Convex DKD-505, Elite 3 Service Manual

CONVEX

Computer

Corporation

3000 Waterview

Parkway

P.O.

Box

833851

Richardson,

TX

75083-3851

United States

of

America

(214)497-4000

CONVEX

Elite

3

Disc

Drive

Service Guide

Order

No. DHW-261

First Edition

April 1993

CONVEX Press Richardson, Texas United States of America

CONVEX

Elite 3 Disc

Drive

Service

Guide

Order

No.

DHW-261

1993

CONVEX Computer Corporation

may

not,

in

whole

or

part, be copied, duplicated, reproduced, translated,

electronically stored,

or

reduced to machine readable form

without prior written consent from

CONVEX

Computer

Corporation. Although the material contained herein has been carefully

reviewed, CONVEX Computer Corporation does

not

warrant

it to

be free of errors or omissions. CONVEX reserves the right to

make

corrections, updates, revisions

or

changes to the information

contained herein. CONVEX does

not

warrant the material

described herein to

be free of patent infringement.

UNLESS

PROVIDED

OTHERWISE

IN

WRITING

WITH

CONVEX

COMPUTER

CORPORATION

(CONVEX),

THE

PROGRAM

DESCRIBED

HEREIN

IS

PROVIDED

AS

IS

WITHOUT

WARRANTY

OF

ANY

KIND,

EITHER

EXPRESSED

OR

IMPLIED,

INCLUDING,

BUT

NOT

LIMITED

TO

THE

IMPLIED

WARRANTIES

OF

MERCHANTABILITY

AND

FITNESS

FOR A PARTICULAR

PURPOSE.

SOME

STATES

DO

NOT

ALLOW

THE

EXCLUSION

OF

IMPLIED

WARRANTIES.

THE

ABOVE

EXCLUSION

MAY

NOT

BE

APPLICABLE

TO

ALL

PURCHASERS

BECAUSE

WARRANTY

RIGHTS

CAN

VARY

FROM

STATE

TO

STATE.

IN

NO

EVENT

WILL

CONVEX

BE

LIABLE

TO

ANYONE

FOR

SPECIAL,

COLLATERAL,

INCIDENTAL

OR

CONSEQUENTIAL

DAMAGES,

INCLUDING

ANY

LOST

PROFl'IS

OR

LOST

SAVINGS,

ARISING

OUT

OF

THE

USE

OR

INABILITY

TO

USE

THIS

PROGRAM.

CONVEX

WILL

NOT

BE

LIABLE

EVEN

IF

IT

HAS

BEEN

NOTIFIED

OF

THE

POSSIBILITY

OF

SUCH

DAMAGE

BY

THE

PURCHASER

OR

ANY

THIRD

PARTY.

CONVEX and the CONVEX logo ("C") are registered trademarks of CONVEX

Computer Corporation. UNIX

is a registered trademark of UNIX System Laboratories, Inc.

Seagate, Seagate Technology,

and

the Seagate logo are registered trademarks

of

Seagate

Technology, Inc. Computer Products/Power Conversion America

is a registered

trademark

of

Computer

Products, Inc.

()

This book is recyclable excluding cover

and

binding.

Printed in the United States of America

Revision

information

for

CONVEX

Elite 3 Disc

Drive

Service

Guide

Edition Document No. Description

First

081-010230-000 Released April

1993.

Contents

l'r4tfCIC:ft

••••••••••••••••••••••••••••••••••••••••••••••••••••••••

lci

Purpose and audience ................................................................ xi

Using this guide .......................................................................... xi

Notational conventions .............................................................. xi

Notes

and

cautions .................................................................... xii

Associated documents .............................................................. xii

Ordering documents ................................................................. xii

Technical assistance .................................................................. xiii

FCC notice .................................................................................. xiii

Electrostatic discharge protection .......................................... xiii

1

Description

and specifications ............... 1

Drive specifications ..................................................................... 1

de power requirements ......................................................... 3

Front panel indicators and switches ................................... 4

Control board jumpers .......................................................... 6

I/0

board jumpers ................................................................. 8

Power supply specifications ....................................................... 9

Elite 3 chassis ...............................................................................

11

2 Unpacking and

installation

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

13

Unpacking and inspection ....................................................... 13

Inspection .............................................................................. 13

Unpacking .............................................................................

13

Damage claims .....................................................................

14

Preparations .......................................................................... 14

Installing power strip

and

disc tray ..................................

14

Installation ..................................................................................

14

Installing the Elite 3 chassis ................................................ 16

Connecting a single drive ................................................... 17

Connecting multiple disc drives ........................................

20

Physical configuration ......................................................... 22

3 Integration and

testing

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

27

Software integration .................................................................. 27

General integration procedure ........................................... 27

Contents v

I ioconfig file ..........................................................................

28

I

etc/

disktab .......................................................................... 30

Testing the Elite 3 disc drive

with

idcfmt

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

31

Verifying format ...................................................................

31

Formatting a drive ...............................................................

32

4 Maintenance

and

IPB

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

33

Troubleshooting ..........................................................................

33

Elite 3 disc drive error codes ...............................................

33

Fault symptom code

(FSC)

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

35

Sector errors ..........................................................................

44

FSC

codes reported

during

autoconf

............................. 45

WM

messages ..................................................................... 45

Determining disc usage ....................................................... 46

Removal

and

replacement procedures ................................... 48

Elite 3 chassis ........................................................................ 48

Disc drive ............................................................................... 48

Power supply ........................................................................ 49

Operator panel ...................................................................... 49

Fan assembly .........................................................................

50

Air filter .................................................................................

51

Illustrated parts breakdown

(IPB)

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

52

A idcfmt ( 1

D)

man

page

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

55

vi

Elite 3 Disc

Drive

Service

Guide

Figures

Figure 1 Figure2 Figure 3 Figure4 Figures Figure6 Figure 7 Figures Figure9 Figure 10

Figure

11

Figure

12

Figure

13

Figure

14

Figure

15 Figure 16 Figure 17

Figure

18

Figure

19

Figure20 Figure21 Figure22 Figure23 Figure24

Front panel ...................................................................... 4

Control board jumper location..................................... 6

J/O

board jumper locations .......................................... 8

Power

supply

component location ............................ 10

Elite 3 chassis ............. ...................................... .............

12

Additional power strip location.................................

15

Elite 3 disc tray.............................................................. 16

Connecting a single drive............................................ 17

JDC bulkhead

port

assignments for C3200/C3400.

18 JDC

bulkhead port assignments for C3800 ...............

19

Connecting multiple drives........................................

21 JDC

maximum drive configuration...........................

23

Expansion cabinet drive locations for 32 drives......

24

Expansion cabinet drive locations for 16 drives......

25

/ioconfig example 1 .....................................................

29

/ioconfig example 2 .....................................................

29

Example I

etc/

disktab ................................................. 30

WM

device failure message ...................................... 45

WM

completion message .......................................... 46

Determining disc usage

with

df

I

grep..................

46

Determining disc usage from

/etc/fstab

.................. 47

Elite 3 operator panel cable routing...........................

50

Air filter access..............................................................

51

Illustrated parts breakdown (IPB) .............................

52

Figures

vii

Tables

Table 1 Table2 Table3 Table4 Tables Table6 Table 7 Table8 Table9 Table

10

Table

11

Table

12

Table

13

Drive specifications ......... ........................ ........................ 1

Elite 3 disc drive de

power

requirements..................... 3

Elite 3 front panel indicators

and

switches .................. 5

Control board jumper block

J12

assignments ....... 7

1/0

board jumper block

assignments .................... 8

Power supply characteristics ...............

..

........................ 9

Power supply voltage specifications...........................

10

Power supply

chart ..................................................

11

I

etc/

disktab description ............

..

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

30

Elite 3 disc drive error codes........................................

34

FSC

field descriptions

..

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

35

FSC

descriptions ............................ ....... .................... .....

36

IPB

parts list ....................................................................

53

Tables

ix

Preface

Purpose

and

audience

Notational

conventions

The CONVEX

Elite 3 Disc

Drive

Service

Guide

provides a general overview of the Seagate Elite 3 disc drive and related equipment. This guide describes how

to:

• Install the Elite 3 disc drive

and

related equipment

• Integrate the Elite 3 disc drive into the CONVEX operating system

(ConvexOS)

This document is intended for:

• CONVEX customer support engineers

and

CONVEX

manufacturing personnel

• CONVEX customers who install

and

maintain their

own

Elite 3 disc drives and related equipment

This section discusses notational conventions used

in

this book.

Bold

mono

space

In command examples, text shown

in

bold

monospace

identifies user

input

that

must

be typed exactly as shown.

Mono

space

Italic

In paragraph text,

monospace

identifies:

• Command names

• System calls

• Data structures

and

types

In command examples,

monospace

identifies command output, including error messages.

In command syntax diagrams, text

shown

in

monospace

must

be

typed exactly as

shown. In paragraph text,

italic

identifies:

Preface

xi

Notes

and

cautions

Note

Caution

Associated documents

Ordering documents

•

New

and

important terms

• Titles of documents

In command syntax diagrams,

italic

identifies variables

that

must

be

supplied

by

the user.

This document presents notes

and

cautions in the following

formats:

A Note highlights supplemental information.

A Caution

highlights Information necessary to

avoid

damage

to

the system.

For more information

on

the ConvexOS operating system,

you

can order the following books from CONVEX

Computer

Corporation:

•

ConvexOS

Primer

(DSW-133). This book introduces

new

users to the ConvexOS operating system.

•

ConvexOS

Programmer's

Reference

(DSW-332). This book is

the standard reference for the

ConvexOS operating system.

•

Managing

ConvexOS

Operations

Guide

(DSW-031). This book

is the standard reference for system operation.

•

Managing

ConvexOS

Configuration

Guide

(DSW-030). This

book describes the management of system resources.

To

order the current edition of this

or

any

other

CONVEX

document, send requests to: CONVEX Computer Corporation

Customer Service P.O.

Box

833851

Richardson

TX

75083-3851

USA

Please

include the order number (DSW

or

DHW number)

or

the

exact title of the document.

xii

Elite 3 Disc

Drive

Service

Guide

Technical assistance

FCC

notice

Electrostatic discharge protection

Caution

If

you

have questions that are

not

answered

in

this book, contact

the CONVEX Technical Assistance Center (TAC)

at

the

following locations:

• Within the continental U.S., call 1 (800) 952-0379.

•

From Canada, call 1(800)345-2384.

• From all other locations, contact the local CONVEX office.

You

may

also use the

contact

utility to report

any

problems

with

ConvexOS

or

its associated documentation. For more

information, refer to the contact(l)

man

page

in

ConvexOS

Man

Pages

for

Users,

or

the appendix ''Reporting problems"

in

the

ConvexOS

·Primer

or

Managing

ConvexOS:

Operations

Guide.

This equipment generates, uses

and

can

radiate radio frequency

energy.

If

the equipment is not installed

and

used

in

strict

accordance with the instruction manual,

it

may

cause

interference to radio communications. This equipment has been tested

and

found to comply

with

limits

for a Class A digital device,

pursuant

to

Part

15

of

the

FCC Rules. These limits are designed to provide reasonable protection against harmful interference

when

equipment

is

operated

in

a

commercial environment. When this equipment is operated in a residential area, it is likely

to cause interference. In this case, the interference

must

be

corrected at the operator's expense. Do not connect external equipment to the utility outlets

in CONVEX equipment cabinets. Unauthorized connection voids all agencies' emissions certification.

The Elite 3

and

related assemblies are sensitive to static electricity. Although some devices such as metal-oxide semiconductors are extremely sensitive, all semiconductors, as well as some resistors

and

capacitors,

may

be

damaged

or

degraded

by

exposure to static electricity.

Electrostatic damage to electronic devices

may

be caused

by

the

direct discharge of a charged conductor

or

by

exposure to the

static fields surrounding charged objects.

To

avoid

electrostaHc

damage,

service

personnel

must

observe

the

following

precautions

when

servicing

equipment:

Preface

xiii

xiv

Elite 3 Disc

Drive

Service

Guide

• Ground yourself to the peripheral cabinet

or a grounded

service area whenever working

on

the Elite 3

or

be

exposed.

Connect yourself to ground with a wrist strap. Connection

may be made to any grounded metal assembly

in

the

peripheral cabinet. Remember that you

and

the electronic

devices must both be grounded to avoid potentially

damaging static discharges.

• Tum off power before removing

or

installing

power

cords.

• Shut down ConvexOS before cabling drives to the IDC bulkhead.

•

Do

not remove any circuit boards from the drive.

• Never use an ohmmeter

on

any Elite 3 c.ircuit board.

Description

and

specifications

1

Drive specifications

Table 1 Drive specifications

Characteristics

Size

Interface

Capacity (bytes)

This chapter discusses the features, front panel functions,

and

electromechanical

and

physical specifications of

the

CONVEX Elite 3 disc drive (CONVEX model DKD-505). The Elite 3 disc drive

and

related hardware are designed to

be

used

with

the

CONVEX Integrated Disc Channel (IDC).

Table 1 contains the basic specifications for the Elite 3 disc drive.

Conditions

Specifications

Width

146

mm

(5.75

in.)

Height

83

mm

(3.25

in.)

Length

218

mm

(8.6

in.)

Weight

4.1

kg

(9.0

lb)

IPI-2E

(intelligent peripheral

interface)

Unformatted 3.1512019 Gbytes

Chapter 1 Description

and

specifications

1

Table 1 Drive specifications (continued)

Characteristics

Conditions

Specifications

Physical

Number

of discs

11

Number

of data heads

20 (10 logical)

Servo heads

1

Cylinders

per

head/

disc

2738

(0-2737 logical cylinder, 2737

assembly (HOA)

reserved for defect map, cylinder

0

reserved for topology,

and

cylinders

2735

and

2736 reserved

for diagnostics)

Disc diameter

5-1/4

in.

Transfer rate Disc speed

at

5400 RPM

70---104

MHz

(1.5-24.0

Mbytes/s)

Seek time (time required Average

11.5ms to move heads to a different track address)

Full seek

23.5ms

One

track

1.7ms

Latency (time required to

Average

5.55ms reach a particular track

Maximum

11.17

ms

(disc rotation

at

5373

address after head positioning is complete)

RPM)

Start time (following

Maximum

30s

power

sequence delay,

de

is

applied

and

start

conditions are present) Stop time Maximum

30s

Environmental

Operating temperature

10·c

to45°C

(50"Fto 113"F)

with

a

requirements

maximum change of

20"

C (36 • F)

per

hour

Operating humidity

5%

to

95%

relative

Maximum

wet

bulb

of

26 • C (79 •

F)

2

Elite

3

Disc

Drive

Service

Guide

de

power

requirements

The Elite 3 disc drive uses

+5 V de

and + 12

V de. Table 2 contains

the

de

power specifications for the Elite 3 disc drive.

Table 2 Elite 3 disc drive de power requirements

Power

requirements

Supply

voltage

Current

+5V

+12V

Maximum

operating current

5.lA

3.3A

Average idle current

3.9A

1.76A

Maximum starting current

5.lA

6.3A

Maximum seek current

5.lA

3.3A

VoHage

Regulation

±5%

Absolute maximum applied voltage

6.5V

14.0V

The Elite 3 disc drive, power supply, operator panel,

and

mechanical assemblies

must

be

properly

grounded

to

the

peripheral cabinet to ensure error-free operation

and

conformance with regulatory agency requirements.

Chapter 1 Description

and

specifications

3

Front

panel

indicators

and

switches

The front panel contains four indicator LEDs, a write protect switch and indicator,

and

an

address selection switch. Figure 1

shows the front panel components.

Figure

1 Front panel

Disk

drive

front

panel

(FRONT

VIEW}

Ready

indicator

Active

indicator

Error

indicator

Power

indicator

Address

switch

4

Elite 3 Disc

Drive

Service

Guide

BHOB1002

4/16/93

Table 3 contains descriptions of

the

front panel components.

Table 3 Elite 3 front panel indicators

and

switches

Item

Description

Ready LED

Drive is ready when the

LED

is on. This indicates the drive is

in

the spin

up

state

and

is able to accept

read/write

data

and

head

position

commands.

Active LED

LED

is

on

when the drive is selected

on

the IPI interface.

Error LED

A fault condition exists.

Two

digit error codes are displayed. Refer to

Table

12

for a description of error codes

and

an

explanation

of

how

to

read them. This light is

on

when

power

is applied to the drive until the

drive becomes ready.

Power LED LED is

on

when

power is applied to the Elite 3 chassis.

Write protect

Used to enable

and

indicate write protect.

switch/

indicator

Address selection

Used to select the binary logical IPI address (0-7).

switch.

Chapter 1 Description

and

specifications

5

Control

board

jumpers

The disc control board is located

on

the

top

of the disc enclosure.

Jumpers

J9,

Jll,

J16-J19,

and

JSO

are reserved for manufacturing

use only.

J12

controls drive configuration. Figure 2

shows

the

location of the control board jumpers.

Figure 2 Control board jumper location

~JlO

J5

J20

J15

J12

J16

J17

~

rn

J50

c=::::J

J9

rn

J23~

J11A

~

rn rn

1111111111

J48

JllB

J19

rn

J18

~

AH261011

4/22/93

Jumper block

J12

is a vertical header located

on

the control

board. Drive configuration is controlled

by

installing

configuration jumpers between adjacent pins.

6

Elite 3 Disc

Drive

Service

Guide

Table 4 describes the

jumper

block jumpers.

Table 4 Control board jumper block

J12

assignments

Jumper

Pins

Description

Defautt

1

configuration

J12-1

1&2

Master Spindle Sync enable. When a

jumper

is

Not

installed installed, the drive will unconditionally remain in

Master Spindle Sync Mode.

J12-3

3&4

Local Spin

Up

Enable. When a jumper is Installed

installed, spindle rotation begins after

de

power

is applied

and

the sequencing delay is satisfied. When no jumper is installed rotation begins after the unit receives a spin-up command. (normally not installed)

J12-5

5&6

Sweep Enable (Normally not installed).

Not

installed

J12-7

7&8

Manufacturing Test Mode Enable (jumper

Not

installed

must

be off).

J12-9

9&10

Reserved (jumper

must

be off).

Not

installed

J12-11

11 & 12

Manufacturing Test Mode Disable (jumper

Not

installed

must

be off).

J12-13

13&

14

Spare

Not

installed

J12-15

15&

16

Spare

Not

installed

1

Default configuration is subject

to

change. See Convex Elite 3 Configurator

part

number

204-000026-600.

Chapter 1 Description

and

specifications

7

1/0 board

Jumpers

The

I/0

board has 4 jumper blocks (Wl,W2,W3

and

W4) for

configuring the drive.

Table 5

1/0

board jumper block

assignments

Block Jumper

Pins

Description

Default'

configuration

Wl

Wl-1

1&3

Enable position calibrate

on

seek

Not

installed

Wl

Wl-2

2&4

Enable short

RPS

Not

installed

W2

W2-1

1&3

Disable

read/write

diagnostics

Installed

W2

W2-2

2&4

Reserved

Not

installed

W3

W3-1

1&3

Microcode configuration bit 2

Not

installed

W3

W3-2

2&4

Microcode configuration bit 3 Installed

W4

W4-1

1&3

Microcode configuration bit 0

Not

installed

W4

W4-2

2&4

Microcode configuration bit 1

Not

installed

1

Default configuration

is

subject

to

change. See Convex Elite 3 Configurator part number

204-000<Y26-600.

Figure 3 shows the relative location of the

I/0

board

jumpers.

Figure 3

I/

0 board jumper locations

J031

8

Elite 3 Disc

Drive

Seroice

Guide

J20

W2

W4

I I

DO

OJ

\

\,

Wl

J032

AH261012

4/5/93

Power

supply

specifications

Computer Products Power Conversion America manufactures

the

NFS110-7602P power supply used

with

the Elite 3 disc drive.

Universal input voltage allows the

supply

to be powered

in

any

country without changing jumpers

or

switch settings.

Table 6 contains the power supply characteristics

and

Table 7

contains the power supply voltage specifications.

Table 6 Power supply characteristics

Parameter

Condition

Limits

Input voltage

85

Vac

to

264

Vac

Input frequency range

47

Hz

to 440

Hz

Input surge current Cold start

110

Vac

17 A maximum

Cold start

220

Vac

34Amaximum

Output voltage adjustability

+5.1

v

±3%

Over voltage protection threshold

+5.1

V output

6.25

v ±

0.75

v

Total output power@

50°

C ambient Convection cooled

0Wto80W

temperature

Peak

(60

s)

now

Operational environment

Operational altitude

10,000 feet

Nonoperational

40,000 feet

altitude Operational

o·c

to

+5o·c

temperature Nonoperational

-40°C to

+ss·c

temperature Relative humidity

5%

to95%

Weight

1.55

lb

(0.70

kg)

Chapter 1 Description

and

specifications

9

Table 7 Power supply voltage specifications

Output vottage

+5.1

v

+24V

+12V

-12V

Minimum current

OA

Maximum current@

80

W

BA

3.5A

4.5 A

O.SA

Peak

current

20A

4.SA

9A

1.5A

Ripple p-p@ 50 MHz 50mV 240mV

120mV

Total regulation

±2%

+10/-5% ±3%

±3%

Figure 4 shows the component locations for the

power

supply.

Figure 4 Power supply component location

1

Jl

1

J2

5 A,

250

Vac

Fl

0

D

D D D

0

D D D D D

10

Elite 3 Disc

Drive

Seroice

Guide

11

Voltage

adjust

Tl

C6

0

AH261009

5/5/93

Table 8 contains the

power

supply

connector pinout.

Table 8 Power supply pin chart

Connector

Location

number

Signal

Molex

099-50-3051

Jl

with second

and

fourth pins removed

Jl Jl

Molex

09-50-3131

J2 J2

J2 J2 J2 J2 J2 J2 J2 J2 J2 J2 J2

Elite 3 chassis

Pin 1

acground

Pin2

ac neutral

Pin3

achot

Pin 1

+5.1

v

Pin2

+5.1

v

Pin3

+5.1

v

Pin4

Return

Pin5

Return

Pin6

Return Pin 7 Return Pin8

+12V Pin 9 +12V Pin 10

PFD Pin

11

-12V

12

Removed for key Pin

13

-24

v

The Elite 3 chassis, shown

in

Figure

5,

contains the Elite 3 disc

drive,

power

supply, fan assembly, operator panel, air filter,

interface cables,

and

power plug.

Chapter 1 Description

and

specifications

11

Figure 5 Elite 3 chassis

12

Elite 3 Disc

Drive

Seroice

Guide

AH261013

4-/8/93

Unpacking and

installation

2

Unpacking

and

inspection

Step 1 Step2

Step3

This chapter describes unpacking

and

inspection, identifies

major components of the Elite 3 disc drive

and

mechanical

package,

and

provides installation procedures.

This section gives general guidelines for unpacking

and

inspecting the Elite 3 disc drive

and

related assemblies.

Inspection

All shipping containers are designed to protect their components

under

normal shipping conditions. Carefully inspect each carton for signs of shipping damage as it is unpacked.

If

damage is found after visual inspection, document

the damage with photographs

and

contact the transport carrier

immediately.

Unpacking

The customer's bill of materials lists all equipment shipped from CONVEX.

It

should be used as a checklist to ensure that all

equipment has arrived.

Use the following procedure to unpack

the shipping container: Remove each item from its shipping container. Inspect each item for any signs of shipping

damage

as it is

unpacked.

If

equipment damage is found, document the damage,

and

proceed to the next section.

Save all packing material until after operational checkout of the equipment. This enables equipment to

be returned safely to

CONVEX if required.

Chapter 2 Unpacking

and

installation

13

Installation

Damage

claims

If the equipment is damaged, a damage claim form

must

be

completed. Damage claims should

be

completed

by

the customer and given to the shipping representative. Claim forms are normally obtained from the shipping representative.This section describes the installation

of

the Elite 3 disc drive

in a high

performance peripheral cabinet (HPPC).

This section discusses the preparations for

and

installation of the

Elite 3 disc drive.

·

Preparations

Observe the electrostatic discharge procedures described

in

the

"Electrostatic discharge protection" section

on

page xiii, to

prevent damage

to

the drive

during

installation.

Installing

power

strip

and

disc

tray

An

additional power strip

must

be

added

to the

EXP-105

cabinet

to support the maximum number of Elite 3 disc drives. A

maximum of

24

drives may be installed.

The additional power strip is attached to the rear RETMA rail centered between the existing power strips. Figure 6 shows the location of the additional power strip.

14

Elite 3 Disc

Drive

Service

Guide

Figure 6 Additional power strip location

Additional

power

strip

AH261004

4/16/93

Chapter 2 Unpacking

and

installation

15

Figure 7 Elite 3 disc tray

Alignment/retaining

screws

Step 1

Step2

Step3

Step4

Six disc trays may be installed

in

an

EXP-105

peripheral

expansion cabinet. Each tray, shown in Figure

7,

has four

alignment/retaining screws in the front

and

rear of the tray for

the installation of

up

to four Elite 3 disc drives.

Installing

the

Elite 3 chassis

AH261005

4/5/93

The Elite 3 chassis has an alignment slot located to the

rear,

and

a mounting tab in front of the chassis.

To

install the chassis

in

a

disc tray: Remove the front alignment/retaining screw from the disc tray

and

loosen the rear alignment/retaining screw.

Slide the Elite 3 chassis into the tray, aligning the rear screw with

the slot in the chassis. Install the front alignment/retaining screw

and

tighten the front

and rear screws. Select the appropriate drive address with the address selection

switch on the chassis front panel.

16

Elite

3

Disc

Drive

Seroice

Guide

Caution

Step 1 Step 2

Step3

Connecting a single drive

Each Elite 3 disc chassis

has

3 connectors

at

the rear

of

the

assembly;

an

ac connector,

an

IPI

input

connector,

and

an

IPI

output

connector.

As root, use the

/etc/ebutdown

command

to

halt

ConvexOS

before connecHng a disc

to

the IDC. Failure

to

do

so

may

cause a

system crash

and

loss of data.

To

connect a single drive to

an

JDC:

Install the Elite 3 chassis

in

the peripheral cabinet.

Ensure the ac

power

switch

at

the

rear

of

the

Elite 3 chassis

is

in

the OFF position. Attach the IPI cable to the

mate connector

on

the

rear

of

the

Elite

3 chassis.

Figure 8 Connecting a single drive

To

IDC

port

AH261008

4/5/93

Chapter 2 Unpacking

and

installation

17

Step4

Attach the IPI cable to the appropriate JDC bulkhead connector.

It

is not necessary to remove power from the IDC

when

connecting disc drives to the IDC.

Step 5 Figure 9 shows the factory bulkhead

port

assignments for

C3200/C3400. IDC ports are connected with

port

3, located

at

the top of the bulkhead when assembled in the factory.

Figure 9 IDC bulkhead port assignments for C3200/C3400

0

Port3

0

Port2

0

Port 1

0

Porto

18

Elite 3 Disc

Drive

Seroice

Guide

Figure 10

shows

the

relative

port

assignments

for

the

IDC

bulkhead

in

a C3800 system. IDC

ports

are

connected

with

port

0, located to the left

of

the

bulkhead

when

assembled

in

the

factory.

Figure 10 IDC

bulkhead

port

assignments for C3800

P2 P3

Step6

Step7

Step 8

P4

PS

P2-

Used

for port 0

P3-

Used

for port I

P4-

Used

for port 2

PS-

Used

for port 3

Install

the

IPI terminator to

the

female

connector

at

the

rear

of

the

chassis.

Attach

the

ac

power

cord

and

plug

into

an

available

convenience

outlet

in

the

power

strip. Careful

planning

will

avoid

relocating

power

cords

when

additional disc

drives

are

added.

Turn

on

the

Elite 3 disc

drive

and

integrate

the

drive

into

the

system. Refer to

Chapter

3, ''Integration

and

testing,"

on

page 27

for information

about

integration.

Chapter 2 Unpacking

and

installation

19

Caution

Step 1 Step2

Step3 Step4

Step5

Step6

Connecting multiple disc drives

Each port

on

an

IDC can

support

up

to eight disc drives. These

may

be

any combination of DKD-501, DKD-502, DKD-503,

DKD-504, or DKD-505 disc drives.

As root, use the

/ate/shutdown

command

to

halt ConvexOS

before connecting a disc to the

IDC.

Failure

to

do

so

may

cause

a system crash and

loss

of data.

To

install multiple disc drives

on

an

IDC: Install the Elite 3 chassis in the peripheral cabinet. Ensure the ac power switches

at

the rear of the chassis are

in

the

off position.

Repeat Step 3 through Step 6 for each IDC

port

containing disc

drives. Refer to Figure

11

on

the following page.

Attach the

IPI cable to the IDC bulkhead port.

Attach the

IPI cable to the male connector

at

the rear of

the

first

Elite 3 disc drive in the daisy chain. Install IPI daisy chain cables between additional drives

in

the

daisy chain. Install the terminator

plug

on

the last drive in

the

daisy chain.

20

Elite 3 Disc

Drive

Service

Guide

Figure

11

shows the installation of multiple drives, daisy chain

cables, and termination.

Figure

11

Connecting multiple drives

To

JDC

port

Step

7

Step8

Daisy

chain

cables

AH261006

4/5/93

Attach power cords

and

plug drives into available convenience

outlets on the power strips. Turn on the drives

and

integrate the drives into the system. Refer

to Chapter

3,

"Integration

and

testing,"

on

page 27 for more

information about integration.

Chapter 2 Unpacking

and

installation

21

Physical

configuration

A wide variety of configurations are possible

when

connecting

DKD-501, DKD-502, DKD-503,

DKD-504,

or

DKD-505

drives

to

an

IDC. The Elite 3 drives

may

be

added

to systems

that

currently contain Sabre 7

or

Sabre 5 disc drives. · ·

Each IDC can support

up

to

32

disc drives. The

maximum

cable

length

alln~edJ

between the IDC

and

the last disc

drive

is 100

'Ute.~

1.

~

'"2-S-

~

.

When planning your installation, consider the following guidelines:

22

Elite 3 Disc

Drive

Service

Guide

• Group disc drives attached to a single IDC

port

together.

• Address drives in a consistent manner, starting

with

unit

0

in

the lower left-hand side

of

the cabinet,

and

incrementing

unit addresses horizontally.

• Allow for service access

when

placing the peripheral

cabinets.

• Generally, address switches determine drive numbers.

Figure

12

shows the connection of

32

disc drives to

an

IDC.

Figure

U

IDC

maximum drive configuration

IPI port

r::J. I I I

11

i 1

I

I I I

I

3

1~ve11m:ve11~vell

4

vel

drive

drive drive drive

4

5

6 7

IPI port

T~

..

11~~11~~11~~11

~

..

11~~11~

..

11

~ve

I

2

IPI port

r-1

I I I I I I I I I I I I I I

1

drive drive

drive drive drive

drive

drive drive

0

1 2

3

4

5

6 7

IPI

r-1

I

port

'TI'a

I I

1

0

drive

BJ

drive I drive I drive drive

drive

I

~

..

I

0

2 3 4 5

6

-

IDC

Chapter 2 Unpacking

and

installation

23

A single peripheral expansion cabinet can contain a

maximum

of

24

Elite 3 disc drives. When connecting disc drives to

an

IDC,

disc drives attached to a single

port should

be

grouped

in

horizontal rows to allow the use of the single hot spare cable.

Figure

13

shows the relative locations of 32 Elite 3 disc drives

when

installed in 2 peripheral expansion cabinets.

Figure 13 Expansion cabinet drive locations for

32

drives

I

IPI

port2

drive4

IPI

port2

driveO

IPI

port 1

drive4

IPI

port 1

driveO

IPI-

portO

drive4

IPI

portO

driveO

I

First peripheral

expansion cabinet

Fan assembly

IPI IPI

port2

drives

drive 6

IPI

port2

drive 1 drive 2

IPI

portl

drives

drive 6

IPI

portl

port 1

drive 1 drive2

IPI IPI

portO

drive S drive6

IPI

portO portO

drive 1 drive 2

PDU

24

Elite 3 Disc

Drive

Service

Guide

11

IPI

port2

drive 7

IPI

port2

drive3

IPI

port 1

drive 7

IPI

port 1

drive 3

IPI

portO

port3

drive 7 drive4

IPI

portO

port3

drive 3

drive

0

I

Second peripheral

expansion cabinet

Fan assembly

IPI

port3 port3

drives

drive 6

IPI

port3 port3

drive 1 drive 2

PDU

I

IPI

port3

drive 7

IPI

port3

drive3

I

Although it is possible for

an

IDC to

support

32

Elite 3 disc

drives, a typical configuration

may attach 16 Elite 3 disc drives

toanIDC. Figure

14

shows

the

relative locations of 16 Elite 3 disc

drives

installed

in

a single peripheral expansion cabinet.

Figure 14 Expansion cabinet drive locations for 16 drives

Fan assembly

IPI

IPI IPI

port3

port3 port3

driveO drive 1 drive2

drive3

IPI

IPI IPI

port2

driveO

drive 1 drive2 drive3

IPI

portl

port 1

portl

driveO drive 1 drive 2

drive3

IPI

IPI IPI

IPI

portO

portO portO

portO

driveO

drive 1 drive2

drive3

I

POU

I

Chapter 2 Unpacking

and

installation

25

Integration

and

testing

3

Software

integration

Note

Step

1 Step2 Step3 Step4 Steps Step6 Step

7 Step8 Step

9

Step

10 Step

11

This chapter contains guidelines for integrating Elite 3 disc drives into

ConvexOS

and

information about the CONVEX

diagnostic tests.

Managing

ConvexOS

Configuration

Guide

(DSW-030) describes device integration in detail. This section contains a summary of the procedures used to

add

an Elite 3 disc drive

and

descriptions

of the system files that may require modification.

The

Elite 3 disc

drive

requires

ConvexOS

Vl

0.1

or

later.

Untll

ConvexOS

Vl

1.0

Is

released, a

sysgen

will

be

required

to

support

the

Ellte 3 disc

drive.

General

integration

procedure

The following steps are used to

add

a disc drive to the system: Install the physical hardware. Edit the /ioconfig

file.

Use

idcfmt

to verify operation

and

format.

Boot

ConvexOS to single-user mode.

Verify the entry in the

/etc/disktab

file.

Create block and character special device files using

MAKEDEV.

Create mount points. Configure the partitions with

newfs

or

newst.

Edit /etc/fstab.

Use

preen

to check disc integrity.

Boot

ConvexOS to multiuser mode.

Chapter 3 Integration

and

testing

27

/ioconfig

file

ConvexOS

uses

a configuration file

(/ioconfig)

located

on

the

service processor

unit

(SPU) disc to

identify

system

level

hardware.

The

/ioconfig

file describes

in

hierarchical

fashion

the

connections

between

the

integrated

disc

channels

(IDCs)

and

peripheral

devices. ConvexOS

uses

this

information

to

assign

a

logical

number

to a device

of a given

type.

Each

type

of

device is identified to ConvexOS

by

mnemonic

device

code. The device

and

marketing

code

for

the

Elite 3

disc

drive

is:

DKD-505-Elite

3 disc

drive

Elite 3 entries

in

the

/ioconfig

file

on

the

!:)PU

disc

contain:

• IDC

unit

number-Determined

by

physical

location

of

the

IOC

• IPI

port

number-Determined

by

connecting

to

the

JDC

•

Driver

number-Determined

by

device

driver

type

•

Unit

number-Determined

by

device

address

selection

•

Type-Device

code

for

the

device

• Logical

unit

number-Used

to create

the

desired

file

system

structure

(optional)

ConvexOS

uses

this information

during

autoconf

to

assign

a

logical device

number

of a given

type. This

determines

which

device files

found

in

the I dev

directory

will

be

used

for

each

disc

drive

identified

in

the

/ioconfig

file.

28

Elite 3 Disc

Drive

Seroice

Guide

In

the example /ioconfig file

shown

in

Figure 15, four elite 3 disc

drives are attached to

an

JDC

located

in

CCU slot

0.

Two of the

Elite 3 disc drives are attached to

IPI channel 0

and

two Elite 3 disc

drives are attached to

IPI channel 1. In example 1, the logical

assignments have been

made

sequentially.

Figure

15 /ioconfig example 1

idc

0 ipi

0

drvr

DKC-IP2

ipi

1

unit

o

type

DKD-505

duO

unit

1

type

DKD-505

dul

drvr

DKC-IP2

unit

0

type

DKD-505

du2

unit

1

type

DKD-505

du3

In the /ioconfig file shown

in

Figure 16,

an

additional drive is

added

to IPI

0,

addressed as unit

2,

and

assigned a logical

designation of du4. By

adding

an

additional drive

on

IPI 0

and

specifying the logical designation du4, the file system built

on

existing disc drives is not affected. Figure 16 /ioconfig example 2

idc

0

ipi

0 drvr

DKC-IP2

ipi

1

unit

0

type

DKD-505

duo

unit

1

type

DKD-505

dul

unit

2

type

DKD-505 du4

drvr

DKC-IP2

unit

O

type

DKD-505

du2

unit

1

type

DKD-505

du3

Chapter 3 Integration

and

testing

29

/etc/disktab

The I

etc/

disktab file describes disc types, disc geometry, file

system partition sizes,and default block

and

fragment sizes. Do

not

change

/etc/disktab

if

an

entry for the Elite 3 disc

drive

exists. Figure

17

contains the

/etc/disktab

entry for

an

Elite 3 disc

drive.

Figure

17

Example

/etc/disktab

rdkd-SOSIDKD-5051Elite3-2HPISeagate

ST43200K 3.0SGB

IPI-2

disk:\ :ty=winchester:spilS:set2048:nst53:nttlO:nct2734:rmt5400\ :pat70180:bat16384:fat2048:\ :pbt280555:bbl16384:fbi2048:\ :pct1402490:bct65536:fci8192:\ :pdf70540:bdt32768:fdt4096:\

:pei420320:bet16384:fet2048:\ :pfl140285:bfi16384iffi2048:\ :pgi631145:bgi16384:fgi2048:\ :phi420610:bhi16384:fht2048:

Table 9 shows

/etc/disktab

types

and

descriptions.

Table 9 /etc/disktabdescription

Name

Type/description

ty

Type of disc

se

Number of bytes

per

sector

sp

Number of spare sectors

per

cylinder

ns Number of sectors

per

track

nt

Number of tracks

per

cylinder

nc Number

of

cylinders

per

disc

rm

Disc speed (revolutions

per

minute)

p[a-h]

Partition sizes (sectors)

b[a-h]

Partition block sizes (bytes)

f[a-h]

Partition fragment sizes (bytes)

30

Elite 3 Disc

Drive

Service

Guide

Testing

the

Elite 3

disc drive with

idcfmt

Caution

The Elite 3 disc drives are tested

by

the

idcfmt

diagnostic

program. There are two versions of

idcfmt.

The first version of

idcfmt

runs

on

the SPU

when

ConvexOS is not running. The

second version of

idcfmt

is

run

under

ConvexOS

as

root.

Appendix A contains the

idcfmt

man

page. The diagnostic

program:

• Formats disc drives

• Verifies previously formatted disc drives

• Performs disc maintenance

Some subtests

In

idcfmt

are

data

destructive. Refer to the

ldcfmt

(10) man

page

for a

detailed

description

of

idcfmt.

Verifying format

Elite 3 disc drives are formatted before shipping to the customer site.

Use the

verify

option of

idcfmt

to ensure that the

drive

is correctly cabled to the system

and

has been formatted,

and

that each entry in the defect list points to a sector that has been

slipped.

idcfmt

is invoked by typing its name, a

-d,

followed

by

the

CCU slot number, IPI port number,

and

drive

unit

of the drive to

be used. A list of options may follow, then zero

or

more

commands.

If

no commands are given

on

the command line,

idcfmt

displays a prompt

and

expects commands to be entered

from standard input. In the following example,

idcfmt

verifies the format

of

the

drive attached to the IDC in slot

0,

port

0,

unit

2.

idcfmt

-d

0 0 2

verify

Chapter 3 Integration

and

testing

31

Caution

No

write operations occur

when

idcfmt

is

used

with

the

verify

option.

To

verify that the drive is formatted:

• Read one copy of all the data

on

the topology cylinder. Verify

that the checksums

and

magic numbers are correct.

• Read the remaining copies of the topology

data

and

verify

that they match the data from the first step.

• Read all logical blocks

on

the disc

and

verify that there

are

'

no header CRC

or

data ECC errors.

If

any

errors are found,

they are listed.

,

•

Read all the sector headers from the disc

and

verify that

there is a one to one correspondence of entries

in

the

defect

list with sectors that are slipped.

• Verify that each entry in the defect list points to a slipped sector.

Formatting

a drive

The

format

command option is used

with

idcfmt

to format a

drive that has never been formatted.

If

a partial format is found,

the format resumes

at

the appropriate point,

with

the

manufacturers

and

grown defect lists that were

in

effect

when

the format was interrupted.

If

the drive has no valid format

on

it, then the program reads the manufacturers defect

data

from

the drive. Under most circumstances a format is not required

in

the field.

Using

the

fo:mat

opHon

of

the

idcfmt

command

will destroy

any

existing flies on the disc drive.

In the following example,

idcfmt

formats

an

Elite 3 disc

drive

attached to the

JDC

in

slot

0,

port 0

and

addressed as

unit

2:

idcfmt

-d

0 0 2

fonnat

32

Elite 3 Disc

Drive

Service

Guide

'

Maintenance and

IPB

4

Troubleshooting

This chapter contains maintenance precautions, removal

and

replacement procedures,

and

an

illustrated

parts

breakdown

(IPB).

This section contains information for interpreting

the

hardware

errors originating from the disc

drive

and

indicated

through

the

error light message codes. Information

about

the

fault

symptom

codes

(F'SC)

associated

with

the

disc

subsystem

is

also provided.

Elite 3 disc

drive

error

codes

The Elite 3 disc drive generates message codes for certain

circumstances

and

presents these codes to the JlO connector. This

connector is used to

drive

the error light

on

the

front panel.

Message codes consist

of

two digits

composed

of

the

numbers

1-9. The message code starts

with a long

pause

followed

by

short

pulses for the first digit, a short pause,

and

then

more

short

pulses for the second digit. The code

can

be

determined

by

counting the short pulses

that

represent the

two

digits.

In addition to being presented to

the

error

indicator on.the front

panel, FSCs associated with the

drive

messages

are

recorded

in

the

/mnt/errlog

file

on

the

SPU disc.

Chapter 4 Maintenance

and

IPB

33

Table

10

contains descriptions of the error codes

generated

by

the Elite 3 disc drive.

Table

10

Elite 3 disc drive error codes

Error

Description

Corrective

action

code

11

Invalid micro code

ID-Switch

settings are

Check jumper installation

on

disc

incorrect

control board

21

Illegal

condition-Master

terminate

indicated,

but

the hardware did not indicate

the sending of

90

ending status

22

External RAM failure

23

Unexpected vectored

interrupt/trap

occurred

24

CSAW failure detect in power

up

Cycle power; if failure continues,

initialization

replace disc drive

25

Buffer memory failure detected in power

up

31

BIPIP IPC failure during power

up

32

BIPIP

SFC

failure

during

power

up

33

BIPIP

BUF

failure

during

power

up

34

BIPIP

VIB

failure

during

power

up

35

VIC failure

during

power

up

36

BIPIP

ECC

failure

during

power

up

34

Elite 3 Disc

Drive

Service

Guide

Fault

symptom

code

CFSC)

Fault symptom code (FSC) is a standardized

method

for

reporting error conditions

on

the

CONVEX JDC

and

attached

JPJ

disc products. The

FSC

is a hexadecimal code printed to the console

and

errlog

upon

detection of

an

error condition. The format

of

the

FSC is:

FSC

OxNNNN

idc

N

port

N

unit

N

cyl

OxNNN

trk

OxNNN

sec

OxNNN

p N

cnt

N

Table

11

contains a description of

the

FSC

fields.

Table

11

FSC

field descriptions

FSCfield

Description

FSCOxNNNN Hexadecimal representation of the fault symptom code. idc

The

JDC

channel number that reported the error.

port The port

on

the

JDC

that reported the error.

unit The address of the drive that reported the error.

cylOxNNN

The hexadecimal cylinder number associated with the reported error. This field is always reported even if the error

was

not a

data

or

seek

error.

trkOxNNN The hexadecimal track number associated

with

the reported error. This

field is always reported even

if

the error

was

not a

data

or

seek error.

secOxNNN

The hexadecimal sector number associated with the reported error. This field is always reported even if the error

was

not a

data

or

seek error.

p

The partition number in which the error occurred. Partition 1 is a, 2 is b, 3 is c,

and

so on.

cnt

The number of times the operation was retried. Error threshold is

11.

If

the cnt field is less than

11

the retry succeeded.

Chapter 4 Maintenance

and

IPB

35

Table 12 contains a list

of

fault

symptom

codes

used

in

the disc

subsystem, the definition

of

each fault

symptom

code,

and

an

explanation of the condition

that

causes

the

code

to

generate.

Table

12

FSC descriptions

FSCcode

Definition

Explanation

Ox.0000

Normal success

No

error condition. This

should

never

appear

in

an

status

FSC

error message.

Ox0006

Device driver

not

This error is returned

if

an

invalid

port

number

is

found

specified in the CMI message dst_id field.

The

IDC

supports ports

0-3.

Any

other

number

is invalid.

Check /ioconfig for invalid IPI

p0rt

numbers.

Ox.1001

Device

not

found

Not

used

by

IDC.

Ox.1002

Unit

not

found This error is returned if the

unit

specified

cannot

be

selected

or

does

not

respond to a logical reset. Check

connection

and

drive

address

setting.

Ox.1003

Unit already

An

attempt

was

made

to move from CMI

boot

level

to

connected

physical level,

but

the

unit

was

already

at

physical

level.

Ox.1004

Unit disconnected

The

unit

was not connected

at

CMI physical level for

an

operation that is only valid

at

the physical level.

This can occur while:

•Attempting

to disconnect the

unit

• Attempting to

put

the

unit

online

• Attempting to slip a sector

•Attempting

to

unslip

a sector

•Attempting

physical level

I/0

operations

Ox.1005

Unit already online

An

attempt

was

made

to

move

from

the

CMI physical

level to the logical level,

but

the

unit

was

already

at

the logical level.

Ox1006

Unit offline

An

attempt was

made

to move from

the

CMI logical

level to the physical level,

but

the

unit

was

not

at

the

logical level.

Ox1007

to

Not

used

by

IDC.

OxlOOa

36

Elite 3 Disc

Drive

Service

Guide

Table 12 FSC descriptions (continued)

FSCcode

Definition

Explanation

OxlOOb

Function time-out

The IDC timed

out

on

a read

or

write operation,

or

an

exceeded

expected attention from a

unit

did

not

arrive

prior

to

time-out.

OxlOOd

Function aborted

by

If

the IDC receives a DEBUG_FLIMBS message, all

CCU

pending operations

on

the specified device are

aborted

and

the error is posted.

OxlOOf

Invalid CMI function

An

invalid CMI function code

was

passed to the IDC.

code

Oxl012 to

Not

used

by

IDC.

Oxl015

Ox1016

CMI revision

An

invalid CMI revision level

was

passed

in

the CMI

unsupported

by

CCU device class field. For

the

IDC, the revision level

must

be

zero.

Ox1017

Defective CMI

The CMI message

was

defective.

message (badparms)

Oxl019 to

Not

used

by

IDC.

Ox

I Ole

OxlOlf Controller not The unit was not initialized

and

an

attempt

was

made

initialized

to connect, disconnect, set geometry, transition offline, transition online, slip a sector,

unslip

a sector,

perform

physical

or

logical

1/0,

perform

long read

or

write,

or

read the configuration.

Ox1020to

Not

used

by

IDC.

Ox1027 Ox1028

Memory allocation This code is never printed

in

an

FSC error message.

It

(memalloc)

failed

is

used

with adb88 to indicate the breakpoint table is

full.

Ox1029

Not

used

by

IDC.

Ox102a

Unit not ready

The error is returned if the

drive

status cannot

be

read

or

if the device cannot

be

selected after reset.

Ox102b

Could

not seek to A seek error occurred.

required location

Ox102c

Recoverable IDC

An

ECC error occurred

and

the

data

was corrected.

ECCerror

Chapter 4 Maintenance

and

IPB

37

Table

12

FSC descriptions (continued)

FSCcocle

Definition

Explanation

Ox102d

Bad header (chksum)

This indicates a header CRC error.

Ox102e

to

Not

used

by

IDC.

Ox102f

Ox1030

Bad device address

Not

used during normal operation

of

the

IDC.

Ox1031

Bad parity

of

Used

by

adb

to indicate attempt to access nonexistent

nonexistent memory

memory.

Ox1032

to

Not

used

by

IDC.

Ox1049 Ox104a

IOS reset failed

A selective reset of the

unit

failed.

If

the CMI message

can

be examined, the status.extend field contains

the

DICE

EXREG

for the failed operation.

Ox104b

Readformat

A read format command failed.

If

the CMI message

command failed

can

be examined, the status.extend field contains

the

DICE

EXREG

for the failed operation.

Ox104c

Read configuration

A read configuration command failed.

If

the

CMI

command failed

message can

be

examined, the status.extend field

contains the DICE

EXREG

for the failed operation.

Ox104d

Tried to slip

bad

An attempt was

made

to slip a sector

that

is already

sector marked bad.

Ox104e

Tried to unslip a An attempt was

made

to unslip a sector

that

is

already

good sector marked good.

Ox104f

Missed sync byte A missed sync byte

data

exception occurred. This will

occasionally

happen

if

the IDC software is

not

keeping

up

with the drive.

Ox1050

Drive reported

an

General purpose error message from disc drive. Sync

error

detected

by

one head

but

not

both.

Ox1051

Parity error

A parity error occurred

on

the IPI bus,

in

the B

bus

octet

or

in

an

external field.

Ox1052

Disc format

1/0

was attempted

on

a disc that has

not

been

inconsistent

formatted properly.

Ox1053

Read status

A read status command failed.

If

the CMI message can

command failed

be examined, the status.extend field contains

the

DICE

EXREG

for the failed operation.

38

Elite 3 Disc

Drive

Seroice

Guide

Table

12

FSC

descriptions (continued)

FSCcode

Definition

Explanation

Ox1054

Recalibrate command

A drive recalibration command failed (load slave

failed

function, command

Ox28).

Ox1055to

Not

used

by

IOC.

Ox1057

Ox1058

Header

parity error

A parity error occurred

in

the header field.

Ox1059

Header

miscompare

The header

on

the disc

did

not match the

header

error

expected

by

the software. This will occasionally

happen if the

IOC software is

not

keeping

up

with

the

disc.

If

this error code is reported frequently, there

may be a disc hardware problem.

Ox1060

Drive verify header

This can be a data error

or

a true miscompare.

miscompare

Ox1061

Busy doing

General purpose error message from disc drive.

something else

Ox1062

Out

of spare General purpose error message from disc drive.

cylinders

Ox1063

Usually a warning, General purpose error message from disc drive. not fatal

Ox1066

Read specification A drive read disc specification values

command

command failed.

Ox1067

Read buffer control A read buffer control command failed. command failed

Ox1068

Read formatter A read slave formatter parameters command failed. parameters failed

Ox1069

Recoverable drive The drive reported

it

detected

and

corrected a

data

ECC error error

on

the given sector.

Ox106a

Drive sent wrong

The drive sent a different sector

than

the

one

the IOC

sector

expected. The expected location of the missed sector is

printed

and

stored

in

the status.extend field of the

CMI message. This becomes a fatal error after

11

retries.

Ox106b

Drive buffer

The drive data buffer is temporarily unavailable. This

over

/underflow

becomes a fatal error after

11

retries.

Chapter 4 Maintenance

and

!PB

39

Table

1l

FSC

descriptions (continued)

FSC

code

Definition

Explanation

Ox2004

Geometry not set

The geometry

has

not been set,

and a command

that

requires geometry

has

been issued.

Ox2005

Unable to read

This error should not occur

during

normal operation.

topology

map

Ox2006

Unable to read either

There are two sections reserved

by

the

IDC for storing

ucode

temp

section

slip sector state information.

If

they cannot

be

read,

this error is returned.

Ox2007

Unable to read

any

An

error occurred

during

the translation of a logical

headers

on

cylinder 0

block to a physical cylinder,

trac~,

and

sector.

Ox4000

Select failed A drive selection command failed.

Ox4001

Deselect failed A drive deselection command failed.

Ox4002

Load drive format

A load format command failed.

failed

Ox4003

Section invalid

The checksum

or

magic

number

of

a section is bad.

If

the CMI message can

be

examined,

the

status.extend

field contains the DICE

EXREG

for

the

failed

operation.

Ox4004

PIGAfailed

An

unexpected interrupt occurred

during

a PIGA

transfer between

main

memory

and

the

data

buffer.

If

the CMI message can

be

examined, the status.extend

field contains the DICE

EXREG

for

the

failed

operation.

Ox4005

Load buffer control

A drive load buffer control

command

failed.

failed

Ox5000

Unable to disable

A disable class 2 interrupt attention

command

failed

class 2 interrupt

(load slave function

OxtA).

If

the

CMI message can

be

examined, the status.extend field contains

the

DICE

EXREG

for the failed operation.

Ox5001

Enable sync master

A enable sync command failed (load slave function,

failed

command

Ox2f).

If

the CMI message can

be

examined,

the status.extend field contains the DICE

EXREG

for

the failed operation.

Ox5002

Unit not correct type

The device connected to

the

IDC indicates it is

not

a

of

device

disc. The device

type reported

by

the read

configuration command

must

be

one

(type 1 is disc).

40

Elite 3 Disc

Drive

Service

Guide

Table 12

FSC

descriptions (continued)

FSCcode

Definition

Explanation

Ox5003

Uncorrectable IDC

An

uncorrectable ECC

error

occurred.

ECCerror

Ox5004

Load

head

command

The load head address

command

failed.

failed

Ox5005

Load target

The load

RPS

target sector

address

command

failed.

command

failed

Ox5006

Enable alternate sync

An

enable alternate sync byte

command

failed (load

byte failed slave function command

Ox35).

If

the CMI message

can be examined, the status.extend field contains

the

DICE

EXREG

for the failed operation.

Ox5007

Enable

primary

sync

An

enable primary sync byte

command

failed (load

byte failed

slave function command

Ox36).

If

the CMI message

can be examined, the status.extend field contains

the

DICE

EXREG

for the failed operation.

Ox5008

Uncorrectable drive The drive reported a

data

error

and

was

unable to

ECCerror

correct it with ECC.

Ox5009

Unable to enable

An

enable class 2 interrupt attention

command

failed

class 2 interrupt

(load slave function

command

OxlB).

If

the

CMI message can be examined, the status.extend field contains the DICE

EXREG

for

the

failed operation.

Ox500a

Unable to flush drive

A drive load

end

of write

command

failed.

If

the CMI

data

buffer

message can be examined, the status.extend field contains the DICE

EXREG

for

the

failed operation.

OxSOOb

Unable to enable

An

enable slave ECC correction

command

failed (load

ECC correction

slave function command

Ox21).

If

the

CMI message

can be examined, the status.extend field contains

the

DICE

EXREG

for the failed

~peration.

OxSOOc

Unable to enable

An

enable slave ECC/CRC reporting

command

failed

ECC reporting

(load slave function

command

Ox2D).

If

the

CMI

message can be examined,

the

status.extend field

contains the DICE

EXREG

for the failed operation.

Ox500d

Unable to load buffer

A drive load buffer available

interrupt

delay

interrupt delay

command failed.

If

the CMI message can be examined,

the status.extend field contains the DICE

EXREG

for

the failed operation.

Chapter 4 Maintenance

and

IPB

41

Table 12 FSC descriptions (continued)

FSCcode

Definition

Explanation

Ox500e

Unable to

tum

off

A disable slave ECC/CRC correction

and

reporting

driveECC

command failed (load slave function

Ox2C).

If

the CMI

message can

be examined, the status.extend field

contains the DICE

EXREG

for the failed operation.

0500£

Unable to enable

An

enable single

burst

ECC/CRC reporting

command

single burst report

failed (load slave function command

Ox8C).

If

the CMI

message can

be examined, the status.extend field

contains the

DICE

EXREG

for the failed operation.

Ox5010

Unable to disable

A disable single burst ECC/CRC reporting

command

single burst report

failed (load slave function command

Ox8D).

If

the

CMI

message can

be examined, the status.extend field

contains the DICE

EXREG

for the failed operation.

Ox5011

Unable to disable

A disable internal position calibration initiation

auto thermal comp

command failed (load slave function

command

Ox39).

If

the CMI message can be examined, the status.extend

field contains the DICE

EXREG

for

the

failed

operation.

Oxbadd

Used

in

send

General purpose error message from disc drive.

message

Ox6000

Buffer overrun

or

General purpose

JDC

error message.

underrun

error

Ox6001

CCU has detected a

If

this error is returned the IDC has crashed.

Only

fatal error debug messages are accepted after a crash.

Ox7000

Interface

not

initially The DICE indicated

an

interface

was

not

initially

an

idle idle error. The IDC software attempted a deselect

operation

and

the interface

did

not

return idle,

or

an

invalid state transition has taken place.

Ox7001

Interface not

The

DICE

indicated

an

interface

did

not

return

to idle

returned to idle

error. The IDC software attempted a deselect operation

and

the interface

did

not

return idle,

or

an

invalid state transition has taken place.

Ox7002

Invalid command

The operation failed

due

to

an

invalid DICE

command

status

status.

Ox7003

Register filed CRC

A

CRC

error was detected

in

the DICE register field.

error

42

Elite 3 Disc

Drive

Service

Guide

Table

12

FSC descriptions <continued)

FSCcode

Definition

Explanation

Ox7004

Register field parity

A parity error

was

detected

in

the DICE register field.

error

Ox.7005

Invalid transfer

The operation failed

due

to

an

invalid DICE transfer

status

status.

Ox7100

Bus B parity error

General purpose IDC error message.

Ox7101

Spontaneous drive

General purpose IDC error message.

deselection

Ox7102

DICE

command

General purpose IDC error message.

time-out exceeded

Ox7103

DICE header parity General purpose IDC error message. error

Ox7104

DICE external parity General purpose IDC error message. error

Ox7105

IPI parity error General purpose IDC error message.

Ox7106

IPI

bus

control General purpose IDC error message.

rejected

Ox7107

IPI

odd

transfer size General purpose IDC error message.

Ox7108

IPI 3

command

General purpose IDC error message.

exception

Ox7109

IPI 3

command

General purpose IDC error message.

time-out

Ox710a

IPI 3 machine General purpose JDC error message. exception

Ox710b

Invalid slave

ending

General purpose IDC error message.

status

Ox710c

IPI 3 alternate

port

General purpose IDC error message.

exception

Ox710d

IPI 3 intervention General purpose

JDC

error message.

request exception

Ox710e

IPI 3

vendor

General purpose IDC error message.

exception

Ox710f

IPI sync error

General purpose JDC error message.

Chapter 4 Maintenance

and

IPB

43

Table

12.

FSC

descriptions (continued)

FSCcode

Ox7110

Ox7111

Definttlon

Explanation

IPI 3 incomplete

General purpose IDC error message.

substatus Failed

due

to

bad

General purpose error message from disc drive.

spot

on

media

Sector

errors

The

slip

command

in

idcfmt

is used to repair

header

CRC

or

data

ECC

errors.

If

a sector

in

the user

data

area

of

the disc is slipped, this should be done with exclusive access to the disc. Sectors in the topology

and

diagnostic areas require exclusive

access to the disc. A description

of

the topology

is

included

in

the idcfmt (10)

man

page in Appendix A.

When slipping a sector with

idcfmt

running from ConvexOS,

idcfmt

attempts to lock the disc

during

the slip operation.

The

file system containing the sector to

be

slipped

should

be

unmounted. Fault symptom codes

(FSC)

are printed to the console

and

the

errlog upon detection of

an

error. The following list identifies

some of the

FSCs

that indicate a sector

or

sectors should be

slipped.

• Ox102c-Recoverable IDC ECC error

• Ox102d-Bad header (checksum)

• Ox1069-Recoverable drive ECC error

• Ox5003-Uncorrectable IDC ECC error

• Ox5008-Uncorrectable drive ECC error

Multiple occurrence of recoverable ECC errors

on

a single sector

should be slipped before they become unrecoverable ECC errors.

In the following example,

idcfmt

is used to slip sector 6

of

track 5

on

cylinder

10

in the drive addressed as

unit

2, which

is

attached to port 0 of the IDC

in

CCU slot

0.

idcfmt

-d

0 O 2

slip

10 5 6

For additional information about the use

of

idcfmt,

refer to

the

idcfmt(lD) man page in Appendix A.

44

Elite

3

Disc

Drive

Service

Guide

FSC

codes

reported

during

autoconf

The

autoconf

routine is executed during the boot process.

autoconf

attempts a probe

and

attach

on

each device identified

in the /ioconfig

file

on

the SPU disc. During the probe

and

attach

of

an

Elite 3 disc drive each sector

in

cylinder 0 is read.

If

an

Elite

3 has not been formatted with

idcfmt,

multiple errors are

reported. The

FSC

codes you may see include:

• Ox102d-Bad header CRC error

• Ox104f-Missed sync byte

• Ox2005-Unable to read topology

map

• Ox4003-Section invalid

• Ox6000-Buffer overrun

or

underrun

error

The

FSC

Ox102d

is also reported when the Elite 3 drive has been formatted by the vendor with a format that is incompatible with the

ConvexOS device driver.

WM

messages

When a disc in a redundant stripe fails,

VVM

checks to see

if

a

hot spare is available for the affected partitions.

If

so,

WM

reconstructs the data from that disc

on

the specified hot spare

device and continues operations. Figure

18

shows the message printed by

WM

when

a disc

failure is detected.

Figure

18

WM

device failure message

/Failure:

device

/dev/dxxx

in

stripe

/dev/stx

failed

\.

After printing the device failure message to the console,

VVM

attempts to instruct the

vvmdaemon,

to begin the automatic

reconstruction process.

If

VVM

successfully contacts the

daemon,

and

the daemon is able to perform the reconstruction,

no

human

intervention is required.

Chapter 4 Maintenance

and

IPB

45

Figure

19

show the message

VVM

prints

upon

reconstruction

of

data from a failed disc.

Figure

19

VVM

completion message

r

failure:

mvst

(8) command

completed

successfully

If

there is insufficient space in available hot spare disc,

or

VVM is unable to automatically reconstruct the.data from the failed device, it may be possible to manually reconstruct the data. Refer to

Managing

ConvexOS

Operations

Guide

(DSW-031), for

more information

on

disc failure recovery.

Determining

disc

usage

ConvexOS provides a variety utilities for monitoring disc usage. For example, in Figure

20,

df

is used with

grep

to determine

which files are mounted

on

the dulO disc drive.

Figure

20 Determining disc usage with

df I grep

t

df I grep

dulO

/dev/dulOa

45978

39226

2154

95%

/swi

/dev/dulOg

413364

221690

150336

60%

/master

/dev/dulOh

263304

233736

3236

99%

/cadstaff

However,

df

does not show partitions that are used

as

part

of

swap space. In Figure

21

grep

is used to locate entries

in

/etc/fstab

that define the usage of dulO. In this case

an

entry

is

shown for partition b as swap space. When a partition

on

the

failed drive is used for swap space, ConvexOS should

be

taken

to single-user mode before attempting to replace the failed disc.

46

Elite 3 Disc

Drive

Service

Guide

Figure

21

Determining disc usage from /etc/fstab

#

grep

dulO

/etc/fstab

/dev/dulOa

/swi

4.2

rw

13

1

/dev/dulOb

swap

rw

0 0

/dev/dulOg

/master

4.2

rw

13

3

/dev/dulOh

/cadstaff

4.2

rw

2 2

*

Chapter 4 Maintenance

and

IPB

47

Removal and replacement

procedures

Caution

Step 1 Step2

Step3 Step4 Steps

Step 1 Step2

Step3

Step4

Components in the Elite 3 chassis cannot

be

serviced while

the

chassis is installed

in

a peripheral cabinet.

Elite 3 chassis

To

remove

an

Elite 3 chassis from a peripheral cabinet:

Use appropriate electrostatic discharge protections procedures as

described in the Preface, "Electrostatic discharge protection."

As

root,

use

the

/etc/ahutdown

command

to

halt ConvexOS

before

removing

power

from

an

Ellte 3 disc

drive

Lmed

In

the

ConvexOS

ftle

system.

Fallure

to

do

so

may cause a

system

crash

and

loss

of data.

Tum

the power switch

on

the

rear of the Elite 3 chassis off.

Disconnect

IPI cable(s)

and

terminator. Loosen rear alignment/retaining screw. Remove front alignment/ retaining screw.

Remove the Elite 3 chassis from

the

disc drive tray.

Reverse the procedure to install

an

Elite 3 chassis.

Disc

drive

With the Elite 3 chassis

on

a grounded work area,

perform

the

following steps to remove

an

Elite 3 disc

drive

from

the

Elite 3

chassis: Loosen screws

on

the Elite 3 chassis cover

and

remove

the

cover.

Disconnect the

IPI interface cable, operator panel cable,

and

power

supply cable.

Remove the two

upper

mounting screws

on

the

top

of

the

upper

disc mounting bracket.

Remove the two lower mounting screws from the

lower

disc

mounting bracket. Lower bracket mounting screws

are

accessed

through holes

on

the side of the Elite 3 chassis.

Reverse the procedure to install

an

Elite 3 disc drive

in

an

Elite 3

chassis.

48

Elite 3 Disc

Drive

Service

Guide

Step

1 Step2 Step

3

Step4

Step

1

Step

2

Step

3

Step4

Power

supply

With the Elite 3 chassis

on

a grounded

work

area, perform the following steps to remove the power supply from the Elite 3 chassis:

Loosen screws on the Elite 3 chassis cover

and

remove the cover.

Disconnect the

ac

cable from

Jl

on

the power supply.

Disconnect the de power cable from

J2

on

the power supply.

Remove the 4 mounting screws. Mounting screws are accessed through holes on the side of the Elite 3 chassis.

Reverse the procedure to install the power supply in

an

Elite 3

chassis.

Operator panel

With the Elite 3 chassis

on

a grounded

work

area, perform the following steps to remove the operator panel from the Elite 3 chassis:

Loosen screws on the Elite 3 chassis cover

anq

remove the cover. Remove the power supply. Disconnect the operator panel cable from the Elite 3 disc drive.

Make note of the cable routing for subsequent installation as shown

in

Figure

22.

Remove mounting screws

and

operator panel.

Reverse the procedure to install

an

operator panel

in

the Elite 3

chassis.

Chapter 4 Maintenance

and

IPB

49

Figure 22 Elite 3 operator panel cable routing

Step 1

Step2 Step3 Step4

Fan

assembly

AH261010

4/8/93

With the Elite 3 chassis

on a grounded

work

area,

perform

the

following steps to remove the fan from

the

Elite 3 chassis:

Loosen screws

on

the Elite 3 chassis cover

and

remove

the

cover. Remove the power supply. Remove the

power

cable from the fan assembly.

Remove the mounting screws while holding

the

fan.

Reverse the procedure to install a fan in

an

Elite 3 chassis.

50

Elite 3 Disc

Drive

Service

Guide

FigUl'e

23

Air filter access

Air

filter

The air filter can be accessed

by

removing the retaining screws

and

front cover on the Elite 3 chassis.

The air filter should be inspected once a

month

and

cleaned as

required. Figure

23

shows the location of the air filter

and

the front cover retaining screws that are removed to gain access to the air filter.

+12 v

GND

-12

V

AH261001

4/30/93

Chapter 4 Maintenance

and

IPB

51

Illustrated

parts

breakdown

(IPB)

Figure

24

shows the Elite 3 parts.

Figure 24 Illustrated parts breakdown

(IPB)

52

Elite 3 Disc

Drive

Service

Guide

AH261003

4/22/93

Table

13

contains

the

CONVEX

part

numbers

for

the

FR Us

found

in

an

Elite 3 chassis. The CONVEX

part

number

for

the

entire

assembly is

550-002230-200.

Figure 25

IPB

parts list

Item

CONVEX

part

Description

number

1

601-500059-200

Internal

data

cable

2

603-070010-200

de

cable harness

3

603-030042-200

ac cable

harness

4

601-160004-200

Drive to

operator

panel

5

312-000555-001

Air

filter1

6 411-001385-200

Operator

panel

2

7 500-000544-200 Fan assembly

8

200-001046-200

Power

supply

9

204-00026-200 Elite 3 2

HP

3.4 Gbyte

disc

drive

10 603-010029-200

Ground

strap

1

Item 5 in Figure

24

points to the general area where the

air

filter is located.

The

air filter is shown in Figure

23.

2

Item 6 in Figure

24

points to the general area where the operator panel

is

located.

Chapter 4 Maintenance and IPB 53

idcfmt ( 1 D)

man

page

IDCFMT(10)

NAME

CONVEX

Diagnostic

Utilities

Manual

idcfmt

- utility for JDC

disk

formatting,

media

repair,

and

data

editing

SYNOPSIS

idcfmt

-d ccu

port

unit

[options] [commands]

or

id

cf

mt

-d I

dev

I rdun-ctl [options] [commands]

A

IDCFMT(tD)

Idcfmt

is a program

used

to prepare

and

repair

an

IDC

disk

for

use

under

ConvexOS.

There are

two

versions of idcfmt. The first version

runs

on

the

Service Processor

while

ConvexOS is

not

running.

The second version

runs

on

CONVEX

C200/C3200/C3400/C3800

Series machines while ConvexOS

is

running.

Both versions

require

the

user

invoking

the

program

be

the superuser.

Idcfmt

is

invoked

by

typing

its name, a "-d", followed

by

the

CCU

slot

number,

IPI

port

number,

and

drive

unit

number

of the drive to

be

used.

The

ConvexOS

version

allows

a

drive

to

be

specified

through

its control name,

such

as

/dev/rduO-ctl

or

/dev/rdu1-ctl.

A list of

options

may

follow

the

drive

specification,

then

zero

or

more

commands.

If

no

commands

are

given

on

the

command

line,

idcfmt

displays a

prompt

and

expects

the

commands

to

be

entered

from the

standard

input.

OPTIONS

With

the

exception

of

the -v option, the following

options

are

not

normally

required.

These

options

are

intended

for

use

under

rare, exceptional conditions

and

then

only

by

CONVEX personnel. The description for the following

options

make

reference to

various

files

used

by

idcfmt. A description of the layout

of

these files

may

be

found

at

the

end

of

this document.

-t

dkd-999

The

-t

option

overrides

the

automatic identification

of

the

disk

drive

that

idcfmt

normally

performs.

If

the

-t

option

is

not

specified,

idcfmt

queries

the

target

disk

drive

for

its

model

number. This

model

number

is

then

used

as

a search

key

in

the file

/mnt/usr/lib/DB_idc.

Idcfmt

uses the

matching

entry

in

this

file to

Appendix

A

idcfmt

(1D)

man

page

55

determine the geometry of the disk drive.

If

the -t option is specified,

idcfmt

uses

the name provided by the user as the search key into the database. The

model

number obtained from the drive is then verified against the selected entry.

If

the

model number does not match the one in the database, a warning is printed

and

idcfmt proceeds with the database entry selected

by

the user. This option is

useful only in the unlikely event that there are two drives that return the same model number.

-m

manuf

acturers_defects_pathname

The -m option specifies a pathname to a file which contains the manufacturers defect data. This option is useful in the event that

an

otherwise usable

disk

drive

has

an

unreadable defect map. The user may then manually enter the defect

data

from the

paper

copy of the defects supplied with each drive into

an

asci file

and use the -m option to specify the location of the file. Normally, idcfmt will read the defect

map

from the drive anytime it requires knowledge of the defect

data.

For information

on

the format of the manufacturers defect file refer to the FILES

section within this manpage.

-g grown_defects_pathname The -g option is used to specify a list of sectors that are

known

to

be

bad,

but

do

not appear in the manufacturers defect list.

It

should be noted that the

-g

option is not the normal way to handle defects that occur after initial formatting (refer to the description of the

"slip" command within this document). During initial

formatting, idcfmt assumes there are no grown defects unless the

user

specifies

the -g option.

-v volume_name

IDC disks contain a place for a volume name for each disk. The -v option allows

the user to specify what name will be

put

on

the disk

during

initial formatting.

If

the -v option is not used, a volume name of ''NONE" is used.

Idcfmt

does

not

use the volume name for anything other than initializing the name field

during

formatting. The maximum length of the volume name is 32 characters.

-p pattern_file

-y

Normally, during initial formatting, idcfmt pattern tests the drive using

the

set

of patterns specified in the database. The normal patterns

may

be

overridden

by

putting a set of patterns into ale,

and

using the -p option to tell

idcfmt

where

the

file is located.

The -y option forces the answer

to

all

yes/no

prompts to be treated as

if

the

user

has responded with a "yes". The initial release of idcfmt has only one

of

these

prompts. Refer to the following description of the

FORMAT command. This

option is useful if the user wants to format a drive in background,

and

the user

does not wish to answer the

yes/no

prompt interactively. The

prompt

is there for

the

user's

protection. Think carefully before using the -y option.

56

Elite 3 Disc

Drive

Seroice

Guide

COMMANDS

The following commands perform various functions

on

IDC disks. Each

command

is

listed along

with

any

required arguments specific to that command. Some

commands

require exclusive access to the disk. The description for each command lists whether

or

not the command requires exclusive access. The version

of

idcfmt

that

runs

on

the

Service Processor automatically assumes that it has exclusive access to

the

disk. The

version of

idcfmt

that

runs

under

the ConvexOS coordinates exclusive access to the

disk

through a diagnostic ioctl() call to the ConvexOS disk driver. Exclusive access to a

disk

is

denied

if

the disk is currently

in

use

under

the ConvexOS file system,

or

if

any

process

has

opened

any

of the block

or

character special devices associated

with

the

disk.

The description for some of the commands assume that the

user

has

a basic knowledge

of the layout of

an

IDC disk. A description of this layout follows below.

format

Formats the drive

if

it

appears that the drive

has

not

been

previously formatted.

If

the drive

appears

to be partially formatted, the format resumes

at

the

appropriate point. This command will destroy

any

on

the

disk. Exclusive access to the disk is required before this command can execute. This command tries to determine whether

or

not the drive is formatted

by

attempting

to read the topology

map

in the topology area.

If

successful in reading the

topology map, the command prints a message

and

proceeds to

the

format

verification step. The user

may

override this behavior

by

using the FORMAT

command which is the next command described.

If

the

topology

map

is unreadaple, the program assumes that the drive has not been completely formatted.

It

then proceeds to check for a partial format of

the

disk

by

looking for

checkpoint

data

in the topology and diagnostic areas.

If

a valid set

of

checkpoint

data

is found, the command resumes formatting

at

the point specified

by

the

checkpoint data.

If

no

checkpoint

data

is found, the command assumes that the drive

has

never been

formatted

and

proceeds to format the drive accordingly. The steps required to

format a drive are:

o) Identify the type of drive being formatted

by

performing a read configuration

command to the drive. The data returned by the read configuration

command

is

used

to search the

data

base file "DB_idc" for a match. The

data

base file contains

information

about

the disk drive such as number of cylinders,

number

of

tracks, encoding schemes, etc. The user may override this automatic identification of the drive

by

using the -t option

on

the command line.

o) Read the manufacturers defect

data

from the drive. The

user

may

override this

step

by

using

the

-m

option described previously in the OPTIONS section.

o) Assume that there are no grown defects for the drive. The

user

may

override

this step

by

using the -g option described previously in the OPTIONS section.

o) Format the topology area with

no

spare sectors. The topology area is

then

pattern tested with a set of patterns that are specified in the

data

base file

"DB _idc". The user

may

override the normal patterns

by

specifying

the

-p

option

Appendix A

idcfmt

(1D)

man

page

57

on

the

command line. Once pattern testing is complete,

the

topology area

is

re-formatted

with

the

number of spare sectors specified in the

data

base file

"DB_idc". A detailed description of pattern testing can

be

found

in:

the

FILES

section of this manpage. o) Format the user, diagnostic, and spare areas

of

the disk with

no

spares.

The

first checkpoint of

the

format is written this step is complete.

o) Pattern test

the user, diagnostic,

and

spare areas of

the

disk

with

the

patterns

specified

in

the

data

base file "DB_idc". The

user

may

override

the

normal

patterns

by

use

of

th~

-p

option described previously

in

the

OPTIONS section.

The process of pattern testing consists of writing each of

the specified

patterns

to

the

area

under

test,

and

then reading each pattern to look for errors

that

might

be

attributed to disk media problems. These include

data

Error Correcting

Code

(ECC) errors, header Cyclical Redundancy Character (CRC) errors,

and

missing

sector sync byte errors.

If

any of these errors are encountered,

the

manufacturers

defect list is searched to see if the cylinder, track

and

sector where

the

error

occurred is already listed.

If

the cylinder, track,

and

sector is

not

in

the

manufacturers defect list, the grown defect list is

then

searched.

If

the

cylinder,

track,

and

sector where the error occurred is

not

found

in

either defect list,

the

location is

added

to

the

grown defect list. After each

pattern

is completed, a

checkpoint

is

written to the disk.

o) Format

the

user, diagnostic,

and

spare areas

of

the disk, reserving a

number

of

sectors

at

the

end

of each cylinder for

use

as

spare sectors.

Any

sector

which

is

listed

in

either

the

manufacturers

or

grown defect lists is re-allocated from

this

pool of spare sectors.

If

the number of defective sectors

in

a cylinder exceeds

the

number

of spare sectors allowed for each cylinder,

the

disk

is

not

usable

under

ConvexOS. A checkpoint is written to the

disk

after this step.

o) Initialize the topology area by writing several copies

of

the topology data. This

includes the topology area map, the defect lists for

the

disk, a copy of

the

disks

geometries, the ConvexOS label, etc. The contents of

the

topology area

is

described below. o) Clear

the

diagnostic cylinder.

o) Verify that the drive received a valid format. Refer to

the

verify

command

below for a description of what steps are taken to verify

that a drive

is

formatted

correctly.

The

disk

in

now

formatted

and

ready for use.

FORMAT

Formats the drive regardless of its current state. This

command

is

the

same

as

the

format command listed above, except that

if

the

disk

is

determined

to

have

a

valid, complete disk format, the user is

prompted

as

to whether

or

not

to

continue.

If

the user answers in the affirmative,

the

format operation proceeds

as

if

the

drive

had

never been formatted.

If

no

valid format is found

on

the

drive,

this command does not look for checkpoint data. The format starts from

the

58

Elite

3

Disc

Drive

Service

Guide

verify

beginning. All the exclusive access requirements of the normal format command also apply to this command.

This command verifies that the disk is properly formatted. This automatically takes place when the drive is initially formatted. This command does

not

write

to the drive

and

may be used

on

a drive with user data. Exclusive access to the disk is not required. The following steps are taken to verify that the drive is formatted:

o) Reads one copy of all the data on the topology cylinder. Verifies that the

checksums

and

magic numbers are correct.

o) Reads the remaining copies of the topology

data

and

verifies that they match

the

data

from the first step.

o) Reads all logical blocks

on

the disk

and

verifies that there are

no

header CRC

or

data

ECC

errors.

If

any errors are found they are listed. The user

may

then

attempt to

fix

the error by using the slip command described below.

o) Reads all the sector headers from the disk

and

verifies that there is a one to one

correspondence of entries

in

the defect lists with sectors that are slipped.

o) Verifies that each entry

in

the defect lists point to a sector that

has

been slipped.

slip cylinder track sector

This command is used to repair header CRC or data ECC errors.

If

a sector

in

the

user

data

area of the disk is to be slipped, it can be

done

without having to have

exclusive access to the disk. Sectors in the topology,

and

diagnostic areas require

exclusive access to the disk.

The default radix used for cylinder track sector is decimal.

If

hexadecimal values

are required, prefix the values with

Ox.

If

a sector in the user area of the disk is to be slipped, idcfmt issues a slip sector

command to the

IDC. This command is

an

atomic (indivisible) operation

implemented

in

the IDC software in a fashion that guarantees

data

integrity at

all times, even in the event of a power failure

during

the operation.

If

the slip

operation is interrupted, the operation automatically completes the next time

the

IDC software accesses the disk.

The preferred procedure before slipping a sector is to first

unmount

the file

system containing the sector to be slipped.

If

the sector to be slipped is in the topology or diagnostic areas of the disk, the

following steps are taken:

o) The topology data is read from the topology area,

and

the sector to

be

slipped

is

added

to the memory copy of the grown defect list.

o) A flag is written to the topology area that keeps the idc software from allowing normal access to user data.

Appendix

A

idcfmt

(1D)

man

page

59

o)

If

the sector to

be

slipped is not

in

the topology area, then the cylinder

that contains the sector is reformatted, causing the specified sector to be reallocated from the pool of spare sectors at the end of the cylinder.

o)

If

the sector to be slipped is in the topology area, then the diagnostic area is

reformatted in order to ensure that the topology area

data

can

be

written to

the

disk. o)

If

the sector to be slipped is in the topology area, a checkpoint containing

the

topology data is written to the diagnostic area.

o) If the sector to be slipped is in the topology area, the topology area is

reformatted

and

the data is copied from the diagnostic area back to the topology

area. The diagnostic area is then reformatted to remove the checkpoint data. o)

If

the sector to be slipped is not

in

the topology area, the defect lists

in

the

topology area are updated to show the new defect data.

o) The flag which keeps the IDC microcode from allowing access to

user

data

is

removed.

If

a slip

of

a sector in the topology

or

diagnostic areas is interrupted, the

user

data

on

the disk is not accessible until the slip is completed. The completing

of

the slip

is accomplished

by

re-issuing the slip command with the same parameters

as

before.

If

the parameters differ after

an

interruption,

idcfmt

prints

out

a message

stating

what

sector was being slipped before the interruption.

unslip

cylinder track sector

This command is the inverse of the slip command described above.

It

is

provided

in

the event that user mistyped the cylinder track

and

sector

number

during

a

slip command,

or

to allow the user to slip

and

then unslip a sector

in

an

attempt

to

fix

a header CRC

or

data

ECC

error. As with the slip command, the

unslip

command does not require exclusive access to the disk

if

the sector to

be

unslipped is in the user data area.

If

the sector to be unslipped is in the topology

or

diagnostic areas, exclusive access to the disk is required. The algorithm

used

by

the unslip command is identical to the one used for the slip command, except

that the specified sector is removed from the defect list instead of being

added

to

it.

WARNING!!!!

UNSLIPING A SECTOR FOR

ANY

REASON OTHER THAN THOSE

DESCRIBED

ABOVE

MAY

RESULT

IN CATASTROPHIC AND

UNRECOVERABLE

LOSS

OF

DATA

ON

THE ENTIRE

DISK.

60

Elite 3 Disc

Drive

Service

Guide

EXAMPLES

idcfmt -d 2 0 1 format

THIS

IS

AN

EXAMPLE

OF HOW

TO

FORMAT

99%

OF

ALL

DRIVES. Formats

drive

1,

on

port

0,

of ccu 2 if the drive has never been formatted.

If

a partial

format is found, the format resumes

at

the appropriate point, with the

manufacturers

and

grown defect lists that were in effect

when

the format was

interrupted.

If

the drive has no valid format

on

it, then

the

program reads the

manufacturers defect data from the drive. The examples that follow are for drives with unreadable manufacturers defect

lists or other peculiarities.

If

a drive develops additional defects after it has been

formatted, they may be fixed with the slip command described above.

idcfmt-d

/dev/rdul-ctl

format

Formats the drive pointed to by

/dev/rdu1-ctl.

Everything else is the same as

in

the

first example.

idcfmt -d 2

0 1 -m foo format

Formats the same as the first example, but the user specified defect list

in

the file "foo" is used instead of reading the manufacturers supplied list

on

the disk

drive.

idcfmt -d 2

0 1

FORMAT

Forces a format operation to occur regardless of the current status of

the

drive. The manufacturers defect cylinder is read

and

no grown defect

data

is assumed.

idcfmt -d 2

0 1 -m foo FORMAT

Formats the same as the previous example,

but

the user specified defect list in the

file "foo" is used instead of reading the manufacturers supplied list

on

the disk

drive.

idcfmt -d 2 0 1 verify

Causes a verification of the format of the disk. The steps taken to verify the disk are described previously.

idcfmt -d 2

0 1 slip

10

5 6

Causes sector 6 of track 5 on cylinder

10

to be re-allocated from the pool of spare

sectors

at

the end of the disk.

idcfmt -d 2

0 1 unslip

10

5 6

This command undoes the effects of the command described

in

the previous

example.

PLEASE read the description of the slip

and

unslip command before

using the unslip command.

Appendix A idcfmt

(1D)

man

page

61

LAYOUT

OF

AN

IDC

DISK

An IDC disk has five distinct areas. They are:

1)

The

topology

area

2)

The

user

data

area

3)

The

diagnostic

area

4) A

spare

area

5)

The

manufacturers

supplied

defect

data

Each area is immediately adjacent to each other and they occur in the order listed above.

Each area must begin and end on a cylinder boundary. The size of each area is determined

by

idcfmt during the initial format of the disk from the

data

contained

in

the

database. Refer to the section on

FILES

within this

man

page for more information.

The topology area contains various types of information about the disk. Each section

in

the topology area is recorded in triplicate. The topology area is recorded

in

the default

geometry specified

by

the manufacturer. The other areas are recorded

in

a format

specified

by

CONVEX.

The various sections

in

the topology area are:

a) A map of the rest of the topology area (always located

at

logical block 0)

b) A copy of the manufacturers defect data. This

data

is used to avoid reading

or

writing sectors with defects in them. The defect

data

is also

used

in calculating

what the sector header each sector contains. The disk is unreadable if this

data

is

not present.

c)

The grown defect data. This list of defects performs the same function

as

the list

in

"b" above, except that it contains a record of defects that were

not

listed

in

the

manufacturers list.

d) A copy of the CONVEX specified disk geometry used in the other areas of the

disk.

e) A copy of the manufacturers default geometry.

f)

The ConvexOS disk label. This disk label contains the type of the disk, e.g.

DKD-502, the volume name, e.g. Volume_123, the

date

that the drive

was

formatted, the date the ConvexOS disk label was last changed,

and

a list of

the

logical partitions defined for the drive.

g) Two scratch areas used by the

IDC

software to keep track of the state of a sector

slip operation in the user data area. These areas are interrogated

when

the JDC

software is first booted.

If

the scratch areas indicate that a slip

or

unslip of a

sector in the user area was in progress

but

was interrupted, then the operation

automatically resumes at the interrupted point.

H the scratch areas indicate that

an

exclusive access diagnostic operation was in progress, the IDC software will

not allow access to the user data area until the operation is finished.

62

Elite 3 Disc

Drive

Service

Guide

The

user

area is where user

data

is read

and

written

under

ConvexOS. Its size

is

the entire area of the disk, less the area allocated to

the

topology, diagnostic,

spare,

and

manufacturers defect area.

The diagnostic area is a scratch area

on

the

disk

that

is

used

for

disk

diagnostics,

and

to temporarily hold

user

data

while a sector

in

the

user

area is being

slipped

or

unslipped.

The

spare

area is

an

area

that

is not currently used,

but

is reserved for future

use

by

CONVEX.

The manufacturers defect area is

the

area

on

which

the

disk

drive

manufacturer

records

the

defects for the drive. A copy of this

data

is

kept

in

the

topology area.

This area is never written

by

CONVEX,

thus

the original defect

data

is

always

present.

ERROR CODES

Idcfmt

produces errors messages in the following format

Error:

(Ox123)

message

: Additional infonnation

where

"Ox123"

and

"message" are

one

of the codes

and

messages listed below. The

"Additional information" is other useful information such

as

the

cylinder,

and

track,

sector where

an

error occurred, the name of the disk command

that

failed,

the

name

of

ale that

idcfmt

could

not

open, etc. The information

on

the

"Error:" line

is

intended

to

help the

user

diagnose the problem. The "Additional information" lines are elaborations

on

the

original error message. These lines also contain information to

help

the

CONVEX

development staff assist

in

isolating problems

in

both

hardware

and

software.

Code Message

Oxl Error

in

byte

count.

A read

or

write operation was initiated

but

not

all of

the

data

requested

was

transferred. This will

not

normally

happen

without one

of

the

other

errors listed

below.

Ox3

Various CMI (Common Message Interface) error messages This error code will

be

followed

by

additional lines

that

describe the

type

of

operation in progress

when

an

error occurred,

the

number

of

times

the

operation

was

retried,

and

any

extended status available.

Ox4

Error from fwrite Idcfmt

tried to write to ale

on

the SPU

or

the CONVEX file system

and

an

error

occurred. This is most likely caused

by

ale system being full. This error code

has

nothing to

do

with

the disk

under

test

OxS

Non

numeric value found

when

expecting a

number

It is

most

likely that a

parameter

was mistyped from

the

command

line

or

that

one of

the

files described

in

the

FILES

section contains

an

error.

Appendix A idcfmt

(1D)

man

page

63

Ox8

spu

window

ioctl error

Idcfmt

received

an

error when trying to get SPU UNIX to allocate

or

remap

a

service processor window. This may be caused

by

too

many

processes allocating

too

many

windows

on

the SPU.

Ox9

mbs

error

An

error was received from the Message Based System

(MBS).

This

is

usually

indicative of a hardware malfunction

on

the SPU,

the

lOC,

the

PIA/Pl2,

or

system memory. The user

may

wish to consult the system error

log

to see

if

it

contains more information about the error. See

the

discussion

on

FILES

within

this document.

Oxa

timeout waiting for message Idcfmt

sent a message to the IDC,

but

the JDC failed to respond. This

can

occur

if the JDC encounters a situation that it is not expecting,

arid

it

panics.

It

is

possible that this message can result

when

running

the

online version

and

a

loaded, slow system causes a timeout to occur.

Oxb

main memory too small. Requires

4MB + 2MB

per

disk drive

For operation

on

the SPU,

iddmt

requires a

4MB

overhead area

plus

2MB

per

copy of idcfmt.

If

there are many copies of

idcfmt

in

execution,

and

the

memory

configuration is very small, this error may occur.

Oxc

must

set geometry before this operation

This indicates

an

internal software error

in

idcfmt.

Oxd

error

on

open

of file

Idcfmt

attempted to open a SPU UNIX

or

ConvexOS file,

and

received

and

error.

The file name is displayed

on

a subsequent line. This error can occur if

the

user

mistyped ale name, if the specified file is not present,

or

if

the

SPU

or

ConvexOS

denied idcfmt access to the file.

Oxe

parsedb() error Idcfmt

attempted to extract information from

the

file

/mnt/boot_db

on

the

SPU

and

the

requested information was missing. The

name

of

the

missing piece

of

information is also listed.

Be

sure that .diaginit(lD)

has

successfully completed.

Ox11

attempt to position outside

of

defect area

boundary

The defect list

on

the selected disk drive is unreadable.

Ox12

manufacturers defect list not in CONVEX compatible format A defect list was encountered that is not

in

the format

needed

by

CONVEX.

Ox13

cannot malloc() memory Idcfmt

requested memory from SPU UNIX

or

the ConvexOS

and

received

an

error. This is most likely caused by too many processes

in

execution

at

the

same

time.

64

Elite

3

Disc

Drive

Service

Guide

Ox14,0x15

Set

#1

missing from defect map

Last set in

map

does not have proper value

in

flag field These errors indicate that

the defect

map

on the drive has been partially destroyed,

and

is not usable.

Ox16

must read defects before this operation

This is

an

internal software error in idcfmt. Contact the CONVEX development

staff to report a software error.

Ox17

error from fseek() Idcfmt tried to re-position ale and received

an

error.

Ox18

defect file differs

on

pass2 during reading mfg defects.

Ox19

data miscompare Idcfmt

was

comparing data read from the disk with

known

good

data

and

it

encountered a discrepancy. The address of the

data

along with the expected

and

actual data is displayed.

Oxlc

topology area too small The topology area is too small to hold the required data. Make sure that the

correct diagnostic database is installed. This error is most likely caused

by

attempting to format a drive with the wrong parameter specified

in

the -t option

on

the command line. Oxtd magic number incorrect Oxte topology check

sum

incorrect

These errors indicate corruption of the topology area.

Ox2e

Illegal ecc interleave

An

unknown

ECC interleave was specified

in

the database file. Make sure the

correct version of system diagnostics and diagnostic database are installed.

Ox2f

error from fread ldcfmt

tried to read ale

on

the SPU UNIX or ConvexOS file system

and

received

an

error.

Ox30

Bad format for manufacturers defect

The user specified the -m option and one of the entries

in

the file is incorrect. The

line

number

of the entry in error is also displayed.

Ox31

Manufacturers defect overlap. List not usable The manufacturers defect data contains overlapping defect entries. This is not a

normal occurrence

and

idcfmt can not handle this case.

Ox37

Cmd

line DKD entry not unique in

DB

file

Appendix A idcfmt

(1D)

man

page

65

The user specified -t

on

the command line, and the disk type specified has more

than one entry in the database

file.

Make sure that the correct diagnostic database

is installed.

Ox38

Cmd

line DKD entry not found in

DB

file

The user specified -t on the command line, and the disk type specified

was

not

found

in

the database. Make sure that the correct diagnostic database

is

installed.

Ox3b

Crc incorrect

on

checkpoint data

An media error occurred when trying to read checkpoint data off the drive.

If

the

operation

in

progress was a format command, the user can retry the operation

by

using the

FORMAT

variant of the command.

Ox3c,

Ox3dVersion number incorrect on checkpoint data Magic number incorrect

on

checkpoint data

Checkpoint data was encountered that was written

by

an

incompatible version

of idcfmt. This error is not recoverable. Either reformat the disk using

FORMAT

or

ensure that the disk was not previously interrupted while attempting to

format with another version of idcfmt.

Ox3e

Sequence number incorrect on checkpoint data

An media error occurred when trying to read checkpoint data off the drive.

If

the

operation

in

progress was a format command, the user can retry the operation

by

using the

FORMAT

variant of the command.

Ox3f

Ran

out

of room for checkpoint data

The diagnostic or topology are is too small to hold the checkpoint data. This is a fatal error

and

in most cases the drive is defective.

Ox40

Unknown checkpoint state Checkpoint data was encountered that was written by

an

incompatible version

of idcfmt. This error is not recoverable. This error usually indicates that

the idcfmt software contains errors. Please contact the CONVEX development staff to report this error.

Ox43

topology version number incorrect A topology area was encountered with a format incompatible with this version

of idcfmt. This error can indicate that the idcfmt software contains

an

error

or

that the drive was previously formatted with a different version of idcfmt.

Ox45

Non-hex digit in pattern The user specified a alternate pattern

file

that contains a non-hex digit

in

the

field. See the section on

FILES

within this document.

Ox46

Sector header incorrect A sector header was found with good

CRC,

but

incorrect data. The expected

and

actual header are displayed. This error message usually indicates a faulty drive.

66

Elite 3 Disc

Drive

Service

Guide

Ox47

Header ere incorrect

A sector header

was encountered with bad CRC. This error message usually

indicates a fixable media problem.

Ox48

non-media related cmi error

An

error was encountered during pattern testing of the drive that does

not appear to be related to media errors. The format of the message is the same as for error code

Ox3.

Ox49

can't read ucode fields to resume pending diagnostic operation

The IDC software reports that a diagnostic operation was

in

progress,

but

the

ucode scratch areas cannot be read

in

order to resume the operation.

Ox4a

the following operation must be completed.

An

exclusive access diagnostic operation was interrupted,

and

the

user

tried to

start a different operation without completing the first one.

Ox4b

One

or

more sections in the topology area are unreadable

The user attempted

an

operation which requires that the topology area

be

read,

but

there is at least one topology section which is totally unreadable.

Ox4c

verification after write of diagnostic ucode temp field failed

The user attempted

an

operation which requires idcfmt to write a value to the

topology area scratch sections,

but

the read after write failed to verify.

Idcfmt

will attempt to clear

out

the scratch areas and abort the operation.

Ox4d

disk address

out

of bound

The user attempted to specify a disk address that is outside the allowed range for the drive.

Ox4f

drive model number not unique in data base file Idcfmt encountered more than one record

in

the database with the same model

number. The user may override idcfmt

by

specifying the -t option.

OxSO

drive model number not found in data base file

Idcfmt encountered a a disk model number that it could

not

found

in

the

data

base. Make sure that the correct diagnostic database is installed.

Ox52

can't find

MEMSTART

in

/mnt/boot_db.

Make sure .diagint was executed

Ox53

can't find PCM in

/mnt/boot_db.

Make sure .diagint

was

executed. The file

/mnt/boot_db

appears to

be

incorrect.

Make sure that .diagint(ld) completed execution.

Ox54

can't acquire

spu

unix lock

Ox55

can't get

open

count of

spu

unix lock

Appendix

A

idcfmt<1D)

man

page

67

Ox56

can't set

spu

unix variable

The wrong version of

SPU

UNIX

appears to be installed. Idcfmt

spu

version for

CONVEX C200/C3200/C3400 requires

SPU

UNIX

5.2

or

greater.

Ox57

initall failed to complete successfully Idcfmt was attempting to initialize the system

and

initall(lD) reported

an

error.

The output from initall can

be found in

/tmp/idcfmt.initall.

Ox58

can't get spu unix variable The wrong version of

SPU

UNIX

appears to be installed.

Iddmt

spu version for

CONVEX C200/C3200/C3400 requires

SPU

UNIX 5.2

or

greater.

Ox59

unknown main memory state. Make sure idcfmt is the only thing running There appears to be a program running on the

SPU that is using main memory

in a fashion that is not compatible with

idcfmt.

OxSa

can't release spu unix lock The wrong version of

SPU

UNIX

appears to be installed. Idcfmt spu version for

CONVEX C200/C3200/C3400 requires SPU UNIX

5.2

or

greater.

OxSb

the specified ccu is not logged as being an IDC in

/mnt/boot_db

The user specified a ccu number that is not listed

and

being

an

IDC

in

/mnt/boot_db.

Verify that then ccu number entered is correct,

and

then verify

that

.diaginit(td) finished execution.

OxSc

Grown defect entry in wrong format. Expecting cyl trk sec

Please read the section on

FILES

in this document.

OxSd

Cannot

open/create

device control table. Verify that there is room in

/tmp

OxSe

idcfmt already appears to be running

on

the specified device

OxSf

unable to wire memory for program use

Ox60

the specified drive is in use

Ox61

unable to take drive offline

Ox62

unable to re-attach drive

Ox63

no

such device

Ox64

unable to determine status of drive from the du-driver

Ox65

operation interrupted by SIGINT

Ox66

command line

dkd

number does not match ioconfig

Ox67

unable to obtain unit name from du_driver

or

unable to obtain driver handle from du_driver

Ox69

ConvexOS is running. Execute idcfmt from a system terminal

68

Elite

3

Disc

Drive

Service

Guide

FILES

Ox6a

Slipped sector

not

written

to

top

cyl defect list

Ox6b

Attempt

to

write

topology cylinder

grown

defects section failed

Ox6c

None

of

the

enabled

drives

were successfully initialized

Format

of

manufacturers

defect

file

When

the

user

specifies

the

-m

option

on

the

command

line,

idcfmt

expects

the

specified

file

pathname

to

point

to

an

ascii file

with

the

following

data.

Cylinder

track byte_count_from_index defect_length_in_bits

There

should

be

one

line

in

the

file for each defect entry. All

numbers

are

in

decimal.

Due

the

peculiarity

of

the

way

defect

data

is

encoded

for

multiple

heads

per

track

disk

drives,

it is extremely difficult to

guess

what

sector

an

error

will

occur

for a

given

byte

count

from

index.

The

-m

option

is

provided

so

that

the

user

can

recover defect

data

from

a

paper

copy.

If

the

user

wishes

to force a sector to

be

treated

as

defective

during

the

initial

format

of

the

drive,

he

should

make

an

entry

in

the

grown

defect table.

Format

of

grown

defect

file

When

the

user

specifies

the

-g

option

on

the

command

line,

idcfmt

expects

the

argument

after

the

-g

to

point

to ale

in

the

following format:

Cylinder

track sector

There

should

be

one

line

in

the

file for each defective sector. All

numbers

are

in

decimal.

Format

of

pattern

The

-p

option

on

the

command

line is

used

to

change

the

default

set

of

patterns

used

during

pattern

testing

of a disk

drive. The

argument

after

the

-p

flag

should

point

to

ale

that

contains a

set

of hex

digits

that

will

be

used

to

pattern

test

the

drive.

Thus,

ale

containing

I 4 c 9abdef01

will

cause

four

patterns

to

be

used

for

pattern

testing.

The

first

pattern

will fill

the

disk

with

all

Oxllllllll,

the

second

with

all Ox44444444,

the

third

with

Oxcccccccc,

and

finally,

Ox9abcdef01.

It

should

be

noted

that

entering

a single hex

digit

of

1 is

the

same

as

entering

it

as

11,

1111,

and

11111111.

It

is also possible to

enter a pattern

of

23456789a

and

then

the

pattern

generated

will

be

23456789

a2345678

9a234567

89a23456

789a2345

...

When

the

pattern

is specified

with

a length of 1, 2, 4

or

8 hex digits,

the

initialization

of

the

data

buffer

is

much

faster. Specifying a

pattern

as

111

instead

of

1, 11,

1111,

or

11111111

results

in

over

16x

the

amount

of

time

needed

to initialize

the

buffer.

Appendix A idcfmt

(1D)

man

page

69

Default

pattern

files

Entries in DB_idc specify the default pattern. The 4 pattern files are RLL_1_7, RLL_1_7.2hp, RLL_2_7

and

RLL_2_7.2hp. The pattern files are

kept

in

/usr/lib

in

the

ConvexOS file system,

and

in

/mnt/usr/lib

on

the Service Processor file system.

Default

pattern

files contents RLL_1_7: RLL_1_7.2hp : RLL_2_7: RLL_2_7.2hp :

fl

7 6

Sf

c7 e e3

7c

fl

7 6

Sf

c7 e e3

7c 0 c 1 4 9abcdef0 5 d 789abcde 0 c 1 4 9a9abcbcdedef0f0 5 d 78789a9abcbcdede

Output

from prtlog

during

idcfmt

When idcfmt executes,

it

may

cause error conditions that cause

the

~DC

software to

generate messages. These messages are normally

put

in/mnt/errlog

on

the

Service

Processor. Format

of

DB_idc

Below is a sample of idcfmt' s database. This file is normally kept

in

/usr/lib/DB

_idc

in

the

ConvexOS file system,

and

in

/mnt/usr/lib/DB_idc

on

the Service Processor file system.

The numbers

in

this file have been carefully chosen to yield optimal performance

and

reliability. These numbers should not be changed

by

the user. Changing these

numbers

and

formatting a drive will produce a disk drive that is

not

usable

under

the

ConvexOS.

#

DB

idc -!PI

DISK

DRIVE

PARAMETER

FILE

*

i

Drive

Name

Description

*

----------

-----------

DKD-501

Imprimis(CDC)

97209 Se

1.2

GB.

IPI-2

DKD-502

Imprimis(CDC)

97229

Se-512

1.153

GB.

IPI-2

(2

head

parallel)

DKD-503

Imprimis(CDC)

Se 7

3.0

GB.

IPI-2

(Serial)

*

DKD-504

Imprimis(CDC)

Se 7

3.0

GB.

IPI-2

(2

head

parallel)

DKD-505

Seagate

Elite3

3.3

GB

Enhanced

IPI-2

(2

head

parallel)

Zone

Bit

Recording

#----------------------------

Regular

IPI-2

drive

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

*

a b c d e f g h i j k 1 m n o p q r s

*

-------

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

* DKD-501

CDC

MPI

EMD5

1635 50400 P 2048 15 1 23 9 1 1 1 1 12 10 20 1 10

*

* t u v w

*

-------

# DKD-501

BYTE

3000000

usr/lib/idc_pat_rll_2-7

*

#--------------------------- Zone

bit

recording

drive

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

*

#

a'

b'

c'

d'

e'

f'

g'

h'

i'

j'

k'

l'

m'

n'

o'

p'

q'

*

-------

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

--

70

Elite 3 Disc

Drive

Service

Guide

# DKD-505

SEAG2HD

E3L2

18

2738

2048

10 2 15

2 1 1 1

12

0 0 2

20

#

r'

SI

t I U I

#

-------

# DKD-505

BYTE

16777200

usr/lib/RLL_l_7.2hp

# # #

# #

v'

w'

x'

y'

0

DKD-505

0

z'

0'

l'

200

72799

p

62

#==========================================================================

1

DKD-501

CDC

MPI

EMD5

1635

50400 p 2048

15 1 23

9

1 1 1 1

12

10

20 1 10

-

2 DKD-501

BYTE

3000000

usr/lib/RLL

2 7

1

DKD-502

CDC

MPI

S5L2

1635

50400 p 2048

7

2

45

15

1

1 1 1

12

14

34

2

10

-

2 DKD-502

BYTE

6000000

usr/lib/RLL

2

7.2hp

1

DKD-501 SEAGMPI

EMD5

1635

50400 p 2048

15 1 23

9 1 1 1

1

12

10

20

1

10

-

2 DKD-501

BYTE

3000000

usr/lib/RLL

2 7

1

DKD-502 SEAGMPI

S5L2

1635

50400 p 2048

7

2

45

15

1 1 1

1

12

14

34

2

10

2 DKD-502

BYTE

6000000

usr/lib/RLL

2

7.2hp

1 DKD-503

SEAGlHD

S7Kl

2655

63840

p

2048

19 1 29

15

1 1 1 1

12

11

40 1 20

-

2 DKD-503

BYTE

4644360

usr/lib/RLL

1

7

1 DKD-504

SEAG2HD

S7L2

2655

63840 p 2048

9 2

56

15

1 1 1 1

12

2

40 2 20

2 DKD-504

BYTE

9288720

usr/lib/RLL_1_7.2hp

#==========================================================================

3 DKD-505

SEAG2HD

E3L2

18

2738

2048

10 2 15

2 1 1 1

12

0 0 2

20

4 DKD-505

BYTE

16777200

usr/lib/RLL_l_7.2hp

5 0 DKD-505 0

200

72799

P 62

5 5

5 5 5 5 5

5 5

5

1 DKD-505 2 DKD-505 3

DKD-505 4 DKD-505 5 DKD-505 6 DKD-505 7 DKD-505 8 DKD-505 9 DKD-505 10

DKD-505

11

DKD-505

12 DKD-505

13

DKD-505

201

417

71399

p

61

418

628

69999

p

60

629

810

68599

p

59

811

973

67199 p 57

974

1131

65799

p

56

1132 1288

64399 p 55

1289 1425

62999 p 54

1426 1550

61599 p 53

1551

1658

1659

1769

1770

1871

1872

1968

1969

2095

60199

p

58799

57399 55999 54599

51

p

50

p

49

p

48

p 47

14 DKD-505

2096

2299

53199

5 p

45

5

15

DKD-505

2300

2424

51799

P 44

5

16

DKD-505

2425 2561

50399

P

43

5

17

DKD-505

2562

2737

48999

P 42

#===========================================================================

#

LEGEND

for

regular

IPI-2

drives

# a -

drive

name -Must

be

an

unique

DKD-NNN

number.

# b -

mfg

id/model

number

#

mfg

id

- 4

bytes

obtained

by

read

cog

of

drive,

eg.

"CDC_"

#

model

number

- 8

bytes

obtained

by

read

cog

of

drive,

eg.

"MPI

S5L2"

#

Appendix A idcfmt

(1D)

man

page

71

#

Note:

see

product

speccation

manual

of

IPI-2

drive

for

these

info.

#

Replace

any

spaces

in

the

field

with

underscore

characters.

#

# c -

Total

number

of

cylinders

per

unit

# d - Number

of

bytes

per

track

(see

e)

#

# e - P - (

deld

is

the

physical

bytes

per

track

)

# L - (

deld

is

the

logical

bytes

per

track)

# The

logical

bytes

per

track

is

the

physical

bytes

per

track

multiplied

#

by

the

number

of

heads

per

logical

track.

eg.

DKD-502

entry

can

have

an

#

alternate

speccation

of

'100800'

for

the d field

and

'L'

for

the

#

eeld.

# # f

- Number

of

data

bytes

per

sector

#

g - Number

of

logical

tracks

per

cylinder

#

h

- Number

of

heads

per

logical

track

#

i -

Number

of

sectors

per

logical

track

#

j - Number

of

spare

sectors

per

user

cylinder

#

k

- Number

of

cylinders

reserved

for

diagnostic

use

#

l

- Number

of

cylinders

reserved

for

manufacturers

defect

map

# m -

Number

of

cylinders

reserved

for

describing

drive

topology

# n

- Number

of

cylinders

#

# o -

Header

size

# p -

Header

gap

# q -

Data

gap

reserved

for

spares.

# r -

Ecc

interleave

(options

are

O,

1, 2 or

4)

# s - Number

of

spare

sectors

per

topology

cylinder

# t -

This

should

have

the

same

DKD

name

as

line

1

of

the

entry.

# u -

Data

Interleave,

applicable

only

to

parallel

heads

(BIT,

BYTE,

WORD)

# v -

Drive

data

transfer

rate

in

bytes

# w -

pathname

of

ascii

hex

patternle.

If

the

pathname

starts

#

with a /,

it

will

be

used

as

is.

If

it

does

not,

a

"/mnt/"

#

will

be

prepended

to

the

pathname

when on

the

spu,

and

a

#

"!"

will

be

prepended

when on

the

JP.

#

#---------------------------------------------------------------------------

#

LEGEND

for

enhanced

IPI-2

(zone

bit

recording)

drives

#

a' -Drive

name - Must

be

an

unique

DKD-NNN

number.

#

b' -Mfg

id/model

number

#

c'

- Number

of

zones

in

drive

#

d'

-

Total

number

of

cylinders

per

unit

#

e'

- Number

of

data

bytes

per

sector

(2K

sector

size)

#

f'

- Number

of

logical

tracks

per

cylinder

#

g'

- Number

of

heads

per

logical

track

#

h'

- Number

of

spare

sectors

per

user

cylinder

#

i'

- Number

of

cylinders

reserved

for

diagnostic

use

#

j'

- Number

of

cylinders

reserved

for

manufacturers

defect

map

#

k'

- Number

of

cylinders

reserved

for

describing

drive

topology

#

l'

- Number

of

cylinders

reserved

for

spares.

#

m'

Header

size

#

n'

-

Header

gap

(in

50

nanoseconds

increments

for

buffered

drive)

72

Elite 3 Disc

Drive

Service

Guide

,,

#

*Use

recommended

MTD

value

of

OxlOO.

OxlOO

* 50 =

12.8

microseconds

#

o' -Data

gap

(in

50

nanoseconds

increments

for

buffered

drive)

# * Use

recommended

MTD

value

of

OxlOO.

OxlOO

* 50 =

12.8

microseconds

#

p' -Ecc

interleave

(options

are

O,

1, 2 or

4)

#

q'

- Number

of

spare

sectors

per

topology

cylinder

#

r'

-

This

should

have

the

same

DKD

name

as

line

1

of

the

entry.

#

s'

-

Data

Interleave,

applicable

only

to

parallel

heads

(BIT,

BYTE,

WORD)

#

t'

-

Drive

data

transfer

rate

in

bytes,

assume

16M

xfer

rate

#

u' -pathname

of

ascii

hex

patternle.

If

the

pathname

starts

#

with

a

!,

it

will

be

used

as

is.

If

it

does

not,

a

"/mnt/"

#

will

be

prepended

to

the

pathname

when on

the

spu,

and

a

#

"/"

will

be

prepended

when on

the

JP.

#

v'

- Zone

number,

from 0 to

(number

of

zones

- 1)

#

w'

-

This

should

have

the

same

DKD

name

as

#

x' -Starting

cylinder

in

this

zone

#

y' -Ending

cylinder

in

this

zone

#

z'

-

number

of

bytes

per

track

(see

0')

#

0'

- P - (

z'eld

is

the

physical

bytes

per

track

# L - (

z'eld

is

the

logical

bytes

per

track)

#

l'

- Number

of

sectors

per

logical

track

#----------------------------------------------------------------------------

BUGS

No

known

bugs

as

of

04/93.

Appendix A idcfmt

(1D)

man

page

73

Index

Symbols

/etc/disktab

30

Elite 3 disc

drive

entry

30

Entry description

30

/ioconfig file 28

A

Elite 3

en

tries

28

Example entries 29

associated

documents

xii

c

control board

jumpers

6

description 7 location 6 pin assignment 7

ConvexOS integration 27 ConvexOS requirements 27

CRC

errors 44

D

damage

claims 14

de

power

requirements 3

determining disc usage 46

E

ECC

errors 44

electrostatic discharge protection

xiii

Elite 3 disc

drive

error codes 33

description 34

Elite 3 disc

drive

specifications 1

expansion cabinet

drive

locations for 16 drives

25

expansion cabinet

drive

locations for 32 drives

24

F

fault

symptom

code

(FSC) 35, 44 descriptions 36 field descriptions 35 format 35 reported

during

autoconf 45

reported from

unformatted

disc 45

FCC

Notice

xiii

1/0

board

jumpers

8

locations 8

pin assignment 8

IDC bulkhead

port

assignments C3200/3400 18 C3800 19

JDC

drive connections

23

idcfmt

formatting

disc

drive

32

manpage

55

bugs

73

commands 57

error codes

6.3

examples

61

files

69

layout

of

an

idc disk

62

options

55

synopsis 55

slipping sectors 44

verify format

31

illustrated

parts

breakdown

(IPB) 52

inspection

13

installation

14

L

cabling a single

drive

17

cabling multiple disc drives 20

disc tray 14 Elite

3 chassis 16

power

strip 14

preparations 14

logical

unit

number

28

N

notational conventions

xi

0

ordering

documents

xii

Index

75

p

physical configurations 22 power

supply

cable

assignments

11

component locations

10

voltage specifications 10

power

supply

specifications

input

voltage 9

R

operational environment 9

output

voltage adjustability 9

weight 9

reading Elite 3 disc error codes

33

removal

and

replacement procedures

48

air

filter

51

s

Elite 3 chassis

48

Elite 3 disc drive

48

fan assembly 50

operator panel 49 power

supply

49

sector errors 44 specifications

capacity

1

T

environmental requirements 2 latency 2

physical

2

seek

time 2

size 1

start

time 2

transfer rate 2

technical assistance

xiii

troubleshooting

33

u

unpacking

13

v

VVM messages 45

device failure message

45

successful reconstruction

45

76

Elite 3 Disc

Drive

Service

Guide

1111111111

I

IHI

Convex DKD-505, Elite 3 Service Manual

Specifications and Main Features

Frequently Asked Questions

User Manual