in Closed Position (Default) for the Netra CT Server 6–7

FIGURE 6-5 Reconfiguring a Netra CP2500 cPSB Board for Use as a cPCI Board 6–8

FIGURE 6-6 Loosening the Ejection Lever Captive Screws 6–12

FIGURE 6-7 Unlocking the Ejection Levers 6–13

FIGURE 6-8 LEDs on the System Status Panel (Netra CT 810 Server) 6–13

Page 12

FIGURE 6-9 LEDs on the System Status Panel (Netra CT 410 Server) 6–14

FIGURE 6-10 Aligning the Board with the Board Cage Cutouts 6–17

FIGURE 6-11 Tightening the Ejection Lever Captive Screws 6–18

FIGURE 6-12 Connectors for the Single-Wide 6U Alarm Card 6–20

FIGURE 6-13 Connectors for the Double-Wide 3U Alarm Card 6–21

FIGURE 6-14 Connectors on the Netra CP2140 Host CPU Board 6–22

FIGURE 6-15 Connectors on the Netra CP2500 Host CPU Board 6–23

FIGURE 6-16 Supported and Unsupported Netra CP2140 Host Rear Transition Module for a Netra CT

Server 6–25

FIGURE 6-17 Netra CP2500 Rear Transition Module-Host (RTM-H) 6–26

FIGURE 6-18 Locating the Host Rear Transition Module in a Netra CT 810 Server

(Top View) 6–27

FIGURE 6-19 Locating the Host Rear Transition Module in a Netra CT 410 Server

(Top View) 6–28

FIGURE 6-20 Aligning the Module with the Rear Cage Cutouts 6–29

FIGURE 6-21 Supported Locations for Satellite CPU or I/O Rear Transition Module Sets in a Netra CT 810

Server (Top View) 6–31

FIGURE 6-22 Supported Locations for Satellite CPU or I/O Rear Transition Module Sets in a Netra CT 410

Server (Top View) 6–31

FIGURE 6-23 Supported Location for the Alarm Rear Transition Module in a Netra CT 810 Server (Top

View) 6–36

FIGURE 6-24 Supported Location for the Alarm Rear Transition Module in a Netra CT 410 Server (Top

View) 6–37

FIGURE 6-25 Connectors on the Alarm Rear Transition Module 6–39

FIGURE 7-1 Locating the Hard Drive LEDs on the System Status Panel (Netra CT 810 Server) 7–6

FIGURE 7-2 Drive Bay Cover Locations 7–7

FIGURE 7-3 Removing the Removable Media Module from a Netra CT 810 Server 7–13

FIGURE 8-1 System Status Panel Locations 8–2

FIGURE 8-2 Removing the System Status Panel (Netra CT 810 Server) 8–3

FIGURE 8-3 Removing the System Status Panel (Netra CT 410 Server) 8–4

FIGURE 8-4 Positioning the System Status Panel (Netra CT 810 Server) 8–5

FIGURE 8-5 Positioning the System Status Panel (Netra CT 410 Server) 8–5

xii Netra CT Server Service Manual • February 2007

Page 13

FIGURE 8-6 Locating the System Controller Board LEDs on the System Status Panel (Netra CT 810

Server) 8–7

FIGURE 8-7 Locating the System Controller Board LEDs on the System Status Panel (Netra CT 410

Server) 8–7

FIGURE 8-8 Removing a System Controller Board 8–8

FIGURE 8-9 Inserting a System Controller Board 8–9

FIGURE 8-10 Locating the Air Filters (Netra CT 810 Server) 8–11

FIGURE 8-11 Locating the Air Filters (Netra CT 410 Server) 8–12

FIGURE 8-12 Locating the Power Supply Unit LEDs on the System Status Panel (Netra CT 810 Server) 8–

FIGURE 8-13 Unlocking a Power Supply Unit 8–15

FIGURE 8-14 Removing a Power Supply Unit 8–16

FIGURE 8-15 Locating the Fan Tray LEDs on the System Status Panel (Netra CT 810 Server) 8–18

FIGURE 8-16 Locating the Fan Tray LEDs on the System Status Panel (Netra CT 410 Server) 8–18

FIGURE 8-17 Locating the Fan Trays in a Netra CT 810 Server 8–19

FIGURE 8-18 Locating the Fan Trays in a Netra CT 410 Server 8–19

FIGURE 8-19 Removing a Fan Tray from a Netra CT 810 Server 8–20

FIGURE 8-20 Removing a Fan Tray from a Netra CT 410 Server 8–20

FIGURE 9-1 Power Supply Units and Power Sources 9–3

FIGURE 9-2 Loosening the Screws at the Top and Bottom of a Netra CT 810 Server 9–5

FIGURE 9-3 Loosening the Screws at the Top and Bottom of a Netra CT 410 Server 9–6

FIGURE 9-4 Removing or Inserting a Netra CT 810 Server 9–7

FIGURE 9-5 Removing or Inserting a Netra CT 410 Server 9–8

FIGURE 10-1 Drive Bay Cover Locations 10–3

FIGURE 10-2 Removing or Replacing a Cold-Swappable Power Supply Unit From a Netra CT 810

Server 10–6

FIGURE 10-3 Removing or Replacing a Cold-Swappable Power Supply Unit From a Netra CT 410

Server 10–7

FIGURE 10-4 Removing a Cold-Swappable Power Supply Unit From a Netra CT 410 Server 10–8

FIGURE 11-1 Illustrated Parts of Chassis and Components 11–3

FIGURE 11-2 Illustrated Parts Breakdown, Netra CT 810 Server 11–5

FIGURE 11-3 Illustrated Parts Breakdown, Netra CT 410 Server 11–7

Figures xiii

Page 14

FIGURE A-1 Connectors on the Netra CP2140 Host CPU Board A–2

FIGURE A-2 RJ-45 Ethernet Connector Diagram A–3

FIGURE A-3 TTY A Connector A–4

FIGURE A-4 Connectors on the Netra CP2500 Host CPU Board A–5

FIGURE A-5 Netra CP2500 Board PMC Port Connectors A–6

FIGURE A-6 PMC Connector Slot Connector Pins A–6

FIGURE A-7 Front Panel Serial Port Diagram A–11

FIGURE A-8 Backplane Connector Contact Numbering A–13

FIGURE A-9 SW3301 DIP Switch Location A–21

FIGURE A-10 Connectors on the Netra CP2140 Rear Transition Module A–24

FIGURE A-11 RJ-45 Ethernet Connector Diagram A–26

FIGURE A-12 Connectors on the Netra CP2500 Rear Transition Module-Host (RTM-H) A–28

FIGURE A-13 Serial Port Connector Pins A–29

FIGURE A-14 Ethernet Port Connector Pins A–31

FIGURE A-15 On-Board Connectors and Interfaces for the Netra CP2500 RTM-H A–33

FIGURE A-16 On-Board Connectors and Interfaces for the Netra CP2500 RTM-S A–34

FIGURE A-17 I

FIGURE A-18 J0501 and J0502 Connector Pins A–35

FIGURE A-19 CompactPCI RJ3 Connector (J0201) Pins A–38

FIGURE A-20 CompactPCI RJ4 Connector (J0301) Pins A–40

FIGURE A-21 CompactPCI RJ5 Connector (J0401) Pins A–42

FIGURE A-22 Connector Ports in the 6U Single-Wide Alarm Card A–44

FIGURE A-23 Alarm Port A–45

FIGURE A-24 RJ-45 Ethernet Connector Diagram A–46

FIGURE A-25 RJ-45 Ethernet Connector Diagram A–47

FIGURE A-26 Connectors on the Alarm Card (Netra CT 410 Server) A–48

FIGURE A-27 RJ-45 Ethernet Connector Diagram A–49

FIGURE A-28 RJ-45 Ethernet Connector Diagram A–50

FIGURE A-29 Alarm Port A–51

FIGURE A-30 Connectors on the Alarm Rear Transition Module A–52

C Serial Bus Access Header Pins A–34

xiv Netra CT Server Service Manual • February 2007

Page 15

FIGURE A-31 RJ-45 Ethernet Connector Diagram A–53

FIGURE A-32 COM Port 1 and 2 A–54

FIGURE A-33 Alarm Port A–55

Figures xv

Page 16

xvi Netra CT Server Service Manual • February 2007

Page 17

Tables

TABLE 1-1 Device Names for I/O Board Slots in the Netra CT 810 Server 1–5

TABLE 1-2 Device Names for I/O Board Slots in the Netra CT 410 Server 1–5

TABLE 4-1 System Status Panel LEDs for the Netra CT 810 Server 4–4

TABLE 4-2 System Status Panel LEDs for the Netra CT 410 Server 4–5

TABLE 4-3 CompactPCI Board LED States and Meanings 4–6

TABLE 4-4 Meanings of Power and Okay to Remove LEDs 4–7

TABLE 4-5 Meanings of Power and Fault LEDs 4–8

TABLE 4-6 Apost Tests and Values through diagconf 4–15

TABLE 4-7 Selected OpenBoot PROM On-Board Diagnostic Tests 4–28

TABLE 5-1 Netra CT System Hot-Swap Modes 5–2

TABLE 5-2 CompactPCI Board LED States and Meanings on the System Status Panel 5–4

TABLE 11-1 FRUs for the Chassis 11–2

TABLE 11-2 FRUs for the Netra CT 810 Server 11–4

TABLE 11-3 FRUs for the Netra CT 410 Server 11–6

TABLE A-1 Ethernet Connector Pinouts, CPU Board (J2301) A–3

TABLE A-2 TTY A Connector Pinouts A–4

TABLE A-3 PMC Jn1 Connector Interface A–7

TABLE A-4 PMC Jn2 Connector Interface A–9

TABLE A-5 PMC Jn4 Connector Interface A–10

TABLE A-6 Serial Micro DB9 Connector Pinouts A–12

xvii

Page 18

TABLE A-7 CompactPCI J2/P2 Connector Pin Assignments A–14

TABLE A-8 CompactPCI J1/P1 Signal Descriptions A–15

TABLE A-9 CompactPCI J2/P2 Connector Pin Assignments A–16

TABLE A-10 CompactPCI J2/P2 Signal Descriptions A–17

TABLE A-11 CompactPCI J3/P3 Connector Pin Assignments A–18

TABLE A-12 CompactPCI J3/P3 Signal Descriptions A–18

TABLE A-13 CompactPCI J5/P5 Connector Pin Assignments A–19

TABLE A-14 Serial COM Port and RS232 Level CompactPCI J5/P5 Signal Descriptions A–20

TABLE A-15 Miscellaneous CompactPCI J5/P5 Signal Descriptions A–20

TABLE A-16 SW3301 Switch Descriptions A–22

TABLE A-17 SCSI Port Pinouts, CPU RTC A–25

TABLE A-18 Signal Interfaces for TTY A and B Port Connectors A–26

TABLE A-19 Ethernet A and B Connector Pinouts, CPU RTM A–27

TABLE A-20 Serial Port A A–29

TABLE A-21 Serial Port B A–30

TABLE A-22 ENET0 and ENET1 A–32

TABLE A-23 ENET2 and ENET3 A–32

TABLE A-24 I

TABLE A-25 J0501 Pin Assignments A–35

TABLE A-26 J0502 Pin Assignments A–36

TABLE A-27 CompactPCI RJ3 Connector (J0201) Pin Assignments A–38

TABLE A-28 CompactPCI RJ4 Connector (J0301) Pin Assignments A–40

TABLE A-29 CompactPCI RJ5 Connector (J0401) Pin Assignments A–42

TABLE A-30 Alarm Port Pinouts A–45

TABLE A-31 Ethernet Port 1 and 2 Pinouts A–46

TABLE A-32 COM Port 1 Pinouts A–47

TABLE A-33 COM Port 2 Pinouts A–47

TABLE A-34 Ethernet Port 1 and 2 Pinouts A–49

TABLE A-35 COM Port 1 Pinouts A–50

TABLE A-36 COM Port 2 Pinouts A–50

C Serial Bus Access Header Pin Assignments A–34

xviii Netra CT Server Service Manual • February 2007

Page 19

TABLE A-37 Alarm Port Pinouts A–51

TABLE A-38 Ethernet Port 1 and 2 Pinouts A–53

TABLE A-39 COM Port 1 and 2 Connector Pinouts, Alarm Rear Transition module A–54

TABLE A-40 Alarm Port Pinouts A–55

TABLE B-1 Connecting an ASCII Terminal to a CPU Board or Rear Transition Module B–2

TABLE B-2 Connecting an ASCII Terminal to an Alarm Card or Alarm Rear Transition Module B–3

TABLE B-3 Connecting a Solaris Workstation to a CPU Board or Rear Transition Module B–4

TABLE B-4 Connecting a Solaris Workstation to an Alarm Card or Alarm Rear Transition Module B–4

TABLE B-5 Connecting a PC Laptop to a CPU Board or Rear Transition Module B–5

TABLE B-6 Connecting a PC Laptop to an Alarm Card or Alarm Rear Transition Module B–6

TABLE C-1 Netra CT Server Error Messages C–1

TABLE D-1 Netra CT Server Chassis Physical Specifications D–1

TABLE D-2 Physical Specifications, Netra CT 810 Server D–2

TABLE D-3 Physical Specifications, Netra CT 410 Server D–2

TABLE D-4 Power Requirements D–2

TABLE D-5 Environmental Specifications D–3

Tables xix

Page 20

xx Netra CT Server Service Manual • February 2007

Page 21

Preface

The Netra CT Server Service Manual contains procedures for the removal and replacement of the field-replaceable units (FRUs) in a Netra™ CT server.

The intended reader of this manual is a service provider or experienced system administrator who has experience installing hardware—systems and components—and has used the Solaris™ Operating System (Solaris OS). The reader should be comfortable with LAN fundamentals and with networking in general.

Before performing the procedures described in this book, you should have completed the installation and setup of the Netra CT server as described in the Netra CT Server Installation Guide.

Safety and Compliance

All Netra CT servers are shipped with the Netra CT Server Safety and Compliance Manual, which specifies the environmental and electrical safety requirements for the product and contains compliance certification for various countries.

xxi

Page 22

How This Book Is Organized

Part 1 “Preparing for Service”

■ Chapter 1 describes the procedures you must perform before installing, removing,

or replacing field-replaceable units (FRUs) in a Netra CT server.

■ Chapter 2 provides hardware and software procedures for powering the Netra CT

server on and off.

■ Chapter 3 provides instructions for handling Compact PCI (cPCI) boards and

assemblies correctly.

Part 2 “Troubleshooting the System”

■ Chapter 4 provides troubleshooting procedures for the Netra CT server.

Part 3 “Replacing Hot-Swappable FRUs”

■ Chapter 5 describes hot-swap software commands.

■ Chapter 6 gives instructions for removing and replacing hot-swappable boards.

■ Chapter 7 contains instructions for removing and replacing hard drives and

removable media.

■ Chapter 8 gives instructions for removing and replacing hot-swappable

subassemblies.

Part 4 “Replacing Cold-Swappable FRUs”

■ Chapter 9 provides instructions for removing and replacing the Netra CT 810

server and Netra CT 410 server.

■ Chapter 10 gives instructions for removing and replacing cold-replaceable

subassemblies.

Part 5 “Illustrated Parts List”

■ Chapter 11 gives the illustrated parts list for the Netra CT server.

Part 6 “Appendixes, Glossary, and Index”

■ Appendix A lists the connector pinouts for boards in the Netra CT server.

■ Appendix B provides instructions for connecting a terminal console to the server.

■ Appendix C gives the error messages for the Netra CT server.

■ Appendix D gives the system specifications for the Netra CT server.

xxii Netra CT Server Service Manual • February 2007

Page 23

Using UNIX Commands

This document contains only limited information on basic UNIX® commands and procedures such as shutting down a system, booting a system, and configuring devices.

Refer to one or more of the following for this information:

■ Solaris Handbook for Sun Peripherals (shipped in AnswerBook2™ form, available in

printed form as an at-cost option)

■ AnswerBook2™ online documentation for the Solaris OS

■ Other software documentation that you received with your system

Preface xxiii

Page 24

Typographic Conventions

Typeface or Symbol Meaning Examples

AaBbCc123 The names of commands, files,

and directories; on-screen computer output

AaBbCc123 What you type, when

contrasted with on-screen computer output

AaBbCc123 Book titles, new words or terms,

words to be emphasized

Command-line variable; replace with a real name or value

Edit your .login file. Use ls -a to list all files. % You have mail.

% Password:

Read Chapter 6 in the User’s Guide. These are called class options. You must be superuser to do this.

To delete a file, type rm filename.

Shell Prompts

Shell Prompt

C shell machine_name%

C shell superuser machine_name#

Bourne shell and Korn shell $

Bourne shell and Korn shell superuser #

xxiv Netra CT Server Service Manual • February 2007

Page 25

Related Documentation

The documents listed as online are available at:

http://www.sun.com/products-n-solutions/hardware/docs/

Application Title Part Number Format Location

Installation Netra CT Server Release Notes 819-2739-xx PDF

HTML

Installation Netra CT Server Installation Guide 819-2740-xx PDF

HTML

Service Netra CT Server Service Manual 819-2741-xx PDF

HTML

Introduction Netra CT Product Overview 819-2742-xx PDF

HTML

System Admin

Safety Netra CT Server Safety and Compliance

Installation Netra CP2500 Board Release Notes 819-1748-xx PDF

Installation Netra CP2500 Board Installation and

Installation Netra CP2500 Board Programming Guide 819-1749-xx PDF

Netra CT Server System Administration Guide

Manual

Technical Reference Manual

819-2743-xx PDF

HTML

819-2746-xx PDF

HTML

819-1747-xx PDF

HTML

Online

Safety Netra CP2500 Board Safety and Compliance

Guide

Installation Netra CP2500 RTM Release Notes 819-1752-xx PDF

Installation Netra CP2500 RTM Installation and

Reference Guide

You may want to refer to documentation on the following software for additional information: the Solaris OS and the Netra High Availability (HA) Suite.

819-1750-xx PDF

HTML

819-1753-xx PDF

HTML

Online

Preface xxv

Page 26

Documentation, Support, and Training

Sun Function URL Description

Documentation http://www.sun.com/documentation/ Download PDF and HTML documents,

and order printed documents

Support and Training

http://www.sun.com/supportraining/ Obtain technical support, download

patches, and learn about Sun courses

Third-Party Web Sites

Sun is not responsible for the availability of third-party web sites mentioned in this document. Sun does not endorse and is not responsible or liable for any content, advertising, products, or other materials that are available on or through such sites or resources. Sun will not be responsible or liable for any actual or alleged damage or loss caused by or in connection with the use of or reliance on any such content, goods, or services that are available on or through such sites or resources.

Sun Welcomes Your Comments

Sun is interested in improving its documentation and welcomes your comments and suggestions. You can submit your comments by going to:

http://www.sun.com/hwdocs/feedback

Please include the title and part number of your document with your feedback:

Netra CT Server Service Manual, part number 819-2741-10

xxvi Netra CT Server Service Manual • February 2007

Page 27

PA R T

1 Preparing for Service

Preparing for FRU Installation and Replacement Chapter 1

Powering the Server Off and On Chapter 2

Handling Boards and Assemblies Chapter 3

Page 28

Page 29

CHAPTER

Preparing for FRU Installation and Replacement

This chapter describes the steps you need to take before you install, remove, or replace a field-replaceable unit (FRU) in your Netra CT server. This chapter is divided into the following sections:

■ Section 1.1, “Tools Required” on page 1-2

■ Section 1.2, “Attaching the Antistatic Wrist Strap” on page 1-2

■ Section 1.3, “FRU Categories” on page 1-3

■ Section 1.4, “Device Names” on page 1-5

1-1

Page 30

1.1 Tools Required

You need the following tools to install, remove, or replace components in a Netra CT server:

■ No. 1 Phillips screwdriver

■ No. 2 Phillips screwdriver

1.2 Attaching the Antistatic Wrist Strap

Most field-replaceable units have sensitive electronic components that require antistatic precautions.

1. Get the antistatic wrist strap and electrostatic discharge mat from the shipping kit.

2. Place the electrostatic discharge mat close to the system.

3. Unwrap the first two folds of the wrist strap and wrap the adhesive side firmly against your wrist.

4. Peel the liner from the copper foil at the opposite end of the wrist strap, and attach the copper end of the strap to a bare metal area on the Netra CT server or on the chassis, as shown in

FIGURE 1-1.

1-2 Netra CT Server Service Manual • February 2007

Page 31

A BORT

RESET

ALARM/USER

H ERN

HDD1

mi cr o sy

st e m s

C O

H O

W A

ALARM

B ORT

RESET

ALARM

/ USER

READ

T H E

R N

E T

COM 2

COM 1 ENET 1 ENET 2

i c r o s y s

t em

COM 2

H OT S W

COM 1 ENET 1 ENET 2

C O

ALARM

H O

A P

T SW

A P

COM 2

COM 1 ENET 1 ENET 2

ALARM

H O T S W A

FIGURE 1-1 Attaching the Antistatic Wrist Strap

1.3 FRU Categories

In general, the FRUs in a Netra CT server can be divided into two categories:

■ Hot-installable/replaceable (referred to hereafter as hot-swappable), meaning that

you can install or remove and replace the FRU while the server is running, without interrupting the operation of the server. This category of FRU sometimes requires that you enter hot-swap software commands before and after an installation or a removal/replacement to incorporate the new FRU in the system correctly. For information on hot-swap commands, see Chapter 5.

■ Non-hot-swappable, or cold-swappable, meaning that you must halt (and, in some

cases, also power off) the server before installing or removing and replacing a FRU. Note that you must halt power only to the server where you would be removing and replacing the FRUs; any other servers installed in the same chassis can remain running.

Redundant power supplies and hard drives in a Netra CT 810 server are a variation of the hot-swappable category. You can replace a single power supply in a redundant configuration while the server is running, without having to enter any software commands, or a single hard drive after you have unmounted the drive or disengaged it if you are running a disk management software package.

Chapter 1 Preparing for FRU Installation and Replacement 1-3

Page 32

1.3.1 Hot-Swappable FRUs

The FRUs listed below are hot-swappable:

■ I/O boards

■ Satellite CPU boards and satellite RTMs

■ Alarm cards for both the Netra CT 810 server and the Netra CT 410 server

■ Alarm rear transition modules

■ Host CPU board

■ Host CPU rear transition module (rear-access model only)

■ Either of the two hard drives in a Netra CT 810 server

■ DVD or DAT (digital audio tape) drives occupying the removable media module

■ Fan trays

■ System status panels

■ System controller boards

■ Either of the two power supply units in a Netra CT 810 server

■ Power supply unit air filters

■ Main air filters

The instructions for removing and replacing these hot-swappable FRUs are covered in Part 3.

1.3.2 Cold-Swappable FRUs

The FRUs listed below are cold-swappable:

■ Single power supply in a Netra CT 410 server

■ Single or lone remaining power supply in a Netra CT 810 server

■ Single hard drive in a Netra CT 410 server

■ Single or lone remaining hard drive in a Netra CT 810 server

■ Power distribution units

■ Powered-off or empty servers

The instructions for removing and replacing these cold-swappable FRUs are covered in Part 4.

1. The alarm card andalarm reartransition module are hot-swappable only if thehost CPU board and system

controller board are alsoinstalled in the Netra CTserver.

2. The host CPUboard andthe host CPU rear transition modules are hot-swappableonly if the alarm card and

system controllerboard are alsoinstalled in the Netra CTserver.

3. The system controller board ishot-swappable only if the alarmcard andhost CPUboard areinstalled in the

Netra CTserver.

1-4 Netra CT Server Service Manual • February 2007

Page 33

1.4 Device Names

The following sections provide information about device names for I/O slots and hard drives.

1.4.1 Device Names for I/O Board Slots in the Netra CT Servers

TABLE 1-1 gives the device names for the I/O board slots in the Netra CT 810 server,

and

TABLE 1-2 gives the device names for the I/O board slots in the Netra CT 410

server.

TABLE 1-1 Device Names for I/O Board Slots in the Netra CT 810 Server

Slot Number Device Name

2 /devices/pci@1f,0/pci@1/pci@1/pci@f

3 /devices/pci@1f,0/pci@1/pci@1/pci@e

4 /devices/pci@1f,0/pci@1/pci@1/pci@d

5 /devices/pci@1f,0/pci@1/pci@1/pci@c

6 /devices/pci@1f,0/pci@1/pci@1/pci@b

7 /devices/pci@1f,0/pci@1/pci@1/pci@a

8 /devices/pci@1f,0/pci@1/pci@1/pci@8

TABLE 1-2 Device Names for I/O Board Slots in the Netra CT 410 Server

Slot Number Device Name

1 /devices/pci@1f,0/pci@1/pci@1/pci@8

2 /devices/pci@1f,0/pci@1/pci@1/pci@f

4 /devices/pci@1f,0/pci@1/pci@1/pci@e

5 /devices/pci@1f,0/pci@1/pci@1/pci@d

Note – For Netra CP2500 boards, pcia and pcib probe lists do not apply.

Chapter 1 Preparing for FRU Installation and Replacement 1-5

Page 34

1.4.2 Device Names for hard drives

Following are the device names for the hard drives installed in the Netra CT servers:

■ Netra CT 810 server:

■ HDD 0—c0t0d0

■ HDD 1—c1t1d0

■ Netra CT 410 server:

■ HDD 0—c0t0d0

1-6 Netra CT Server Service Manual • February 2007

Page 35

CHAPTER

Powering the Server Off and On

This chapter provides instructions on powering the Netra CT server on and off. This chapter contains the following topics:

■ Section 2.1, “Powering On the Server” on page 2-2

■ Section 2.3, “Powering Off the Server” on page 2-7

Note – Whenever you reboot or power your server on and off, the hot-swap states

revert back to the default full hot-swap state for all I/O slots. If you want basic hotswap on an I/O slot, you must manually reset the I/O slot to basic hot-swap after rebooting or powering your server on and off.

Note – You can also power the Netra CT server on and off through the alarm card

using the poweroff and poweron commands. Refer to the Netra CT Server System Administration Guide for more information.

2-1

Page 36

2.1 Powering On the Server

Note – You must have the host CPU board, alarm card, and system controller board

installed in the Netra CT server before you can power it on; the server will not power on properly unless these three are installed. Once the system is powered on and running, then you can hot-swap any of the three components as documented in this manual.

Note – Do not reboot the server if the alarm card is in the process of resetting.

The system power on sequence is as follows:

1. The alarm card comes up and powers on the power supplies

2. The alarm card powers on I/O boards

3. The alarm card powers on the host CPU board, and the Solaris OS boots

4. The alarm card powers on satellite CPU boards supported by Sun Microsystems, such as the Netra CP2160 and the Netra CP2500.

The power on sequence must finish before you can use the alarm card CLI commands poweron cpu_node and poweroff cpu_node.

2-2 Netra CT Server Service Manual • February 2007

Page 37

1. Verify that the power supply locking mechanisms on the server you are about to

power on are in the locked (), or down, positions.

FIGURE 2-1 shows the location of the power supply locking mechanisms on the Netra

CT server. Make sure you lock both power supplies if you are powering on a Netra CT 810 server.

COM 1 ENET 1 ENET 2COM 2

FIGURE 2-1 Locating the Power Supply Locking Mechanism on the Netra CT 810 Server

When the power supply locking mechanism is locked, the green LEDs on the power supplies flash, indicating that the power supplies are powered on, but the server has not been powered on yet. Also, the power distribution unit (PDU) LEDs on the system status panel should be in the following states:

Green LED ( ) is ON Amber LED ( ) is OFF

This status tells you that the power supply locking mechanisms are in the locked (down) position and the PDUs are receiving power and functioning properly.

Chapter 2 Powering the Server Off and On 2-3

Page 38

2. Locate the system status panel.

The locations of the system status panel in the Netra CT 810 server and Netra CT 410 server are illustrated in

FIGURE 2-2.

DD0

A B

O R

HDD0

S E

L A

R M

/U S

E R

E A

D Y

A B

O R

E T

H E

R N

S E

L A

R M

/U S

E R

E A

D Y

i c r o s y s

t e m

COM 2

COM 1 ENET 1 ENET 2

C O

ALARM

H O

T S W

O T

S W

A P

COM 2

COM 1 ENET 1 ENET 2

H O

T S W

B O

E T

ALARM

O T

H E

R N

E T

E S

E T

A L

A R

M /

U S

C O

S W

i c r

o s y

t e m

T H

E R

E T

HDD1

A P

m i c r

o s y s t

e m s

O M

H O

T S

A P

ALARM

COM 2

COM 1 ENET 1 ENET 2

H O

T S W

System status panels,

Netra CT 410 servers

System status panel,

Netra CT 810 server

FIGURE 2-2 System Status Panel Locations

3. Locate the system power button on the system status panel, then press the system power button to power on the server.

FIGURE 2-3 shows the system power button location for the Netra CT 810 server, and FIGURE 2-4 shows the system power button location for the Netra CT 410 server.

2-4 Netra CT Server Service Manual • February 2007

Page 39

System power LED

FIGURE 2-3 System Power Button and System Power LED Locations (Netra CT 810

System power button

Server)

System power LED

System power button

FIGURE 2-4 System Power Button and System Power LED Locations (Netra CT 410

Server)

4. Verify that the system power LEDs on the system status panel is on, indicating that the system is completely powered on.

FIGURE 2-3 shows the system power LED location for the Netra CT 810 server, and FIGURE 2-4 shows the system power LED location for the Netra CT 410 server.

5. Verify that the green power ( ) LED on the power supplies are on, not blinking, indicating that they are powered on and functioning properly.

6. Connect to the console and boot the server.

See Appendix B for instructions on connecting a terminal to your server, if necessary.

Chapter 2 Powering the Server Off and On 2-5

Page 40

2.2 Verifying Full Power-Up

The Netra CT server may take several minutes to completely power up and finish the configuration process, depending on the number of I/O boards installed in the system. Do not remove or install any boards in the Netra CT server until you have verified that the system is completely powered up.

To verify that the system is completely powered up and is fully configured, as superuser, enter:

# prtpicl -c fru -v | more

Output similar to the following is displayed:

chassis (fru, 3d00000008) :ChassisType SUNW,NetraCT-810 :State configured :_class fru :name chassis ...

■ If the entry in the State field is configured, then the system is completely

powered up and configured. You can safely remove and install boards at this point.

■ If the entry in the State field is not configured (for example, configuring or

unconfigured), then the system is still coming up. Do not remove or install any

boards in the Netra CT server in this state. You must wait until the state is

configured.

2-6 Netra CT Server Service Manual • February 2007

Page 41

2.3 Powering Off the Server

You can power off the Netra CT server in three ways:

■ Graceful hardware power-down

■ Ungraceful hardware power-down

■ Graceful software power-down

If you do not have a terminal hookup to your server and your CPU is functioning normally, initiate a graceful hardware power-down for your server. For a graceful hardware power-down, press the system power button for less than four seconds. This will start the orderly power-down sequence in such a manner that no persistent OS data structures are corrupted. In the orderly power-down, applications in service are terminated and no further services are invoked by the CPU.

If you do not have a terminal hookup to your server and your CPU is not functioning normally, you cannot initiate a graceful hardware power-down; you must perform an ungraceful hardware power-down. For an ungraceful hardware power-down, press the system power button for more than four seconds. This immediately shuts the system down without making any attempts to keep the persistent OS data structures from being corrupted.

A graceful software power-down is a normal power-down using a terminal connected to your server, where you log onto the server as superuser, then execute a software command to bring the server down safely.

■ For hardware power-down instructions, go to Section 2.3.1, “To Perform a

Hardware Power-Down” on page 2-7.

■ For software power-down instructions, go to Section 2.3.2, “To Perform a

Software Power Down” on page 2-8.

2.3.1 To Perform a Hardware Power-Down

1. In preparation for bringing down the operating system and powering off the server, ensure that all significant application activity is stopped on the server.

2. Go to the front of the Netra CT server and locate the system status panel.

The locations of the system status panels in the Netra CT 810 server and Netra CT 410 server are illustrated in

3. Locate the system power button on the system status panel.

FIGURE 2-3 shows the system power button location for the Netra CT 810 server, and FIGURE 2-4 shows the system power button location for the Netra CT 410 server.

FIGURE 2-2.

Chapter 2 Powering the Server Off and On 2-7

Page 42

4. Press the system power button and release it to go through a graceful hardware power-down.

This action starts the orderly power-down sequence in such a manner that no persistent OS data structures are corrupted. Applications in service might be terminated and no further services are invoked by the CPU. The system power LED blinks for several seconds, then it goes off.

5. Verify that the green power ( ) LEDs on the power supplies are blinking, indicating that the system is in the standby mode.

FIGURE 2-3 shows the system power LED location for the Netra CT 810 server, and FIGURE 2-4 shows the system power LED location for the Netra CT 410 server.

6. If you want to completely power off the Netra CT server, push the purple power supply unit locking mechanisms up into the unlocked ()position (

FIGURE 2-1).

Note – You must unlock the locking mechanism on both power supply units on the

Netra CT 810 server in order to completely power off that server.

The green power ( ) LEDs on the power supply units should go off (unlit), indicating that the system is now completely powered off.

2.3.2 To Perform a Software Power Down

1. In preparation for bringing down the operating system and powering down the server, ensure that all significant application activity is stopped on the server.

2. Log in as superuser at the system console and enter:

# cd / # shutdown -i0 -g0 -y

When the server is finished shutting down, the ok prompt is displayed.

3. Connect to the console and boot up the server.

See Appendix B for instructions on connecting a terminal to your server, if necessary.

4. At the ok prompt, enter:

ok power-off

2-8 Netra CT Server Service Manual • February 2007

Page 43

5. Go to the front of the Netra CT server and locate the system status panel (see

FIGURE 2-2).

The locations of the system status panel in the Netra CT 810 server and Netra CT 410 server are illustrated in

FIGURE 2-2.

6. Verify that the green power ( ) LEDs on the power supplies are blinking, indicating that the system is in the standby mode.

FIGURE 2-3 shows the system power LED location for the Netra CT 810 server, and FIGURE 2-4 shows the system power LED location for the Netra CT 410 server.

7. If you want to completely power off the Netra CT server, push the purple power supply unit locking mechanisms up into the unlocked ()position (

FIGURE 2-1).

Note – You must unlock the locking mechanism on both power supply units on the

Netra CT 810 server to completely power off the server.

The green power ( ) LEDs on the power supply units turn off (unlit), indicating that the system is now completely powered off.

Chapter 2 Powering the Server Off and On 2-9

Page 44

2-10 Netra CT Server Service Manual • February 2007

Page 45

CHAPTER

Handling Boards and Assemblies

This chapter gives instructions on how to safely handle the CompactPCI boards and assemblies. This chapter is divided into the following sections:

■ Section 3.1, “Handling Boards” on page 3-2

■ Section 3.2, “Handling Assemblies” on page 3-3

Consult the Netra CT Server Safety and Compliance Manual for safety information prior to performing the procedures in this chapter.

3-1

Page 46

3.1 Handling Boards

Each Netra CT server in a chassis has a CompactPCI bus. All of the boards in a server—the CPU board, alarm card, and I/O boards—are CompactPCI boards.

Caution – The system is sensitive to static electricity. To prevent damage, always

connect an antistatic wrist strap between you and the system.

Caution – Do not flex the CompactPCI boards; the surface-mounted components

can break.

To minimize the amount of flexing, observe the following precautions:

■ When removing a board from an electrostatic discharge bag, keep it vertical until

you lay it on the electrostatic discharge mat.

■ Do not place a board on a hard surface. Use a cushioned antistatic mat. The

connectors and components have very thin pins that bend easily.

■ Be careful of small parts located on the component side of a board.

■ Do not use an oscilloscope probe on the components. The soldered pins are easily

damaged or shorted by the probe point.

■ Transport a board in an antistatic bag.

Caution – The heat sinks on a board can be damaged by incorrect handling. Do not

touch the heat sinks while installing or removing a board. Hold a board only by the edges. If a heat sink is loose or broken, obtain a replacement board.

Caution – The heat sinks on a board can be damaged by improper packaging.

When storing or shipping a board, ensure that the heat sinks have sufficient protection.

Caution – The system controller board and its modules have surface-mount

components that can be broken by flexing the board.

3-2 Netra CT Server Service Manual • February 2007

Page 47

3.2 Handling Assemblies

Assemblies have their own set of handling requirements, similar to the requirements for CompactPCI boards.

Caution – The system is sensitive to static electricity. To prevent damage to the

board, always connect an antistatic wrist strap between you and the system.

Chapter 3 Handling Boards and Assemblies 3-3

Page 48

3-4 Netra CT Server Service Manual • February 2007

Page 49

PA R T

2 Troubleshooting the System

Troubleshooting the System Chapter 4

Page 50

Page 51

CHAPTER

Troubleshooting the System

This chapter provides instructions for troubleshooting the Netra CT server. You can troubleshoot the system several ways.

■ Section 4.1, “Troubleshooting the System Using the System Status Panel” on

page 4-2

■ Section 4.2, “Troubleshooting the System Using prtdiag” on page 4-9

■ Section 4.3, “Troubleshooting the System Using Diagnostic Software” on

page 4-14

■ Section 4.4, “Troubleshooting the System Using the Power-On Self Test (POST)”

on page 4-16

■ Section 4.5, “Troubleshooting the System Using the Alarm Card Software” on

page 4-18

■ Section 4.6, “Troubleshooting a Power Supply Using the Power Supply Unit

LEDs” on page 4-18

■ Section 4.7, “Troubleshooting a Host CPU Board” on page 4-20

In addition, Appendix C lists the error messages that might appear when you are operating or servicing a Netra CT server.

4-1

Page 52

4.1 Troubleshooting the System Using the System Status Panel

You can use the system status panel to troubleshoot the Netra CT server.

4.1.1 Locating and Understanding the System Status Panel

The system status panel on the Netra CT server give the majority of troubleshooting information that you need for your server. system status panels on the Netra CT servers. panel for the Netra CT 810 server, and the Netra CT 410 server.

FIGURE 4-1 shows the locations of the

FIGURE 4-2 shows the system status

FIGURE 4-3 shows the system status panel for

System status panels,

Netra CT 410 servers

System status panel,

Netra CT 810 server

FIGURE 4-1 System Status Panel Locations

A B

O R

HDD0

S E

L A

R M

/ U S

E R

E A

D Y

A B

O R

E T

H E

R N

S E

L A

R M

/ U S

E R

E A

D Y

i c r o s y s

t e m

COM 2

COM 1 ENET 1 ENET 2

C O

ALARM

H O

T S W

O T

S W

A P

COM 2

COM 1 ENET 1 ENET 2

H O

T S W

B O

E T

ALARM

O T

H E

R N

E T

E S

E T

A L

A R

M /U

C O

S W

i c r

o s y

t e m

T H

E R

E T

HDD1

A P

m i c r

o s y s t

e m s

O M

H O

T S

A P

ALARM

COM 2

COM 1 ENET 1 ENET 2

H O

T S W

4-2 Netra CT Server Service Manual • February 2007

Page 53

FIGURE 4-2 System Status Panel (Netra CT 810 Server)

FIGURE 4-3 System Status Panel (Netra CT 410 Server)

4.1.2 Using the System Status Panel LEDs to Troubleshoot the System

When you first power-on the Netra CT server, some or all of the green Power LEDs on the system status panel flash on and off for several seconds. Do not attempt to troubleshoot the system until after the LEDs have gone through their initial poweron testing.

Each major component in the Netra CT 810 server and Netra CT 410 server has a set of LEDs on the system status panel that gives the status on that component. Each component has either green Power and amber Okay to Remove LEDs ( green Power and amber Fault LEDs (

Green Power LED

FIGURE 4-4 Power and Okay to Remove LEDs

FIGURE 4-4)or

FIGURE 4-5).

Amber Okay to Remove LED

Chapter 4 Troubleshooting the System 4-3

Page 54

Green Power LED

FIGURE 4-5 Power and Fault LEDs

TABLE 4-1 lists the LEDs for each component in the Netra CT 810 server, and TABLE 4-2 lists the LEDs for each component in the Netra CT 410 server. Note that the

Amber Fault LED

boards in the Netra CT servers all have the green Power LED, and they have either the amber Okay to Remove LED or the amber Fault LED, not both.

TABLE 4-1 System Status Panel LEDs for the Netra CT 810 Server

LED LEDs Available Component

HDD 0 Power and Okay to

Upper hard drive

Remove

HDD 1 Power and Okay to

Lower hard drive

Remove

Slot 1 Power and Okay to

Host CPU board installed in slot 1

Remove

Slots 2 - 7 Power and Okay to

I/O board or satellite CPU board (

●) installed in slot 2 - 7

Remove

Slot 8 Power and Okay to

Alarm card (

■) installed in slot 8

Remove

SCB Power and Fault System controller board (behind the system status panel)

FAN 1 Power and Fault Upper fan tray (behind the system status panel)

FAN 2 Power and Fault Lower fan tray (behind the system status panel)

RMM Power and Okay to

Removable media module

Remove

PDU 1 (DC only) Power and Fault Left most power distribution unit (behind the server)

PDU 2 (DC only) Power and Fault Right most power distribution unit (behind the server)

PSU 1 Power and Okay to

Left most power supply unit

Remove

PSU 2 Power and Okay to

Remove

4-4 Netra CT Server Service Manual • February 2007

Right most power supply unit

Page 55

TABLE 4-2 System Status Panel LEDs for the Netra CT 410 Server

LED LEDs Available Component

Slot 1 Power and Okay to

Alarm card(■) installed in slot 1

Remove

Slot 2 Power and Okay to

I/O board or satellite CPU board (

●) installed in slot 2

Remove

Slot 3 Power and Okay to

Host CPU board installed in slot 3

Remove

Slot 4 and 5 Power and Okay to

Remove

HDD 0 Power and Okay to

I/O boards or satellite CPU boards ( 5

hard drive

●) installed in slot 4 and

Remove

SCB Power and Fault System controller board (behind the system status panel)

FAN 1 Power and Fault Upper fan tray (behind the system status panel)

FAN 2 Power and Fault Lower fan tray (behind the system status panel)

PDU 1 (DC only) Power and Fault Power distribution unit (behind the server)

PSU 1 Power and Okay to

Power supply

Remove

■ TABLE 4-3 lists the LED states and meanings for any CompactPCI board installed

in a slot in the Netra CT 810 server or Netra CT 410 server.

■ TABLE 4-4 lists the LED states and meanings for any component other than a

CompactPCI board that has the green Power and amber Okay to Remove LEDs.

■ TABLE 4-5 lists the LED states and meanings for any component other than a

CompactPCI board that has the green Power and amber Fault LEDs.

Note – Do not use the information in TABLE 4-4 to troubleshoot a power supply unit

in a server that has only one power supply unit (a Netra CT 410 server or a Netra CT 810 server with only one power supply). To troubleshoot the power supply in a single power supply system, use the LEDs on the power supply itself. See

Section 4.6, “Troubleshooting a Power Supply Using the Power Supply Unit LEDs” on page 4-18 for more information. The information given in

TABLE 4-4 applies to all

other components in the Netra CT 810 server or Netra CT 410 server, including the power supplies in a two-power supply Netra CT 810 server.

Chapter 4 Troubleshooting the System 4-5

Page 56

Green Power LED state

TABLE 4-3 CompactPCI Board LED States and Meanings

Amber Okay to Remove LED state Meaning Action

Off Off The slot is empty or the system

thinks that the slot is empty because the system didn’t detect the component when it was inserted.

Blinking Off The component is coming up or

going down.

On Off The component is up and

running.

If there is a component installed in this slot, then one of the following boards is faulty:

• the board installed in the slot

• the alarm card

• the system controller board Remove and replace the failed board to clear this

state.

Do not remove the component in this state.

4-6 Netra CT Server Service Manual • February 2007

Page 57

TABLE 4-3 CompactPCI Board LED States and Meanings

Amber Green Power LED state

Okay to

Remove

LED state Meaning Action

Off On The component is powered off. You can remove the component in this state.

Blinking or Off

On The component is powered on,

but it is offline for some reason (for example, a fault was detected on the component).

On On The component is powered on

and is in use, but a fault has been detected on the component

Wait until the green Power LED stops blinking. If it does not stop blinking after several seconds, enter

cfgadm and verify that the component is in the unconfigured state, then perform the necessary

action, depending on the component:

• Alarm card—When the green Power LED is OFF and the amber

Okay to Remove LED is ON,

you can remove the alarm card.

• All other boards—Power off the slot through the alarm card software, then remove the component.

Deactivate the component using one of the following methods:

• Use the cfgadm -f -c unconfigure command to deactivate the component. Note that in some cases, this may cause the system to panic, depending on the nature of the component’s hardware or software.

• Halt the system and power off the slot through the alarm card software, then remove the component.

The green Power LED will then give status information:

• If the green Power LED goes off, then you can remove the component.

• If the green Power LED remains on, then you must halt the system and power off the slot through the alarm card software.

TABLE 4-4 Meanings of Power and Okay to Remove LEDs

LED State Power LED Okay to Remove LED

On, Solid

Component is installed and configured.

Component is Okay to Remove. You can remove the component from the system, if necessary.

Chapter 4 Troubleshooting the System 4-7

Page 58

TABLE 4-4 Meanings of Power and Okay to Remove LEDs

LED State Power LED Okay to Remove LED

On, Flashing

Component is installed but is

Not applicable. unconfigured or is going through the configuration process.

Off

TABLE 4-5 Meanings of Power and Fault LEDs

LED State Power LED Fault LED

On, Solid

On, Flashing

Component was not recognized by the system or is not installed in the slot.

Component is installed and configured.

Component is installed but is

Component is not Okay to Remove.

Do not remove the component while

the system is running.

Component has failed. Replace the

component.

Not applicable. unconfigured or is going through the configuration process.

Off

Component was not recognized by

Component is functioning properly. the system or is not installed in the slot.

4-8 Netra CT Server Service Manual • February 2007

Page 59

4.2 Troubleshooting the System Using

prtdiag

You can troubleshoot the system using the prtdiag command. Log onto the server console and, as root, enter:

# /usr/platform/sun4u/sbin/prtdiag

If you have a Netra CT 810 server, the output on the console is similar to the following:

CODE EXAMPLE 4-1 prtdiag Output for a Netra CT 810 Server

System Configuration: Sun Microsystems sun4u SPARCengine CP2000 model 140 (UltraSPARC-IIi 648MHz) Memory size: 512 Megabytes platform is : SUNW,NetraCT-810 =============================== FRU Information =============================== FRU FRU FRU Green Amber Miscellaneous Type Unit# Present LED LED Information

---------- ----- ------- ----- ----- -------------------------Midplane 1 Yes Netra ct800

Properties:

Version=0 Maximum Slots=8

SCB 1 Yes on off System Controller Board

Properties:

Version=2

hotswap-mode=basic SSB 1 Yes System Status Panel CPU 1 Yes on off CPU board

temperature(celsius):38

I/O 2 Yes on off CompactPCI IO Slot

Properties:

auto-config=disabled Board Type:Unknown Devices:

pci

pci108e,1000 SUNW,hme SUNW,isptwo

I/O 3 Yes on off CompactPCI IO Slot

Properties:

auto-config=disabled

Chapter 4 Troubleshooting the System 4-9

Page 60

CODE EXAMPLE 4-1 prtdiag Output for a Netra CT 810 Server (Continued)

Board Type:Unknown

Devices:

pci

pci108e,1000 SUNW,hme SUNW,isptwo

I/O 4 Yes on off CompactPCI IO Slot

Properties:

auto-config=disabled Board Type:Unknown

Devices:

pci

pci108e,1000 SUNW,hme SUNW,isptwo

I/O 5 Yes on off CompactPCI IO Slot

Properties:

auto-config=disabled Board Type:Unknown

Devices:

pci

pci108e,1000 SUNW,hme SUNW,isptwo

I/O 6 Yes on off CompactPCI IO Slot

Properties:

auto-config=disabled

I/O 7 Yes on off CompactPCI IO Slot

Properties:

auto-config=disabled Board Type:Unknown Devices:

pci

pci108e,1000 SUNW,qfe pci108e,1000 SUNW,qfe pci108e,1000 SUNW,qfe pci108e,1000 SUNW,qfe pci1176,608

I/O 8 Yes on off CompactPCI IO Slot

Properties:

auto-config=disabled Board Type:Alarm Card Devices:

4-10 Netra CT Server Service Manual • February 2007

Page 61

CODE EXAMPLE 4-1 prtdiag Output for a Netra CT 810 Server (Continued)

pci

ebus

ethernet PDU 1 Yes on off Power Distribution Unit PDU 2 Yes on off Power Distribution Unit PSU 1 Yes on on Power Supply Unit

condition:ok temperature:ok

ps fan:ok

supply:on

PSU 2 Yes on on Power Supply Unit

condition:ok temperature:ok ps fan:ok supply:on

FAN 1 Yes on off Fan Tray

condition:ok fan speed:low

FAN 2 Yes on off Fan Tray

condition:ok fan speed:low

HDD 0 Yes on off hard drive

condition:ok

HDD 1 Yes on off hard drive

condition:ok

RMM Yes on on Removable Media Module

condition:Unknown

System Board PROM revision:

--------------------------OBP 3.14.1 2000/04/28 12:56

Chapter 4 Troubleshooting the System 4-11

Page 62

If you have a Netra CT 410 server, the output on the console is similar to the following:

CODE EXAMPLE 4-2 prtdiag Output for a Netra CT 410 Server

System Configuration: Sun Microsystems sun4u SPARCengine CP2000 model 140 (UltraSPARC-IIi 648MHz) Memory size: 512 Megabytes platform is : SUNW,NetraCT-410 =============================== FRU Information =============================== FRU FRU FRU Green Amber Miscellaneous Type Unit# Present LED LED Information

---------- ----- ------- ----- ----- -------------------------Midplane 1 Yes Netra ct400

Properties:

Version=0 Maximum Slots=5

SCB 1 Yes on off System Controller Board

Properties: Version=2

hotswap-mode=basic SSB 1 Yes System Status Panel I/O 1 Yes on off CompactPCI IO Slot

Properties:

auto-config=disabled

Board Type:Alarm Card Devices:

pci

ebus ethernet

I/O 2 Yes off off CompactPCI IO Slot

Properties:

auto-config=disabled CPU 3 Yes on off CPU board

temperature(celsius):38 I/O 4 Yes on off CompactPCI IO Slot

Properties:

auto-config=disabled

Board Type:Unknown Devices:

pci

pci108e,1000 SUNW,hme SUNW,isptwo

I/O 5 Yes on off CompactPCI IO Slot

Properties:

auto-config=disabled

Board Type:Unknown Devices:

4-12 Netra CT Server Service Manual • February 2007

Page 63

CODE EXAMPLE 4-2 prtdiag Output for a Netra CT 410 Server (Continued)

pci

pci108e,1000 SUNW,qfe pci108e,1000 SUNW,qfe pci108e,1000 SUNW,qfe pci108e,1000

SUNW,qfe PDU 1 Yes on off Power Distribution Unit PSU 1 Yes on off Power Supply Unit

condition:ok temperature:ok ps fan:ok supply:on

FAN 1 Yes on off Fan Tray

condition:ok fan speed:low

FAN 2 Yes on off Fan Tray

condition:ok fan speed:low

HDD 0 Yes on off hard drive

condition:ok

System Board PROM revision:

--------------------------OBP 3.14.1 2000/04/28 12:56

Chapter 4 Troubleshooting the System 4-13

Page 64

4.3 Troubleshooting the System Using Diagnostic Software

Software packages, such as Sun VTS, allow you to run diagnostic tests on your system. SunVTS is a validation test suite that is provided as a supplement to the Solaris operating environment. The individual tests can stress a device, system, or resource so as to detect and pinpoint hardware and software failures and to provide users with informational messages for resolving problems. SunVTS runs at the operating system level.

The following tests are useful when troubleshooting a Netra CT server:

■ alarm2test—is part of SunVTS, but it is used to test the alarm card installed in

the Netra CT server by invoking the alarmdiag test on the alarm card. alarm2test runs at the operating system level.

■ obdiag—is similar to the alarm2test, in that it invokes the alarmdiag test on

the alarm card; however, obdiag is run from the firmware level, not the operating system level.

■ Apost—is part of the embedded firmware image on the alarm card. It runs a

basic test on the alarm card to verify that the alarm card is operating properly before bringing up Chorus on the alarm card.

A new utility called diagconf, which is part of the embedded firmware image on the alarm card, is now available. You can use diagconf to set or display the configuration settings for Apost, allowing you to make the tests run on the alarm card more or less thoroughly before the embedded firmware is brought up on the alarm card.

To display the values currently set for Apost, access the alarm card command line interface (CLI) and enter the following command:

hostname cli> diagconf -d

4-14 Netra CT Server Service Manual • February 2007

Page 65

Output similar to the following is displayed, giving you the values currently set for the Apost test on the alarm card:

diag-switch False verb-mode True stop-on-error False diag-level Max mfg-mode Off hdr-checksum 0xaa time-stamp 0 record-format-ver 49 post-version 02 reset-status 0xd0000000 post-status ... post-msg Watchdog Reset-------- POST Passed-------------------

Some values are hard-set and cannot be changed by a user, while others can be changed to make the test more or less thorough. To change the value for a test, enter the following command:

hostname cli> diagconf -s command value

where command is the name of the command that you want to change, and value is the value you want to change.

The following table lists the Apost tests that can be changed by a user and the allowable values for each. Any tests not listed in

TABLE 4-6 are either hard-set and

cannot be changed, or should not be changed by a user.

TABLE 4-6 Apost Tests and Values through diagconf

Command Value

diag-switch • True—Turns the diag-switch test on.

• False—Turns the diag-switch test off.

verb-mode • True—Turns the verb-mode test on.

• False—Turns the verb-mode test off.

stop-on-error • True—Stops the Apost testing when the first error is encountered.

• False—Continues Apost testing, regardless of the number of errors encountered.

diag-level • Off—Turns the diag-level test off.

• Min—Sets the diag-level test to the minimum level of testing.

• Max—Sets the diag-level test to the maximum level of testing.

Chapter 4 Troubleshooting the System 4-15

Page 66

For more information on these and other tests in the SunVTS test suite, refer to the Computer Systems Release Notes Supplement for Sun Hardware document or the SunVTS documentation on the Solaris on Sun Hardware Answerbook, both included with your Solaris operating environment.

4.4 Troubleshooting the System Using the Power-On Self Test (POST)

When you first power-up the Netra CT server, some or all of the green Power LEDs on the system status panel flash on and off for several seconds. The green Power LED for the I/O slot holding the host board (slot 1 in the Netra CT 810 server and slot 3 in the Netra CT 410 server) lights solid green while the green Power LEDs for the remaining components are flashing on and off; this status is an indication that the CPU board passed the power-on self test (POST).

Before any processing occurs on a system, the system must successfully complete the POST. Messages are displayed for each step in the POST process. If there is a critical failure, the system does not complete POST and does not boot. To monitor this process, you must be connected to the TTY A port on the CPU board or rear transition module. See Section 5.2.1, “Logging In to the Netra CT Server” on

page 5-5.

OpenBoot PROM (OBP) variables control the console port. The variables and their possible settings are described as follows.

To see the console output device, enter:

ok printenv output-device

The screen displays output similar to the following:

output-device ttya

The possible settings for this variable are as follows:

■ ttya (default)

■ ttyb

■ screen

■ rsc

4-16 Netra CT Server Service Manual • February 2007

Page 67

Both ttya and ttyb represent the serial ports on the CPU board. screen represents the display attached to the first frame buffer installed in the system (not present on the Netra CT server). rsc is used by the alarm card.

To see the console input device, enter:

ok printenv input-device

The screen displays output similar to the following:

input-device ttya

The possible settings for this variable are:

■ ttya (default)

■ ttyb

■ keyboard

■ rsc

ttya and ttyb represent the serial ports on the CPU board. keyboard represents the standard system keyboard (not present on the Netra CT server). rsc is used by the alarm card. If no system keyboard is connected, the console port defaults to ttya.

Note – Be sure the two variables are consistent with each other. For example, do not

set the output-device to screen and the input-device to ttya.

Another OBP variable controls the behavior of the POST process called diag- level. By default, this variable is set to max, which means POST runs more thorough (verbose) tests against the hardware. This variable can be set to min, which runs a less stringent set of tests against the hardware. A minimum level of POST testing takes less time, so the Solaris operating environment can boot more quickly on a machine with diag-level set to min.

To run the maximum amount of POST tests, enter:

ok setenv diag-level max

To run the minimum amount of POST tests, enter:

ok setenv diag-level min

Chapter 4 Troubleshooting the System 4-17

Page 68

4.5 Troubleshooting the System Using the Alarm Card Software

For information on troubleshooting using the alarm card software, refer to the Netra CT Server System Administration Guide (819-2743-xx).

4.6 Troubleshooting a Power Supply Using the Power Supply Unit LEDs

Two LEDs are on each power supply unit: a green ( ) LED and an amber ( ) LED. Use the LEDs on the power supply unit to troubleshoot each power supply unit. Because there is one power supply unit in the Netra CT 410 server and two power supply units in the Netra CT 810 server, the actions to take are different. The following sections provide guidelines for each server.

4.6.1 Troubleshooting the Power Supply Unit in the Netra CT 410 Server

Following are the states of the LEDs on the power supply unit in the Netra CT 410 server:

■ Green, flashing—The power supply unit is in the standby mode; the power

supply unit is powered on, yet it is not supplying power to the server.

■ Green, solid—Both the server and the power supply unit are powered on and

functioning properly.

■ Amber—A fault was found in the power supply unit. Replace the power supply

unit. See Section 10.2, “Cold-Swappable Power Supply Unit” on page 10-5 for those instructions.

■ Off—One of the following conditions apply:

■ The power supply locking mechanism is in the upper, unlocked position.

■ The accompanying cable is disconnected from the power distribution unit.

■ The accompanying power distribution unit failed.

■ The power supply unit failed.

4-18 Netra CT Server Service Manual • February 2007

Page 69

4.6.2 Troubleshooting the Power Supply Units in the Netra CT 810 Server

When both power supply units in a Netra CT 810 server are up and running properly, the green ( )LEDs on both power supply units is ON (note that these are the LEDs on the power supply units themselves, not the LEDs on the system status panel).

If a power supply unit fails, the amber ( ) LED on the power supply unit might light, depending on the type of failure that occurs:

■ If a soft-fault occurs, such as a stuck fan or a temperature warning, a notification

of the error is given; however, the amber ( ) LED on the power supply unit does not light for a soft-fault condition. The power supply unit is still supplying power to the system during a soft-fault condition.

■ If a hard-fault occurs, such as a voltage problem, a notification of the error is

given. In addition, the amber ( ) LED on the power supply unit does light for a hard-fault condition. The power supply unit does not supply power to the system during a hard-fault condition.

If one power supply unit fails (either a soft-fault or a hard-fault), but the other power supply unit is still functioning normally, replace the faulty power supply unit as soon as possible. If both power supply units fail, the action to take varies depending on which of the two types of faults occurred:

If Then

Both power supply units go through a soft-fault Replace one power supply unit at a time, in

order to keep the system up and running.

One power supply unit goes through a soft-fault and the other power supply unit goes through a hard-fault

Both power supply units go through a hard-fault The system is down. Replace at least one of

Replace the power supply unit that has a hard-fault first, in order to keep the system up and running.

the power supply units to bring the system back up again.

Chapter 4 Troubleshooting the System 4-19

Page 70

4.7 Troubleshooting a Host CPU Board

This section describes how to troubleshoot problems related to the host board. The information provided here primarily covers situations when the system containing the host board does not boot up or when the board is not fully functional after boot. Only general troubleshooting tips are provided here. No component-level troubleshooting information is included in this section.

The following topics are covered:

■ General troubleshooting tips

■ General troubleshooting requirements

■ Mechanical failures

■ Power-on failures

■ Failures subsequent to power-on

■ Troubleshooting during POST/OBP and during boot process

Also, the following diagnostic procedures are described:

■ OpenBoot PROM on-board diagnostics

■ OpenBoot diagnostics

4.7.1 General Troubleshooting Tips

Caution – High voltages are present in the Netra CT server. To avoid physical

injury, follow all the safety rules specified in the Netra CT Server Safety and Compliance Manual when opening the enclosure and/or removing and installing a

board.

The following general troubleshooting tips are useful in isolating issues related to a host board:

1. Make sure the host board is installed properly in the correct slot in the Netra CT server.

The CPU board is installed in slot 1 in the Netra CT 810 server and in slot 3 in the Netra CT 410 server.

4-20 Netra CT Server Service Manual • February 2007

Page 71

2. Make sure all the necessary cables are attached properly to the host rear transition module.

The following are possible board and rear transition module combinations:

■ Netra CP2140 and Netra CT CPU transition card (CTC, hereafter referred to as

rear transition module)

■ Netra CP2500 and Netra CP2500 RTM-H (rear transition module for host)

FIGURE 4-6 shows the connectors on the CTC. FIGURE 4-7 shows connectors on the

RTM-H.

SCSI connector

TTY A serial port

Ethernet channel A

Ethernet channel B

Cutout for PIM connector

TTY B serial port

FIGURE 4-6 Connectors on the Netra CP2140 Host Rear Transition Module

Chapter 4 Troubleshooting the System 4-21

Page 72

Cut out for PIM connector

ENET 1

10/100/1000 Ethernet RJ45 connectors

ENET 0

COM

Serial ports

USE ONLY

WITH CP2500

ENET 3

ENET 2

10/100/1000 Ethernet RJ45 connectors

FIGURE 4-7 Connectors on the Netra CP2500 Rear Transition Module Host (RTM-H)

Board

HOST RTM

4.7.2 Warning, Critical, and Shutdown Temperatures

The following temperatures apply to the Netra CP2500 board:

■ Warning: 221˚F (105˚C)

■ Critical: 230˚F (110˚C)

■ Shutdown: 239˚F (115˚C)

4-22 Netra CT Server Service Manual • February 2007

Page 73

4.7.3 General Troubleshooting Requirements

The following devices are required to take some of the recommended actions in this section:

■ Network interface

■ TTYA and TTYB connection or an ASCII terminal connection to serial port

■ Parallel port interface

■ Loopback connectors

4.7.4 Mechanical Failures

Symptom

Unable to insert the CPU board into the backplane.

Action

1. Verify that no mechanical and physical obstructions exist in the slot where the CPU board is going to be installed.

2. Make sure no pins on the board connectors or the CompactPCI backplane connectors are bent or damaged.

3. Verify that the front panel screws are seated and not preventing the board from seating properly.

4.7.5 Power-On Failures

This section provides examples of power-on failure symptoms and suggests actions.

● Make sure the CPU board is installed properly.

■ If the CPU board is a host Netra CP2140 or satellite Netra CP2160 and both the

Ready and Alarm LEDs on the CPU board are green, the board is partially functional and capable of running POST (power on self-test). However, if none of the LEDs is green, the board is not functional. In this case, contact your Sun supplier or field service engineer.

Chapter 4 Troubleshooting the System 4-23

Page 74

■ If the CPU board is a host or satellite Netra CP2500 and the Ready LED is green

and the Alarm LEDs is either off or amber, the board is partially functional and capable of running POST (power on self-test). If the LEDs are not green and either off or amber, the board is not functional. In this case, contact your Sun supplier or field service engineer.

4.7.6 Failures Subsequent to Power-On

Symptom

Cannot connect successfully to a TTY serial port; there are no POST messages and unable to send keyboard input.

Action

1. Check the TTY cable for proper setup.

2. If you do not see any output after connecting the TTY terminal to the rear transition module, remove the terminal, then connect it to the COM port of the CPU board and try again.

4.7.7 Troubleshooting During POST/OpenBoot PROM and During Boot Process

This section describes problems encountered while running POST and OBP and during the boot process.

Symptom

POST error message displays:

cannot establish network service

Action

This might be a hardware address problem.

4-24 Netra CT Server Service Manual • February 2007

Page 75

● Add or check the media access control (MAC) address to the server and the IP

address at the server.

Symptom

POST detects Ecache error and a message similar to the following is displayed:

STATUS =FAILED TEST =Memory Addr w/ Ecache SUSPECT=U5201 and U5202 MESSAGE=Mem Addr line compare error addr 00000000.00000000 exp 00000000.00000000 obs 88888888.88888888

Action

This might be a mounting issue with the CPU Mylar film, socket, or heatsink, which could have occurred during transportation or due to severe vibration.

● Contact Sun’s Customer Care Center.

Caution – Any attempt to disassemble or replace the aforementioned devices voids

the warranty.

Chapter 4 Troubleshooting the System 4-25

Page 76

4.7.8 OpenBoot PROM On-Board Diagnostics

Note – For Netra CP2500 boards, pcia-probe-list does not apply.

The following OBP variables are specific to the Netra CT server:

■ pcia-probe-list—Probes the bus that runs the first ethernet port (front

connection) and standard I/O devices (by default: 1, 2)

■ pcib-probe-list—Probes the bus that runs the second ethernet port (rear

connection) (by default: 1, 2, 3)

■ cpci-probe-list—Probes the bus that runs connections to all cPCI slots in the

ct400 or ct800 (by default: 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, a, b, c, d, e, f)

The following section describes the OBP on-board diagnostics. To execute the OBP on-board diagnostics, the system must be at the ok prompt. The OBP on-board diagnostics are listed as follows:

■ watch-clock

■ watch-net and watch-net-all

■ probe-scsi

■ test alias name, device path, -all

4.7.8.1 watch-clock

The watch-clock command reads a register in the TOD chip and displays the result as a seconds counter. During normal operation, the seconds counter repeatedly increments from 0 to 59 until interrupted by pressing any key on the keyboard. The following identifies the watch-clock output message.

ok watch-clock Watching the seconds register of the real time clock chip It should be ticking once a second Type any key to stop 49 ok

4-26 Netra CT Server Service Manual • February 2007

Page 77

4.7.8.2 watch-net and watch-net-all

The watch-net and watch-net-all commands monitor Ethernet packets on the Ethernet interfaces connected to the system. Good packets received by the system are indicated by a period (.). Errors such as the framing error and the cyclic redundancy check (CRC) error are indicated with an X and an associated error description.

CODE EXAMPLE 4-4 identifies the watch-net-all output message.

CODE EXAMPLE 4-3 watch-net Output Message

ok watch-net Hme register test --- succeeded. Internal loopback test -- succeeded. Transceiver check -Using Onboard Transceiver - Link Up. passed Using Onboard Transceiver - Link Up. Looking for Ethernet Packets. . is a Good Packet. X is a Bad Packet. Type any key to stop.

..................................................

................................................................

........................................................

CODE EXAMPLE 4-3 identifies the watch-net output message and

CODE EXAMPLE 4-4 watch-net-all Output Message

ok watch-net-all /pci@1f,0/pci@1,1/network@1,1 Hme register test --- succeeded. Internal loopback test -- succeeded. Transceiver check -- Using Onboard Transceiver - Link Up. passed Using Onboard Transceiver - Link Up. Looking for Ethernet Packets. . is a Good Packet. X is a Bad Packet.

Type any key to stop. ........ ........

........................................................

................................................................

....................................

Chapter 4 Troubleshooting the System 4-27

Page 78

4.7.8.3 probe-scsi

The probe-scsi command transmits an inquiry command to SCSI devices connected to the system unit on-board SCSI interface. If the SCSI device is connected and active, the target address, unit number, device type, and manufacturer name is displayed.

CODE EXAMPLE 4-5 probe-scsi Output Message

CODE EXAMPLE 4-5 identifies the probe-scsi output message.

ok probe-scsi Primary UltraSCSI bus: Target 0 Unit 0 Disk SEAGATE ST32272W 0876 Target 6 Unit 0 Removable Read Only device TOSHIBA CD-ROM XM-6201TA1037 ok

4.7.8.4 test alias name, device path, -all

The test command, combined with a device alias name or device path, enables a device self-test program. If a device has no self-test program, the message: No selftest method for device name is displayed. To enable the self-test program for a device, type the test command followed by the device alias or device path name.

TABLE 4-7 lists test alias name selections, a description of the selection, and

preparation.

TABLE 4-7 Selected OpenBoot PROM On-Board Diagnostic Tests

Type of Test Description Preparation

test screen Tests system video graphics

hardware and monitor.

test floppy Tests diskette drive response to

commands.

test net Performs internal/external

loopback test of the system autoselected Ethernet interface.

4-28 Netra CT Server Service Manual • February 2007

Diag-switch? NVRAM parameter must be true for the test to execute.

A formatted diskette must be inserted into the diskette drive.

An Ethernet cable must be attached to the system and to an Ethernet tap or hub or the external loopback test fails.

Page 79

TABLE 4-7 Selected OpenBoot PROM On-Board Diagnostic Tests (Continued)

Type of Test Description Preparation

test ttya test ttyb

test keyboard

test -all Sequentially test system-

Outputs an alphanumeric test pattern on the system serial ports: ttya, serial port A; ttyb, serial port B.

Executes the keyboard self-test. Four keyboard LEDs should flash

configured devices containing self-test.

4.7.9 OpenBoot Diagnostics

OpenBoot Diagnostics is an interactive tool that tests various hardware and peripheral devices. When obdiag is typed at the ok prompt in OBP, the menu shown in

obdiag performs root-cause failure analysis on the referenced devices by testing internal registers, confirming subsystem integrity, and verifying device functionality. To run obdiag:

CODE EXAMPLE 4-6 is displayed on the screen.

A terminal must be connected to the port being tested to observe the output.

once and a message is displayed: Keyboard Present.

Tests are sequentially executed in device-tree order (viewed with the show-devs command).

Caution – Prior to running obdiag, do not run any other OBP command that might

change the hardware state of the board. After obdiag tests are run, always reset the system to bring it to a known state.

1. At the ok prompt, enter obdiag.

This displays the OBDiag menu as shown in

CODE EXAMPLE 4-6.

Chapter 4 Troubleshooting the System 4-29

Page 80

2. At the OBDiag menu prompt, enter a number from the menu (such as 17 to enable toggle script-debug messages).

CODE EXAMPLE 4-6 OBDiag Menu

0 .... PCI/Cheerio

1 .... EBUS DMA/TCR Registers

2 .... Ethernet

3 .... Ethernet2 <Inactive>

4 .... Parallel Port

5 .... Serial Port C (on optional I/O board) <Inactive>

6 .... Serial Port D (on optional I/O board) <Inactive>

7 .... NVRAM

8 .... Floppy

9 .... Serial port A

10 ... Serial port B 11 ... RAS 12 ... User Flash1 13 ... User Flash2 14 ... All Above 15 ... Quit 16 ... Display this Menu 17 ... Toggle Script-debug 18 ... Enable External Loopback Tests 19 ... Disable External Loopback Tests Enter (0-14 tests, 15 -Quit, 16 -Menu) ===>

You can type the relevant numbers to run all or some of the tests. If an error is detected, an error message is displayed on the screen. For example, if an error is detected while testing the floppy disk drive, a message similar to the following is displayed on the screen:

TEST= floppy_test STATUS= FAILED SUBTEST= floppy_id0_read_test ERRORS= 1 TTF= 66 SPEED= 440 MHz PASSES= 1 MESSAGE= Error: Recalibrate failed. floppy missing, improperly connected, or defective.

Some of the OBDiag menu options are described in detail in the following paragraphs.

4-30 Netra CT Server Service Manual • February 2007

Page 81

4.7.9.1 PCI/PCIO

The PCI/PCIO diagnostic performs the following:

■ vendor_ID_test: Verifies that the PCIO ASIC vendor ID is 108e.

■ device_ID_test: Verifies that the PCIO ASIC device ID is 1000.

■ mixmode_read: Verifies that the PCI configuration space is accessible as half-

word bytes by reading the EBus2 vendor ID address.

■ 2_class_test: Verifies the address class code. Address class codes include

bridge device (0 x B, 0 x 6), other bridge device (0 x A and 0 x 80), and programmable interface (0 x 9 and 0 x 0).

■ status_reg_walk1: Performs walk1 test on status register with mask 0 x 280

(PCIO ASIC is accepting fast back-to-back transactions, DEVSEL timing is 0 x 1).

■ line_size_walk1: Performs tests “a” through “e.”

■ latency_walk1: Performs walk1 test on latency timer.

■ line_walk1: Performs walk1 test on interrupt line.

■ pin_test: Verifies that the interrupt pin is logic-level high (1) after reset.

CODE EXAMPLE 4-7 shows the PCI/PCIO output message.

CODE EXAMPLE 4-7 PCI/PCIO Output Message

Enter (0-14 tests, 15 -Quit, 16 -Menu) ===> 0

TEST= all_pci/PCIO_test SUBTEST= vendor_id_test SUBTEST= device_id_test SUBTEST= mixmode_read SUBTEST= e2_class_test SUBTEST= status_reg_walk1 SUBTEST= line_size_walk1 SUBTEST= latency_walk1 SUBTEST= line_walk1 SUBTEST= pin_test

Enter (0-14 tests, 15 -Quit, 16 -Menu) ===>

Chapter 4 Troubleshooting the System 4-31

Page 82

4.7.9.2 EBus DMA/TCR Registers

The diagnostic EBus DMA/TCR registers performs the following:

■ The dma_reg_test: Performs a walk1 bit test for control status register, address

■ The dma_func_test: Validates the direct memory access (DMA) capabilities and

first in, first out (FIFO). The test is executed in a DMA diagnostic loopback mode. It initializes the data of transmitting memory with its address, performs a DMA read and write, and verifies that the data received is correct. This diagnostic is repeated for four channels.

CODE EXAMPLE 4-8 shows the EBus DMA/TCR registers output message.

CODE EXAMPLE 4-8 EBus DMA/TCR Registers Output Message

Enter (0-14 tests, 15 -Quit, 16 -Menu) ===> 1

TEST= all_dma/ebus_test SUBTEST= dma_reg_test SUBTEST= dma_func_test

Enter (0-14 tests, 15 -Quit, 16 -Menu) ===>

4.7.9.3 Ethernet

The Ethernet diagnostic performs the following:

■ my_channel_reset: Resets the Ethernet channel.

■ hme_reg_test: Performs walk1 on the following registers set: global register 1,

global register 2, bmac xif register, bmac tx register, and the mif register.

■ MAC_internal_loopback_test: Performs Ethernet channel engine internal

loopback.

■ 10_mb_xcvr_loopback_test: Enables the 10Base-T data present at the transmit

MII data inputs to be routed back to the receive MII data outputs.

■ 100_mb_phy_loopback_test: Enables MII transmit data to be routed to the MII

receive data path.

■ 100_mb_twister_loopback_test: Forces the twisted-pair transceiver into

loopback mode.

4-32 Netra CT Server Service Manual • February 2007

Page 83

CODE EXAMPLE 4-9 shows the Ethernet output message.

CODE EXAMPLE 4-9 Ethernet Output Message

Enter (0-14 tests, 15 -Quit, 16 -Menu) ===> 2

TEST= ethernet_test SUBTEST= my_channel_reset SUBTEST= hme_reg_test SUBTEST= global_reg1_test SUBTEST= global_reg2_test SUBTEST= bmac_xif_reg_test SUBTEST= bmac_tx_reg_test SUBTEST= mif_reg_test Test only supported for National Phy DP83840A SUBTEST= 10mb_xcvr_loopback_test selecting internal transceiver Test only supported for National Phy DP83840A SUBTEST= 100mb_phy_loopback_test selecting internal transceiver Test only supported for National Phy DP83840A SUBTEST= 100mb_twister_loopback_test selecting internal transceiver Test only supported for National Phy DP83840A

Enter (0-14 tests, 15 -Quit, 16 -Menu) ===>

4.7.9.4 Parallel Port

The parallel port diagnostic performs the dma_read. This enables the enhanced capability port (ECP) mode configuration, ECP DMA configuration, and FIFO test mode. It transfers 16 bytes of data from the memory to the parallel-port device, then verifies that the data is in FIFO test. output message.

CODE EXAMPLE 4-10 Parallel Port Output Message

Enter (0-14 tests, 15 -Quit, 16 -Menu) ===> 4

TEST= parallel_port_test SUBTEST= dma_read

Enter (0-14 tests, 15 -Quit, 16 -Menu) ===>

CODE EXAMPLE 4-10 shows the parallel-port

Chapter 4 Troubleshooting the System 4-33

Page 84

Note – The Netra CP2500 board has no parallel port.

4.7.9.5 Serial Port A

The serial port A diagnostic invokes the uart_loopback test. This test transmits and receives 128 characters and checks the transaction validity. identifies the serial port A output message.

CODE EXAMPLE 4-11 Serial Port A Output Message

Enter (0-14 tests, 15 -Quit, 16 -Menu) ===> 9

TEST= uarta_test

Enter (0-14 tests, 15 -Quit, 16 -Menu) ===>

Note – The serial port A diagnostic stalls if the TIP line is installed on serial port A.

CODE EXAMPLE 4-12 identifies the serial port A output message when the TIP line is

installed on serial port A.

CODE EXAMPLE 4-11

CODE EXAMPLE 4-12 Serial Port A Output Message With TIP Line Installed

Enter (0-14 tests, 15 -Quit, 16 -Menu) ===> 9

TEST= uarta_test UART A in use as console - Test not run.

Enter (0-14 tests, 15 -Quit, 16 -Menu) ===>

4.7.9.6 Serial Port B

The serial port B diagnostic is identical to the serial port A diagnostic.

CODE EXAMPLE 4-13 identifies the serial port B output message.

4-34 Netra CT Server Service Manual • February 2007

Page 85

Note – The serial port B diagnostic stalls if the TIP line is installed on serial port B.

CODE EXAMPLE 4-13 Serial Port B Output Message

Enter (0-14 tests, 15 -Quit, 16 -Menu) ===> 10

TEST= uartb_test

Enter (0-14 tests, 15 -Quit, 16 -Menu) ===>

4.7.9.7 NVRAM

The NVRAM diagnostic verifies the NVRAM operation by performing a write and read to the NVRAM.

CODE EXAMPLE 4-14 NVRAM Output Message

Enter (0-14 tests, 15 -Quit, 16 -Menu) ===> 7

TEST= nvram_test SUBTEST= write/read_patterns SUBTEST= write/read_inverted_patterns

CODE EXAMPLE 4-14 identifies the NVRAM output message.

Enter (0-14 tests, 15 -Quit, 16 -Menu) ===>

Chapter 4 Troubleshooting the System 4-35

Page 86

4.7.9.8 All Above

The All Above diagnostic validates the system unit. CODE EXAMPLE 4-15 shows an example of the All Above output message.

CODE EXAMPLE 4-15 All Above Output Message

Enter (0-14 tests, 15 -Quit, 16 -Menu) ===> 14

TEST= all_pci/cheerio_test SUBTEST= vendor_id_test SUBTEST= device_id_test ... SUBTEST= bmac_xif_reg_test SUBTEST= bmac_tx_reg_test SUBTEST= mif_reg_test SUBTEST= mac_internal_loopback_test selecting internal transceiver Test only supported for National Phy DP83840A ... SUBTEST= 100mb_twister_loopback_test selecting internal transceiver Test only supported for National Phy DP83840A TEST= ethernet2_test TEST= parallel_port_test SUBTEST= dma_read TEST= uarta_test ... SUBTEST= write/read_patterns ... ttya in use as console - Test not run. TEST= usi_test ttyb in use as console - Test not run. TEST= ras_test env-monitor = disabled SUBTEST= obd-init-i2c-test ... TEST= flash_test SUBTEST= flash-supported? TEST= flash_test SUBTEST= flash-supported?

Enter (0-14 tests, 15 -Quit, 16 -Menu) ===>

4-36 Netra CT Server Service Manual • February 2007

Page 87

PA R T

3 Replacing Hot-Swappable FRUs

Hot-Swap Software Commands Chapter 5

Removing and Replacing Hot-Swappable Boards Chapter 6

Removing and Replacing Hard Drives and Removeable Media Chapter 7

Removing and Replacing Hot-Swappable Subassemblies Chapter 8

Page 88

Page 89

CHAPTER

Hot-Swap Software Commands

This chapter provides information about hot-swap and the cfgadm utility. This chapter contains the following topics:

■ Section 5.1, “Understanding Hot-Swap” on page 5-2

■ Section 5.2, “Using the cfgadm Utility” on page 5-5

5-1

Page 90

5.1 Understanding Hot-Swap

Certain FRUs in the Netra CT server are hot-swappable (see Section 1.3.1, “Hot-

Swappable FRUs” on page 1-4). Hot-swap, a key feature of the PICMG standard,

means that a CompactPCI board that meets the PICMG standard can be reliably inserted into or extracted from a powered and operating CompactPCI platform without affecting the other functions of the platform.

The Netra CT system hot-swap modes are listed in

TABLE 5-1 Netra CT System Hot-Swap Modes

Type of Hot-Swap Description

Basic The hardware connection/disconnection process is performed

automatically by the hardware, while the software connection process requires user assistance through the cfgadm (1M) command.

Full Both the hardware and the software connection process are

performed automatically.

High Availability High-availability hot-swap provides the ability to control the

hardware connection process. This provides a higher degree of control than just indicating insertion and extraction of a board. The hardware connection process is controlled by software on highavailability systems, such as the Netra CT server.

TABLE 5-1.

The Netra CT system is configured for full hot-swap by default. You can change the mode of the slot for the CPU boards and I/O boards to basic or full hot-swap using the cfgadm(1M) command. You might want to change the hot-swap state of a slot to basic, for example, if you need to insert or remove a third-party I/O board that does not have full hot-swap support.

Note that whenever you reboot or power your server on and off, hot-swap reverts back to the default full hot-swap state for all I/O slots.

5-2 Netra CT Server Service Manual • February 2007

Page 91

5.1.1 How High-Availability Hot-Swap Works

By default, a Netra CT server is configured to accept any cPCI FRU unless you specifically set an allowable plug-in for a slot (refer to the Netra CT Server System Administration Guide for more information.)

When a board is inserted into a Netra CT server, the alarm card checks the midplane FRU ID information for allowable FRUs for that slot, then checks the inserted board’s FRU ID to make sure the board is allowed in the particular slot. If the board is allowed in the slot, the alarm card powers up the board. If the board is not allowed in the slot, the alarm card sends an SNMP trap to the MOH application.

If a host or satellite board is in use (has applications currently running), the alarm card CLI power commands, such as poweron or poweroff, do not work for that board.

5.1.2 Hot-swap with Boards That Do Not Have Full Hot-swap

You might want to change the hot-swap state of a slot from full to basic if you need to insert or remove a third-party I/O board that does not have full hot-swap support.

To determine the current hot-swap state of a slot, use the prtconf(1M) command. To enable or disable the hot-swap on a slot, use the cfgadm (1M) command. For many cfgadm commands, you must know the attachment-point ID for the I/O slot that you are changing the state.

5.1.3 System Status Panel LED States and Meanings

The LEDs on the system status panel give you information to determine what state a board is in and if it is safe to remove the board from the system. See more information.

TABLE 5-1 for

Chapter 5 Hot-Swap Software Commands 5-3

Page 92

Green Power LED state

TABLE 5-2 CompactPCI Board LED States and Meanings on the System Status Panel

Amber Okay to Remove LED state Meaning Action

Off Off The slot is empty or the system

thinks that the slot is empty because the system didn’t detect the board when it was inserted.

If there is a board installed in this slot, then one of the following is faulty:

• the board installed in the slot

• the alarm card

• the system controller board Remove and replace the failed component to clear

this state.

Blinking Off The board is coming up or going

Do not remove the board in this state.

down.

On Off The board is up and running. Do not remove the board in this state.

Off On The board is powered off. You can remove the board in this state.

Blinking On The board is powered on, but it

is offline for some reason (for example, a fault was detected on the board).

Wait several seconds to see if the green Power LED stops blinking. If it does not stop blinking after several seconds, enter cfgadm and verify that the board is in the unconfigured state, then perform the necessary action, depending on the board:

• Alarm card—You can remove the alarm card in this state.

• All other boards—Power off the slot through the alarm card software, then remove the board.

On On The board is powered on and is

in use, but a fault has been detected on the board.

Deactivate the board using one of the following methods:

• Use the cfgadm -f -c unconfigure command to deactivate the board. In some cases, this might cause the system to panic, depending on the nature of the board hardware or software.

• Halt the system and power off the slot through the alarm card software, then remove the board.

The green Power LED gives status information:

• If the green Power LED goes off, remove the board.

• If the green Power LED remains on, halt the system and power off the slot through the alarm card software.

5-4 Netra CT Server Service Manual • February 2007

Page 93

5.2 Using the cfgadm Utility

You can perform the following hot-swap procedures using the cfgadm utility:

■ Deactivate and activate hot-swappable FRUs

■ Enable and disable full hot-swap for I/O slots in a server

■ Enable and disable basic hot-swap for I/O slots in a server

5.2.1 Logging In to the Netra CT Server

To use the cfgadm utility, you must be able to log in to the server either remotely, where you would log in to the Netra CT server as superuser through another server on the network, or directly, where you would connect a terminal console directly to your Netra CT server.

For more information on connecting a terminal console to your system, see

Appendix B.

5.2.2 Running the cfgadm Utility

For all cfgadm commands, you must know the attachment-point ID for the I/O slot that you are changing state. To list the attachment-point IDs for the slots in a server, log in to the server and, as superuser, enter the following command:

# cfgadm

For a Netra CT 810 server, output similar to the following is displayed:

Ap_Id Type Receptacle Occupant Condition AL-8 mcd/fhs connected configured ok CPU bridge/fhs connected configured ok IO-2 stpcipci/fhs connected configured ok IO-3 unknown empty unconfigured unknown IO-4 stpcipci/fhs connected configured ok IO-5 unknown disconnected unconfigured ok IO-6 unknown empty unconfigured unknown IO-7 unknown empty unconfigured unknown

Chapter 5 Hot-Swap Software Commands 5-5

Page 94

For a Netra CT 410 server, output similar to the following is displayed:

Ap_Id Type Receptacle Occupant Condition AL-1 mcd/fhs connected configured ok CPU bridge/fhs connected configured ok IO-2 unknown disconnected unconfigured unknown IO-4 stpcipci/fhs connected configured ok IO-5 stpcipci/fhs connected configured ok

The attachment-point ID is shown in the first column of the readout; for example, the attachment-point ID for I/O slot 2 in a Netra CT 810 server is

IO-2.

To view the online help for cfgadm, as superuser, enter:

# cfgadm -h pci

Output similar to the following is displayed:

PCI hotplug specific commands:

-x enable_slot ap_id [ap_id...]

-x disable_slot ap_id [ap_id...]

-x enable_autoconfig ap_id [ap_id...]

-x disable_autoconfig ap_id [ap_id...]

To view the man page for the cfgadm utility, enter man cfgadm at the prompt.

5.2.3 Basic and Full Hot-Swap cfgadm Commands

The Netra CT servers are set to full hot-swap by default. Full hot-swap on an I/O slot means that you do not need to manually deactivate or activate the I/O board when replacing it in the server; those processes are handled automatically.

■ To determine the current hot-swap state for all of the I/O slots in your Netra CT

server, go to “Determining the Current Hot-Swap State” on page 7.

■ To change default settings for one or more I/O slots in your Netra CT server to

basic hot-swap, go to “Enabling Basic Hot-Swap on I/O Slots” on page 8.

■ To change settings for one or more I/O slots in your Netra CT server back to full

hot-swap, go to “Enabling Full Hot-Swap on I/O Slots” on page 9.

5-6 Netra CT Server Service Manual • February 2007

Page 95

5.2.3.1 Determining the Current Hot-Swap State

To determine the current hot-swap state for the I/O slots, as superuser, enter:

# prtconf -v -P

For a Netra CT 810 server, output similar to the following is displayed:

cphsc, instance #0 System properties: name=’instance’ type=int items=1 value=00000000 name=’default-hotswap-mode’ type=string items=1 value=’full’ Driver properties:

name=’AL-8-autoconfig’ type=string items=1 dev=none

value=’enabled’

name=’IO-7-autoconfig’ type=string items=1 dev=none

value=’enabled’

name=’IO-6-autoconfig’ type=string items=1 dev=none

value=’enabled’

name=’IO-5-autoconfig’ type=string items=1 dev=none

value=’enabled’

name=’IO-4-autoconfig’ type=string items=1 dev=none

value=’enabled’

name=’IO-3-autoconfig’ type=string items=1 dev=none

value=’enabled’

name=’IO-2-autoconfig’ type=string items=1 dev=none

value=’enabled’

name=’CPU-autoconfig’ type=string items=1 dev=none value=’enabled’ name=’hotswap-mode’ type=string items=1 dev=none value=’full’

■ If you see value ’basic’ underneath the default-hotswap-mode line, then

all of the I/O slots in the Netra CT server are set to basic hot-swap. The entry

value ’disabled’ is shown for every I/O slot in the system in this state.

Chapter 5 Hot-Swap Software Commands 5-7

Page 96

■ If you see value ’full’ underneath the default-hotswap-mode line, then at

least one of the I/O slots in the Netra CT server is set to full hot-swap. You must

look at the entries for individual I/O slots to determine if they are set to basic or full hot-swap mode:

■ If you see value ’enabled’ underneath a <slot#-autoconfig> line, then

that slot is set to full hot-swap.

■ If you see value ’disabled’ underneath a <slot#-autoconfig> line, then

that slot is set to basic hot-swap.

5.2.3.2 Enabling Basic Hot-Swap on I/O Slots

All of the I/O slots in the Netra CT server are set to full hot-swap by default. You can disable full hot-swap on one or more I/O slots, bringing those slots to a basic hot-swap state. That means that if an I/O board becomes faulty and needs replacing, you must manually deactivate the I/O slot using the cfgadm utility before you can remove the board, then you must manually reactivate the I/O slot after replacing the board.

To disable full hot-swap on one or more I/O slots, changing those slots to a basic hot-swap state, as superuser, by enter:

# cfgadm -x disable_autoconfig ap_id

Note – Whenever you reboot or power your server on and off, the hot-swap states

revert back to the default full hot-swap state for all I/O slots. If you want one or more I/O slots set to the basic hot-swap setting, you must manually reset the I/O slots after rebooting or powering your server on and off.

Deactivating a Hot-Swappable FRU

To deactivate an I/O board in a basic hot-swap state, as superuser, enter:

# cfgadm -c unconfigure ap_id

where ap_id is the attachment-point ID. For example, to deactivate an I/O board in slot 4, as superuser, enter:

# cfgadm -c unconfigure IO-4

5-8 Netra CT Server Service Manual • February 2007

Page 97

Reactivating a Hot-Swappable FRU

To connect an I/O board, as superuser, enter:

# cfgadm -c connect ap_id

where ap_id is the attachment-point ID.

To activate an I/O board, as superuser, enter:

# cfgadm -c configure ap_id

where ap_id is the attachment-point ID.

5.2.3.3 Enabling Full Hot-Swap on I/O Slots

If you changed the hot-swap states for one or more I/O slots from the default full hot-swap setting to the basic hot-swap setting, use the cfgadm utility to change the I/O slots back to the default full hot-swap setting.

To enable full hot-swap for an I/O slot in a Netra CT server, as superuser, enter:

# cfgadm -x enable_autoconfig ap_id

where ap_id is the attachment-point ID of the slot that you want to have full hotswap enabled on. For example, to enable full hot-swap for I/O slot 4, as superuser, you could enter the following:

# cfgadm -x enable_autoconfig IO-4

To enable full hot-swap on all slots at once, reboot the server, which automatically reverts settings back to the default full hot-swap for all I/O slots.

Chapter 5 Hot-Swap Software Commands 5-9

Page 98

5-10 Netra CT Server Service Manual • February 2007

Page 99

CHAPTER

Removing and Replacing HotSwappable Boards

This chapter provides instructions for removing and replacing hot-swappable boards. This chapter contains the following topics:

■ Section 6.1, “Boards” on page 6-2

■ Section 6.2, “Rear Transition Modules” on page 6-24

Caution – If you are just powering on your Netra CT server, do not remove or

install any boards in the system until you have verified that the system is completely powered up. See Section 2.2, “Verifying Full Power-Up” on page 2-6 for more information.

Caution – A thermal hazard is present if any I/O slots are left uncovered. If you

remove a board from an I/O slot, you must fill it with either a replacement board or a filler panel.

Note – Consult the Netra CT Server Safety and Compliance Manual and Chapter 3 of

this document for safety and board-handling information before performing the procedures in this chapter.

6-1

Page 100

6.1 Boards

This section contains procedures for the installation, removal, and replacement of the following hot-swappable boards for the Netra CT 410 server and Netra CT 810 server:

■ Host board

■ Satellite board

■ Alarm card

■ I/O board

FIGURE 6-1 shows the locations of these boards in the Netra CT 810 server. FIGURE 6-2

shows the locations of these boards in the Netra CT 410 server.

1. The host board ishot-swappable onlyif the alarm card andsystem controllerboard arealso installedin the Netra CTserver.

2. The alarm card ishot-swappable onlyif the host board andsystem controllerboard arealso installedin the Netra CTserver.

6-2 Netra CT Server Service Manual • February 2007