. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . .
DDS-4 and DAT 72 Tape Drives and
Autoloaders
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . .
SCSI Interface Manual
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . .
Copyright © 2003 Certance LLC. All rights reserved
Part Number 50000697
June 2003, Rev.B
Certance and the Certance logo are trademarks of Certance LLC. Seagate is a
trademark of Seagate Technology LLC. Other product names are trademarks or
registered trademarks of their respective owners.
Certance reserves the right to change, without notice, product offerings or
specifications. No part of this publication may be reproduced in any form without
written permission from Certance LLC.
Important Information About This Manual
All information contained in or disclosed by this document is considered proprietary
by Certance. By accepting this material, the recipient agrees that this material and
the information contained therein are held in confidence and in trust and will not be
used, reproduced in whole or in part, nor its contents revealed to others, except to
meet the purpose for which it was delivered. It is understood that no right is
conveyed to reproduce or translate any item herein disclosed without express
written permission from Certance.
Certance provides this manual "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. Certance reserves the right to
change, without notification, the specifications contained in this manual.
Certance assumes no responsibility for the accuracy, completeness, sufficiency, or
usefulness of this manual, nor for any problem that might arise from the use of the
information in this manual.
Contents Page iii
Contents
Introduction 1
Overview......................................................................................................................1
Drive Models ...............................................................................................................1
About This Manual ......................................................................................................1
SCSI Interface Description 2
Overview......................................................................................................................2
ANSI SCSI Bus Standards ...................................................................................2
Cabling and Connectors.......................................................................................2
Signal Descriptions......................................................................................................3
Command Set Description ..........................................................................................4
ANSI X3.131, 199x Conformance Statement (SCSI-2) ..............................................5
General features...................................................................................................5
Interface standards conformance.........................................................................5
Commands............................................................................................................6
* Autoloader only commands................................................................................6
Messages..............................................................................................................6
Vendor-unique commands....................................................................................6
SCSI Bus Protocol.......................................................................................................6
Waiting and Control Phases........................................................................................7
Arbitrating Systems...............................................................................................7
Selection and Reselection Phases..............................................................................9
Select With Attention ..........................................................................................10
Identify Message.................................................................................................10
Information Transfer Phases.....................................................................................11
Asynchronous Data Transfer..............................................................................11
Synchronous Data Transfer................................................................................13
Command Phase.......................................................................................................14
Data Phase................................................................................................................15
Data-In Phase.....................................................................................................15
Data-Out Phase..................................................................................................15
Status Phase.............................................................................................................15
Message Phase.........................................................................................................15
Message-In Phase..............................................................................................15
Page iv DDS-4 SCSI Interface Manual, Rev. B
Message-Out Phase...........................................................................................15
Command Descriptor Block.......................................................................................16
Logical Unit Support..................................................................................................17
SCSI Message Descriptions and Definitions.............................................................18
SCSI Status Code Descriptions and Definitions .......................................................23
Attention Condition....................................................................................................24
Reset Condition.........................................................................................................24
Unit Attention Condition.............................................................................................24
Buffered Mode...........................................................................................................25
Immediate Function...................................................................................................25
Residual Length Function..........................................................................................25
Disconnect/Reconnect Function................................................................................26
SCSI Memory Address Pointers................................................................................27
Current Data Pointers .........................................................................................27
Saved Data Pointers...........................................................................................27
Early Warning Function.............................................................................................27
Error Reporting..........................................................................................................28
Soft Errors...........................................................................................................28
Hard Errors..........................................................................................................28
Variable and Fixed Mode Recording.........................................................................28
Autoloader Operation ................................................................................................29
General SCSI Information...................................................................................29
SCSI Medium Changer Commands ...................................................................29
General Operation ..............................................................................................31
Loading and Initialization ....................................................................................31
Front Panel Operation.........................................................................................32
Cassette Selection..............................................................................................32
Automatic Loading Mode....................................................................................32
Magazine Ejection...............................................................................................33
Stray Tape Recovery..........................................................................................33
Commands 34
Introduction................................................................................................................34
Command Reference List...................................................................................34
Conventions........................................................................................................35
Command Descriptor Blocks (CDBs)........................................................................35
Command Descriptor Block Formats..................................................................35
Command Descriptor Block Field Descriptions..................................................36
Flag and Link Bit Descriptions ............................................................................36
ERASE (19h).............................................................................................................37
Contents Page v
ERASE Command Descriptor Block...................................................................37
Command Descriptor Block Field Description....................................................37
Completion Status ..............................................................................................37
EXCHANGE MEDIUM (AutoLoader Only) (A6h)......................................................39
EXCHANGE MEDIUM Command Descriptor Block...........................................39
Command Descriptor Block Field Descriptions..................................................39
Completion Status ..............................................................................................40
INITIALIZE ELEMENT STATUS (AutoLoader Only) (07h).......................................41
INITIALIZE ELEMENT STATUS Command Descriptor Block............................41
Completion Status ..............................................................................................41
INQUIRY (12h)..........................................................................................................42
INQUIRY Command Descriptor Block................................................................42
Command Descriptor Block Field Description....................................................42
Standard INQUIRY Data Format........................................................................43
Standard INQUIRY Data Format Field Descriptions..........................................43
INQUIRY Drive Serial Number Data Format Page.............................................45
Drive Serial Number Data Format Field Descriptions ........................................45
Completion Status ..............................................................................................45
LOAD/UNLOAD (1Bh)...............................................................................................46
LOAD/UNLOAD Command Descriptor Block.....................................................46
Command Descriptor Block Field Description....................................................46
Completion Status ..............................................................................................47
LOCATE (2Bh)..........................................................................................................48
LOCATE Command Descriptor Block ................................................................48
Command Descriptor Block Field Description....................................................48
LOCATE and the BT Bit......................................................................................49
Completion Status ..............................................................................................50
LOG SELECT (4Ch)..................................................................................................51
LOG SELECT Command Descriptor Block........................................................51
Command Descriptor Block Field Description....................................................52
Errors Detected in the Command Descriptor Block............................................52
Use of the PC Bits...............................................................................................52
LOG SELECT Parameter Data...........................................................................53
LOG SENSE (4Dh)....................................................................................................58
LOG SENSE Command Descriptor Block..........................................................58
Command Descriptor Block Field Description....................................................59
Using Page Control Bits......................................................................................60
Using the Parameter Pointer Field .....................................................................61
Log Sense Pages ...............................................................................................61
Completion Status ..............................................................................................73
Page vi DDS-4 SCSI Interface Manual, Rev. B
MODE SELECT (15h) ...............................................................................................75
MODE SELECT Command Descriptor Block.....................................................75
Command Descriptor Block Field Description....................................................76
MODE SELECT Parameters...............................................................................76
Parameter List Header Field Descriptions..........................................................77
Parameter List—Block Descriptor.......................................................................77
Parameter List—Block Descriptor Field Descriptions.........................................78
Mode Page Format .............................................................................................78
Completion Status...............................................................................................90
MODE SENSE (1Ah).................................................................................................91
MODE SENSE Command Descriptor Block.......................................................91
Command Descriptor Block Field Description....................................................92
MODE SENSE Parameters ................................................................................93
Parameter List Header Field Descriptions..........................................................93
Parameter List—Block Descriptor.......................................................................94
Parameter List Block Descriptor Field Descriptions ...........................................94
Mode Page Format .............................................................................................95
Completion Status.............................................................................................103
MOVE MEDIUM (Autoloader only) (A5h)................................................................104
MOVE MEDIUM Command Descriptor Block...................................................104
Completion Status.............................................................................................105
PREVENT/ALLOW MEDIA REMOVAL (1Eh).........................................................106
PREVENT/ALLOW MEDIA REMOVAL Command Descriptor Block...............106
Command Descriptor Block Field Description..................................................106
Completion Status.............................................................................................106
READ (08h) .............................................................................................................107
READ Command Descriptor Block...................................................................107
Command Descriptor Block Field Description..................................................107
Description of the READ Command .................................................................108
Completion Status.............................................................................................111
READ BLOCK LIMITS (05h) ...................................................................................112
READ BLOCK LIMITS Command Descriptor Block.........................................112
Command Descriptor Block Field Description..................................................112
Block Size Definition .........................................................................................112
Completion Status.............................................................................................113
READ DATA BUFFER (3Ch) ..................................................................................114
READ DATA BUFFER Command Descriptor Block.........................................114
Command Descriptor Block Field Description..................................................114
READ DATA BUFFER Header.........................................................................115
READ DATA BUFFER Header Field Descriptions ...........................................115
Contents Page vii
Completion Status ............................................................................................115
READ ELEMENT STATUS (Autoloader Only) (B8h)..............................................116
READ ELEMENT STATUS Command Descriptor Block .................................116
Element Status Data.........................................................................................116
Element Status Data Header............................................................................117
Completion Status ............................................................................................120
READ POSITION (34h)...........................................................................................121
READ POSITION Command Descriptor Block ................................................121
READ POSITION Command Descriptor Block Field Descriptions...................121
READ POSITION Data Format.........................................................................122
READ POSITION Data Format Field Descriptions...........................................122
READ POSITION Data Format.........................................................................123
READ POSITION Data Format Field Descriptions...........................................123
Completion Status ............................................................................................125
RECEIVE DIAGNOSTIC RESULTS (1Ch) .............................................................126
RECEIVE DIAGNOSTIC RESULTS Command Descriptor Block....................126
Command Descriptor Block Field Description..................................................126
Completion Status ............................................................................................127
RELEASE UNIT (17h / 57h)....................................................................................129
RELEASE UNIT Command Descriptor Block...................................................129
Command Descriptor Block Field Description..................................................130
Completion Status ............................................................................................130
REPORT DENSITY SUPPORT (44h).....................................................................131
REPORT DENSITY SUPPORT Command Descriptor Block...........................131
Command Descriptor Block Field Description..................................................131
REPORT DENSITY Response Data................................................................131
REPORT DENSITY SUPPORT Header...........................................................132
REPORT DENSITY SUPPORT Data block descriptor.....................................132
DENSITY SUPPORT Data Block Descriptor Field Descriptions......................132
Completion Status ............................................................................................133
REQUEST BLOCK ADDRESS (02h)......................................................................134
REQUEST BLOCK ADDRESS Command Descriptor Block............................134
Command Descriptor Block Field Description..................................................134
REQUEST BLOCK ADDRESS Address Data Format.....................................134
REQUEST BLOCK ADDRESS Address Data Field Description......................135
Completion Status ............................................................................................135
REQUEST SENSE (03h) ........................................................................................136
REQUEST SENSE Command Descriptor Block..............................................136
Sense Data Format...........................................................................................136
Sense Data Field Descriptions .........................................................................137
Page viii DDS-4 SCSI Interface Manual, Rev. B
Definition of Sense Keys...................................................................................138
Additional Sense Code and Code Qualifier......................................................138
Vendor-Specific Code Qualifiers.......................................................................143
Completion Status.............................................................................................148
RESERVE UNIT (16h / 56h)....................................................................................149
RESERVE UNIT Command Descriptor Block ..................................................149
Command Descriptor Block Field Description..................................................150
Completion Status.............................................................................................151
REWIND (01h).........................................................................................................152
REWIND Command Descriptor Block ..............................................................152
Command Descriptor Block Field Description..................................................152
Completion Status.............................................................................................152
SEEK BLOCK (0Ch)................................................................................................154
SEEK BLOCK Command Descriptor Block......................................................154
Command Descriptor Block Field Description..................................................154
Completion Status.............................................................................................155
SEND DIAGNOSTIC (1Dh).....................................................................................156
SEND DIAGNOSTIC Command Descriptor Block ...........................................156
Completion Status.............................................................................................157
SPACE (11h)...........................................................................................................158
SPACE Command Descriptor Block.................................................................158
Command Descriptor Block Field Descriptions................................................159
Space-by-Count Functions ...............................................................................159
Space by Position Functions.............................................................................160
Space and the RSMK Bit..................................................................................160
Completion Status.............................................................................................161
TEST UNIT READY (00h).......................................................................................162
TEST UNIT READY Command Descriptor Block.............................................162
TEST UNIT READY Detailed Operation...........................................................162
Completion Status.............................................................................................163
VERIFY (13h) ..........................................................................................................164
VERIFY Command Descriptor Block................................................................164
Command Descriptor Block Field Description..................................................164
Data Blocks Verified..........................................................................................164
Completion Status.............................................................................................165
WRITE (0Ah) ...........................................................................................................166
WRITE Command Descriptor Block .................................................................166
Command Descriptor Block field description....................................................166
Completion Status.............................................................................................166
WRITE DATA BUFFER (3Bh).................................................................................168
Contents Page ix
WRITE DATA BUFFER Command Descriptor Block.......................................168
Command Descriptor Block Field Descriptions................................................169
WRITE DATA BUFFER Data Header...............................................................169
Completion Status ............................................................................................170
WRITE FILEMARKS (10h)......................................................................................171
WRITE FILEMARKS Command Descriptor Block............................................171
Command Descriptor Block Field Description..................................................171
Completion Status ............................................................................................172
Introduction Page 1
Introduction
1
Overview
The Seagate DDS-4 and DAT 72 DAT drives are designed for computer
environments requiring high performance, high capacity data storage. These drives
are available in 3.5-inch or 5.25-inch internal configurations or as an external
subsystem. The Autoloaders contain a DDS-4 or DAT 72 drive with an enclosure
and tape manipulation mechanism.
Seagate DDS-4 and DAT 72 drives contain an embedded Small Computer Systems
Interface (SCSI) controller. These drives provide synchronous or asynchronous
SCSI and a high speed burst synchronous data transfer rate of 80 Mbytes/second.
The internal drive form factors are tailored for easy installation in today's computers
and the full-featured embedded SCSI controller facilitates easy integration into a
variety of systems.
Drive Models
This manual provides detailed information about the SCSI interface and SCSI
commands that apply to Seagate DDS-4 and DAT 72 tape drives and autoloaders.
These drives offer electronically erasable, programmable, read-only memory (flash
EEPROM) for SCSI firmware, which enables qualified Seagate OEMs to download
revised firmware to the drive via two methods: using the SCSI bus or using a
specialized Seagate firmware tape cartridge.
About This Manual
The information in this manual applies to DDS-4 and DAT 72 drives and
autoloaders. For specific installation and operational information—including SCSI
connection information—refer to the Installation Guide or Product Description
Manual for the specific model of your drive or Autoloader. The following table
outlines the chapters in this manual.
Chapter Title Description
1 Introduction Introduces the SCSI interface and explains
2 SCSI Interface
3 Commands Lists and describes the SCSI commands.
Description
the structure of this manual.
Provides general information about the SCSI
interface for the DDS-4 and DAT 72 drives.
Page 2 DDS-4 SCSI Interface Manual, Rev. B
SCSI Interface Description
2
Overview
The Seagate DDS-4 and DAT 72 drives are designed to operate with the Small
Computer System Interface (SCSI) bus. This chapter discusses SCSI bus operation
as it pertains to drive functions.
SCSI is a standard interface established to support peripheral equipment such as
printers, tape drives, magnetic disks, optical disks for microcomputers and other
computer systems. The SCSI bus can support up to eight devices consisting of any
multiple of host adapters and peripheral devices.
The Seagate DDS-4 and DAT 72 drives comply with SCSI-2 and SCSI-3
specifications. In a few cases, vendor unique features are available. These features
are compatible with the SCSI standards.
The interface is a sixteen-port daisy-chained bus using twenty-seven signal lines:
eighteen data-bit signal lines and nine control lines. The eighteen data-bit lines are
made up of two pairs of eight data signal lines and one parity bit line. The remaining
nine lines provide control and status signals to coordinate data transfer operations
between the host controller and the selected drive.
The drives have an internal SCSI controller integrated into the drive electronics.
Each device ID on the SCSI bus may drive up to 8 logical units (LUN). These drives
support only LUN 0, except for the Autoloader, which also uses LUN 1.
ANSI SCSI Bus Standards
In addition to the information presented in this manual, we recommend that for
SCSI-2, read the ANSI Working Draft Revision 10 standard before writing host
software drivers. Also, see the conformance statements, which are given in the
Product Description Manual for the each model of drive or Autoloader.
Cabling and Connectors
The cabling requirements and pin outs for the SCSI connector for the internal drive
models are given in the respective installation guide and Product Description
Manual for each model of DAT drive or Autoloader.
SCSI Interface Description Page 3
Signal Descriptions
The drive SCSI interface consists of eighteen signals. Nine are control lines and
eighteen are data lines. Data lines include the parity signal option. These signals are
described in the following table.
Signal Name Description
-BSY
-SEL
-C/D
-I/O
-MSG
-REQ
-ACK
-ATN
-RST
DB(7-0)
DB(8-15)
DB(P0)
DB(P1)
NOTE: The BSY and RST signals are the only OR-tied signals. In ordinary bus operation, these
signals may be simultaneously driven by two or more drivers. There is no operational problem in mixing
OR-tied and three-state drivers on signals other than BSY and RST.
Busy OR-tied signal used to show that the data bus is in use
Select Signal used by an Initiator to select a Target or by a Target to
reselect an Initiator. SEL is driven by the Initiator during the
Selection Phase and driven by the Target during a Reselection
Phase.
Control/Data Target-driven signal used to indicate whether Control or Data
information is on the data bus. True (low) indicates Control, and
false (high) indicates Data.
Input/Output Target-driven signal used to control data movement direction on
the data bus with respect to an Initiator. This signal is also used to
distinguish between the Selection and Reselection Phases. True
(low) indicates input to the Initiator, and false (high) indicates
output from the Initiator.
Message Target-driven signal used to indicate the presence of a Message
Phase on the bus. True (low) indicates Message Phase, and false
(high) indicates Data, Command, or Status Phase.
Request Target-driven signal used to indicate a request for a REQ/ACK
data transfer handshake.
Acknowledge Initiator-driven signal used to indicate an acknowledgment for a
REQ/ACK data transfer handshake.
Attention Initiator-driven signal used to indicate the Initiator has a message
to communicate to the Target.
Reset OR-tied signal used to indicate a Reset condition.
Data Bus Eight data-bit signals plus a parity bit signal that form the odd
bytes of the data bus. DB (7) is the MSB and has the highest
priority (ID 7) during the Arbitration Phase.
Data Bus Eight data-bit signals that form the even byte of the data bus.
DB(15) is the MSB
Data Bus The data parity bit for the odd byte of the data bus. Parity is odd
and is a jumper-selectable option. Parity is not valid during the
Arbitration Phase. DB (P) is not to be driven False (high) during
the Arbitration Phase.
Data Bus Data parity bit for the even byte of the data bus.
Page 4 DDS-4 SCSI Interface Manual, Rev. B
Command Set Description
The following table shows the SCSI commands for sequential access devices
implemented by the drive and autoloader.
Code Type Command Group
00h M TEST UNIT READY 0
01h M REWIND 0
02h V REQUEST BLOCK ADDRESS 0
03h M REQUEST SENSE 0
05h M READ BLOCK LIMITS 0
07h A INITIALIZE ELEMENT STATUS 0
08h M READ 0
0Ah M WRITE 0
0Ch V SEEK BLOCK 0
10h M WRITE FILEMARKS 0
11h M SPACE 0
12h M INQUIRY 0
13h O VERIFY 0
15h M MODE SELECT 0
16h M RESERVE UNIT 0
17h M RELEASE UNIT 0
19h M ERASE 0
1Ah M MODE SENSE 0
1Bh O LOAD/UNLOAD 0
1CH O RECEIVE DIAGNOSTIC RESULTS 0
1Dh M SEND DIAGNOSTIC 0
1Eh O PREVENT/ALLOW MEDIUM REMOVAL 0
2Bh O LOCATE 1
34h O READ POSITION 1
3Bh O WRITE DATA BUFFER 1
3Ch O READ DATA BUFFER 1
44h M REPORT DENSITY SUPPORT (SCSI-3) 2
4Ch O LOG SELECT 2
4Dh O LOG SENSE 2
56H O RESERVE UNIT (SCSI-3) 2
57H O RELEASE UNIT (SCSI-3) 2
A5H A MOVE MEDIUM 5
A6H A EXCHANGE MEDIUM 5
B8H A READ ELEMENT STATUS 5
M = Mandatory Command V = Vendor Unique Command
O = Optional Command A = Autoloader Command Only
SCSI Interface Description Page 5
ANSI X3.131, 199x Conformance Statement (SCSI-2)
General features
• Disconnect/reconnect, arbitration (required in SCSI-2).
• Supports LVD and single-ended drivers.
• Fixed and variable block transfer lengths.
• Hard reset.
• Synchronous data transfers
• Parity implemented (switch option).
• Space blocks, filemarks, EOD, and setmarks.
• Supports third-party reservation
• Log Sense and Log Select for managing soft errors reporting.
• MODE SENSE/SELECT page to control and report operation of data
compression in sequential access devices.
• Complies with SCSI-2, Section 16, “Medium Changer Devices*.
Interface standards conformance
The Small Computer System Interface is described in standards including several
versions and a number of individual documents. The original Small Computer
System Interface Standard, X3.131-1986, is referred to herein as SCSI-1. SCSI-1
was revised resulting in the Small Computer System Interface – 2 (X3.131-1994),
referred to herein as SCSI-2. The set of SCSI-3 standards are collectively referred
to as SCSI-3. The applicable SCSI-3 standards are as follows:
SCSI-3 Architecture Model (SAM) X3.270:1996
SCSI Architecture Model - 2 (SAM-2) INCITS.366:2003
SCSI Parallel Interface - 2 (SPI-2) X3.302:1998
SCSI Parallel Interface - 3 (SPI-3) NCITS.336:2000
SCSI-3 Primary Commands (SPC) X3.301:1997
SCSI Primary Commands - 2 (SPC-2) NCITS.351:2001
SCSI-3 Medium Changer Commands (SMC) NCITS.314:1998
SCSI-3 Stream Commands (SSC) NCITS.335.2000
The term SCSI is used wherever it is not necessary to distinguish between the
versions of SCSI.
Page 6 DDS-4 SCSI Interface Manual, Rev. B
Commands
Erase Read Position
Exchange Medium* Release Unit
Initialize Element Status* Receive Diagnostic Results
Inquiry Report Density Support
Load/Unload Request Block Address
Locate Request Sense
Log Select Reserve Unit
Log Sense Rewind
Mode Select Seek Block
Mode Sense Send Diagnostic
Move Medium* Space
Prevent/Allow Media Removal Test Unit Ready
Read Verify
Read Block Limits Write
Read Data Buffer Write Data Buffer
Read Element Status* Write Filemarks
* Autoloader only commands
Messages
Abort Linked Command Complete with flag
Bus Device Reset Message Reject
Disconnect No Operation
Identify Parity Error Message
Ignore Wide Residue Save Data Pointer
Initiator Detected Error Synchronous Data Transfer Request
Linked Command Complete Wide Data Transfer Request
Vendor-unique commands
Seek Block Request Block Address
SCSI Bus Protocol
Communication on the SCSI bus occurs between a host computer's SCSI controller
and a peripheral controller. The host controller is the Initiator, and the peripheral
device is the Target.
Some SCSI bus functions are assigned to the Initiator and others to the Target. The
Initiator arbitrates (enters the Arbitration Phase of operation) for control of the SCSI
bus and enters the Selection Phase to select a specific Target such as a DAT drive.
The Target drive can request transfer of command, data, status, information, or
disconnect from the bus. While the tape drive is disconnected, the bus is free to
accomplish other tasks. While disconnected, the Target drive can process
information obtained from the bus. If the Target wants to reconnect, it arbitrates for
bus control. For example, a Target can reselect an Initiator or another Target to
continue an operation.
SCSI Interface Description Page 7
Data-transfer operations on the SCSI bus are either asynchronous or synchronous.
Asynchronous data transfer operations follow a defined request/acknowledge
(REQ/ACK) handshake protocol. Two eight-bit bytes of information can be
transferred with each REQ/ACK handshake. The default data transfer mode is
asynchronous.
Synchronous data transfer operations are initiated through the SYNCHRONOUS
DATA TRANSFER REQUEST message from the Initiator.
The SCSI bus protocol is divided into three modes or phases of operation: Waiting
Phases, Control Phases, and Information Transfer Phases. These phases are
further subdivided into the eight operational phases as listed in the following table.
Information on these operational phases is contained in the following paragraphs.
Waiting Phases Control Phases Information Phases
1. Bus Free 2. Arbitration 5. Command
3. Selection 6. Data (Data In/Data Out)
4. Reselection 7. Status
8. Message (Message In/
Message Out)
Waiting and Control Phases
The status of the SCSI bus is a function of the control signals. These signals define
the SCSI as in the Waiting Phase (Bus Free Phase), the Control Phases
(Arbitration, Selection, or Reselection), or the Information Transfer Phases
(Command, Data, Status, or Message). Seagate DDS-4 and DAT 72 drives support
both a SCSI system with Arbitration Phase and a SCSI system without Arbitration
Phase (non-arbitrating system).
Arbitrating Systems
In arbitration systems (required in SCSI-2), the sequence of SCSI bus phases
follows the sequence shown in Figure 2-1 on the following page. Operation begins
with the Bus Free Phase. Normal progression is from the Bus Free to the Arbitration
Phase. During Arbitration, Initiators and Targets assert for control of the SCSI bus.
The bus is awarded to the device with the highest priority SCSI bus address.
Page 8 DDS-4 SCSI Interface Manual, Rev. B
Figure 2-1
Phase Sequencing with Arbitration
RESET CONDITION
BUS FREE
PHASE
ARBITRATION
PHASE
SELECTION OR
RESELECTION
PHASE
COMMAND,
DATA, STATUS,
OR MESSAGE
PHASE
Arbitration is won by the highest priority SCSI device when both BSY and SEL are
asserted and a delay of at least 1200 nSec (1 bus clear delay + 1 bus settle delay)
occurs before the Arbitration Phase ends and the Selection Phase begins. Signal
timing is shown in Figure 2-3 and the complete Arbitration Phase protocol is
summarized in the following table.
Figure 2-2
Arbitration and Selection Phase Signal Timing
Step Procedure
1.
The SCSI device first waits for a Bus Free Phase to occur. The Bus
Free Phase is detected when both the BSY and SEL signals are
simultaneously and continuously false for a minimum of a bus settle
delay of 400 nSec.
SCSI Interface Description Page 9
2.
3.
4.
5.
NOTE: The single SCSI ID bit on the data bus corresponds to the unique ID code
of the SCSI device. All other SCSI data bus bits are released by the SCSI device.
Parity is not valid during the Arbitration Phase. During the Arbitration Phase, DB(P)
may be undriven or driven true, but not false.
The SCSI device waits a minimum of one bus free delay of 800 nSec
after detection of a Bus Free Phase before driving any signal. The Bus
Free Phase occurs after BSY and SEL are both false for a bus settle
delay of 400 nSec.
Following the 800 nSec Bus Free Delay in Step 2, the SCSI device
Arbitrates for the SCSI bus by asserting both BSY and its own SCSI ID;
however, the SCSI device does not assert a BSY and its SCSI ID if
more than a bus settle delay (1.8 Sec) has passed since the Bus Free
Phase was last observed.
After waiting at least an Arbitration delay (2.2 Sec) measured after
asserting a BSY signal, the SCSI device examines the data bus. If a
higher priority SCSI ID bit is true, then the SCSI device loses the
Arbitration. The SCSI device releases its signals and returns to Step 1. If
no higher priority SCSI ID bit is true on the bus, then the SCSI drive
wins the Arbitration and asserts the SEL signal. Any other device that
participated in the Arbitration Phase and lost Arbitration releases BSY
and its SCSI ID bit within a bus clear delay after SEL becomes true. A
SCSI device that loses Arbitration returns to Step 1.
The SCSI device that wins Arbitration waits at least one bus clear delay
plus a bus settle delay of 1,200 nSec after asserting the SEL signal and
changing the condition of other signals.
Selection and Reselection Phases
The Selection and Reselection Phases provide a method for establishing a link
between the Initiator and Target. When selected by an Initiator that supports
disconnects, the Target has the option of disconnecting from the SCSI bus. When
the Target needs to again establish the link to its original Initiator, the Target
reselects that Initiator.
When selection is made, no restrictions on the sequences between Information
Transfer Phases exist. A phase type may be followed by the same phase type. For
example, a data phase may be followed by another data phase.
A device that wins Arbitration assumes the roll of Initiator by releasing the I/O signal.
The Initiator sets the data bus to a value that is the OR of its SCSI ID bit and the
Targets ID bit. The Initiator waits at least two deskew delays (90 nSec) before
releasing the BSY signal and an additional bus settle delay (400 nSec) before
looking for a response from the Target.
A device that loses Arbitration must release the Select ID and BSY.
Operation enters the Selection or Reselection Phases when a particular device wins
a request in Arbitration and gains control of the SCSI bus. These phases allow the
device in control of the bus to select another specific device connected to the SCSI
bus for communication.
Page 10 DDS-4 SCSI Interface Manual, Rev. B
For example, the Initiator can select a tape drive to begin an operation, or the tape
drive can reselect an Initiator to continue an operation previously disconnected. The
tape drive does not disconnect from the bus unless the host asserted the host ID bit
during selection and the host sent an Identify message of C0h.
The Selection and Reselection Phases can be terminated for either of the following
reasons:
• A Selection/Reselection time-out occurs. A Target or Initiator did not respond to
a Selection or Reselection Phase within a maximum abort time of 200 Sec.
• A reset signal occurs on the SCSI bus. All sequences are terminated, and
signals are released by all Targets and Initiators.
The Initiator can use the ATN signal to notify the DAT drive that an IDENTIFY
message from the Initiator is ready. To ensure that the Target recognizes the
Attention condition before the Command Phase is entered, the ATN signal must be
low before SEL is asserted and BSY de-asserted.
Select With Attention
The host system can select the drive only using Select With Attention. The Select
With Attention allows the host to send the Identify message to specify that the host
supports disconnect.
The host can also use the Select With Attention to send messages other than the
Identify message. For example, a Bus Device Reset or Abort message can be sent
to cause the drive to reset itself without affecting any other device on the SCSI bus.
The DAT drive responds appropriately to the host messages.
If the drive receives an illegal message, it enters the Message Out Phase and sends
the Reject message to the host. If any reserved bit (bits 5, 4, or 3) is set, an Identify
message is rejected.
Responses to Identify messages with a non-zero LUN are described in the following
subsection.
Identify Message
Because the drive only supports Logical Unit 0, the Identify message is used solely
to specify the Disconnect option. The bit map for the Identify message is as follows:
Bits 7 6 5 4 3 2 1 0
| DIS | Reserved | LUN
If bit 7 is set, the Identify message is indicated. Bit 6 (shown as DIS) is set as
follows:
0 The host does not support disconnects. The drive does not disconnect from the
SCSI bus during the current command.
SCSI Interface Description Page 11
1 The host supports disconnects. The drive disconnects from the SCSI bus
appropriately for the command in progress to relinquish the bus for other units.
Note: LUN = 1 is legal for Autoloader drives only.
Information Transfer Phases
The Command, Data, Status, and Message Phases are grouped together as the
Information Transfer Phases because they are all used to transfer data or control
information on the SCSI bus.
The following table shows the Control Data (C/D), Input/Output (I/O), and Message
(MSG) signals used to distinguish between the various Information Transfer Phases.
The Target drives these three signals and thereby controls all changes from one
phase to another.
Signals -MSG -C/D -I/O Phases Direction of Transfer
-I/O
1 1 0 Data Target to Initiator
-C/D
1 0 0 Status Target to Initiator
-MSG
0 0 0 Message Target to Initiator
Key: 1 = False, 0 = True
1 1 1 Data Initiator to Target
1 0 1 Command Initiator to Target
0 0 1 Message Initiator to Target
Information Transfer Phases use one or more REQ/ACK handshakes to control the
information transfer. Each REQ/ACK handshake allows the transfer of one byte of
information. During the information transfer phases, BSY remains true and SEL
remains false. C/D, I/O, and MSG control signals are valid for a bus settle delay of
400 nSec before assertion of REQ signal at the first handshake and remain valid
until negation of ACK at the end of the last handshake.
Asynchronous Data Transfer
The Target controls the direction of information transfer with the I/O signal. When
I/O is asserted (low), information is transferred from the Target to the Initiator. When
I/O is de-asserted (high), information is transferred from the Initiator to the Target.
Figure -4 shows the data transfer signals for information transfer from Target to
Initiator, and Figure 2-5 shows the data transfer signals for information transfer from
Initiator to Target. Each direction of information transfer is discussed in detail in the
following paragraphs.
Transfer from Target to Initiator
The transfer from Target to Initiator is described in the following table and illustrated
in Figure 2-4.
Timing Point Action
t1
t2
The I/O signal is asserted; the Target drives the data (DB7-0)
and parity signals to their desired values.
The Target then asserts the REQ signal.
Page 12 DDS-4 SCSI Interface Manual, Rev. B
t3
The Initiator reads the data and parity signals after REQ is
asserted and then signals its acceptance of the data by asserting
the ACK.
t4
t5
ACK goes low at the Target, and the Target de-asserts REQ.
The REQ signal is false. The Initiator de-asserts ACK. After the
ACK signal is high, the Target can continue the transfer by
repeating the steps from t1.
Figure 2-4
Signals Used in Transfer from Target to Initiator
Transfer from Initiator to Target
Transfer from Initiator to Target is described in the following table and illustrated in
Figure 2-5.
Timing Point Action
t1
The I/O signal is de-asserted, and the Target asserts REQ to
request information.
t2
The Initiator drives the data and parity signals to their desired
values.
t3
t4
The Initiator asserts the ACK.
The Target de-asserts REQ, signaling its acceptance of data.
The Initiator then releases the data and parity signals.
t5
The Initiator de-asserts ACK. The Target can then continue to
transfer by asserting the REQ signal.
SCSI Interface Description Page 13
Figure 2-5
Signals Used in Transfer from Initiator to Target
Synchronous Data Transfer
Both the Initiator and Target must send a SYNCHRONOUS DATA TRANSFER
REQUEST (SDTR) message in order to set up synchronous data transfer
parameters. The drives supports both SDTR WDTR extended (multibyte) messages.
The host may initiate an SDTR to establish a new synchronous data transfer
agreement or to end a previously arranged synchronous data transfer agreement
and return to asynchronous data transfer mode.
The following table shows the format for the SDTR message.
Byte Value Description
0
1
2
3
4
* Transfer Period can be any value between 0Ah and 64h. Exceptions to m x 4 nsec are
made in the following cases: 1) If the value is 0Ch, which is defined in SCSI-3 as a period
of 50 nsec, 2) If the value is 0Bh, which sets the period to 37.5 nsec, and 3) If the value is
0Ah, which sets the period to 25 nsec
** REQ/ACK Offset can be any value between 4 and 32.
01h Extended message
03h Extended message length
01h Synchronous Data Transfer Request Code
m* Transfer Period (m x 4 nsec)
** REQ/ACK Offset
.
The Transfer Period is the minimum time allowed between leading edges of
successive REQ pulses and of successive ACK pulses to meet the requirements of
the drive for successful reception of data. The host and DAT drive can transfer data
with larger, but never smaller, Transfer Period than specified in the SDTR message.
Page 14 DDS-4 SCSI Interface Manual, Rev. B
The minimum value fo the Transfer Period (25 nsec) is determined by the maximum
burst transfer rate of the SCSI interface hardware of the drive and is 80 Mbytes per
second (06h). The host must specify a Transfer Period that allows it to successfully
receive data from the DAT drive during the Data In Phase. Transfer Period
represents the actual transfer period, in nanoseconds, divided by 4.
The REQ/ACK Offset is the maximum number of pulses that can be sent by the DAT
drive in advance of the number of ACK pulses received from the host, establishing a
pacing mechanism. If the number of REQ pulses is greater than the number of ACK
pulses by the REQ/ACK Offset, the DAT drive stops sending data until after the
leading edge of the next ACK is received.
The ACK/REQ Offset is used to prevent an overflow condition in the host's reception
buffer during the Data In Phase. The REQ/ACK Offset should be set to the size of
the host's reception buffer minus one.
To set up a new synchronous data transfer agreement, the host asserts the ATN
signal and sends an SDTR message. The Transfer Period must be set to one of a
set of specific values equal to or greater than 0Ah and less than the maximum rate
of the host’s reception buffer. The ACK/REQ Offset will be set to a nonzero value
between 4 and the size of the host’s reception buffer minus 1, but it must not exceed
32.
If the Transfer Period and the ACK/REQ Offset are within the ranges described
above, the DAT drive goes to the Message In Phase and returns an SDTR message
with the same Transfer Period and ACK/REQ Offset. This return indicates a
successful completion of the SDTR message exchange. The implied synchronous
data transfer agreement remains in effect until:
• A Bus Device Reset message is received.
• A hard reset condition occurs.
• The successful completion of the next SDTR message exchange.
If the Transfer Period is less than 0Ah or the REQ/ACK Offset less than 4 the drive
returns either a Transfer Period of zero or a REQ/ACK Offset of zero to indicate the
request cannot be supported in a synchronous manner and use of asynchronous
transfer is requested.
If the host specifies a REQ/ACK offset of zero, the DAT drive operates in the
asynchronous data transfer mode.
Command Phase
During the Command Phase, the Target requests command information from the
Initiator. The Target asserts the C/D signal and de-asserts the I/O and MSG signals
thus denoting the Command Phase. The REQ/ACK then handshakes the command
bytes across the SCSI bus (Figure 2-5). The command bytes are also called the
Command Descriptor Block (CDB).
SCSI Interface Description Page 15
Data Phase
The Data Phase is subdivided into the Data-In and Data-Out Phases.
Data-In Phase
During the Data-In phase, the Target requests that data be sent to the Initiator from
the Target. The Target asserts the I/O signal and de-asserts the C/D and MSG
signals thus denoting the Data-In Phase. The REQ/ACK handshakes (Figure 2-4)
then transfer the requested byte count.
Data-Out Phase
During a Data-Out phase, the Target requests that data be sent from the Initiator to
the Target. The Target de-asserts the C/D, I/O, and MSG signals thus denoting the
Data-Out Phase. The REQ/ACK handshakes (Figure 2-5) then transfer the
requested byte count across the SCSI bus.
Status Phase
Message Phase
Message-In Phase
Message-Out Phase
During a Status Phase, the Target requests that status information be sent to the
Initiator from the Target. The Target asserts the C/D and I/O signals and de-asserts
the MSG signal thus denoting the Status Phase. The REQ/ACK handshakes the one
byte status code across the SCSI bus (refer to Figure 2-4).
The Message Phase consists of either the Message-In or Message-Out Phases.
The Seagate DDS-4 and DAT 72 drive supports one-byte messages.
During the Message-In Phase, the Target requests that messages be sent to the
Initiator from the Target. The Target asserts the C/D, I/O, MSG signals thus
denoting the Message-In Phase. The REQ/ACK handshakes the one-byte message
across the SCSI bus.
During the Message-Out Phase, the Target requests that messages be sent from
the Initiator to the Target. The Target invokes this phase in response to the Attention
(ATTN) signal asserted by the Initiator. The drive responds to the ATTN signal at
every phase change. The Target asserts the C/D and MSG signals and de-asserts
the I/O signal, denoting the Message-Out Phase. The REQ/ACK handshakes the
one byte message across the SCSI bus. The Target uses REQ/ACK handshakes
(Figure 2-5) until the ATN signal becomes false, unless an error occurs and the
message is rejected.
Page 16 DDS-4 SCSI Interface Manual, Rev. B
Command Descriptor Block
A request to a peripheral device is performed by sending a Command Descriptor
Block (CDB) to the Target. For several commands, the request is accompanied by a
list of parameters sent during a Data Out Phase. If an invalid parameter is contained
in the CDB, the drive terminates the command without altering the medium.
The Seagate DDS-4 and DAT 72 drives implementation supports Group 0 and
selected Group 1 and Group 2 commands. Group 0 CDBs are six-bytes. Group 1
and 2 CDBs are ten-bytes.
The CDB contains both reserved bit fields and defined bit fields. Defined bit fields
are: Group Code, Command Code, Logical Unit Number (LUN), Vendor Unique
(VU), Flag, and Link, whereas reserved bit fields are defined by zeros such as those
appearing on the last line of the following table. This table shows a typical Group 0,
six-byte, Command Descriptor Block.
Note: The LUN field must be 0 in the CDB. Target LUN numbers are sent in the
Identify Message.
Byte
0 Group Code Command Code
1 LUN Command Dependent
2 Command Dependent
3 Command Dependent
4 Command Dependent
5 VU VU 0 0 0 0 Flag Link
7 6 5 4 3 2 1 0
Bits
The following table shows a typical Command Descriptor Block for Group 1 and 2
commands.
Byte
0 Group Code Command Code
1 LUN Command Dependent
2 Command Dependent
3 Command Dependent
4 Command Dependent
5 Command Dependent
6 Command Dependent
7 Command Dependent
8 Command Dependent
9 VU VU 0 0 0 0 Flag Link
7 6 5 4 3 2 1 0
Bits
The following table describes the CDB fields common to all Group 0, 1, and 2
commands.
CDB Field Description
SCSI Interface Description Page 17
Group Code* This field indicates which of eight possible SCSI command groups is specified.
Seagate drives support Group 0.
Command
Code*
Logical Unit The LUN must be set to zero.
Vendor Unique When set, these bits select vendor unique functions in specified commands.
Flag bit The Flag bit is used only in conjunction with the Link bit and must be set to zero if
Link bit** The Link bit is used to indicate that the Initiator desires automatic linking to the next
* Together, group code and command code make up the op code.
** If the Link bit is used, all applicable commands must have the Immediate bit set to zero or a Check
Condition status is returned and Extended Sense Key is set to Illegal Request.
This field indicates which of 32 possible command codes for a particular group
code is specified.
the Link bit is zero. When the Link bit is set, the value of the Flag bit determines the
appropriate message to send to the Initiator when a linked command completes
successfully. A 0 value indicates that the Linked Command Complete message is
required. A 1 value indicates that the Linked Command Complete with Flag
message is required. Typically, the Flag bit is used to cause an interrupt in the
Initiator at the end of, or at logical intervals in, linked command processing.
command on successful completion of the current command. When the Link bit is
one, on successful termination of the command, the drive returns an Intermediate
Status followed by one of the two Command Complete messages as determined by
the Flag bit. (Refer to description of Flag bit above.)
Logical Unit Support
The Seagate DDS-4 and DAT 72 drive only supports Logical Unit (LUN) 0, except
for the Autoloader, which supports both LUN 0 and LUN 1. However, it responds to
a host command that tries to select or identify any other LUN. The response of the
drive to illegal LUNs varies depending on the command and the manner in which the
host specifies the LUN.
The host can specify an LUN I only by sending an Identify message after the
Selection Phase
Because of the redundancy of specifying the LUN, the Seagate DDS-4 and DAT 72
drive responds appropriately if the LUN is specified in both the Identify message and
in the CDB. If the drive is selected and a valid Identify message is received with
LUN = 0 (or LUN = 1 if the drive is an Autoloader), the LUN field of the CDB is
ignored. Thus, the Identify message overrides the CDB LUN specification.
Because the SCSI INQUIRY command determines what LUNs a particular SCSI
device supports, the response from the drive to this command is unique. With the
INQUIRY command when the drive receives an unsupported LUN, the command
completes normally. The drive indicates that it does not support the specified LUN
by returning a 7F hexadecimal value in byte 0 of the INQUIRY data. Regardless of
what method is used to specify the LUN, the 7F hex value is returned.
If an unsupported LUN is specified for a REQUEST SENSE command, the
command completes normally with Good Status, and the sense data is set to Illegal
Request.
For all other commands, the illegal LUN is detected in the Identify message. If an
illegal LUN is specified in the Identify message, the drive enters the Command
Phase and accepts the CDB. It immediately skips to the Status Phase and posts a
Check Condition. The Sense data is set to Illegal Request. If the Identify message is
correct, the CDB LUN is ignored.
Page 18 DDS-4 SCSI Interface Manual, Rev. B
SCSI Message Descriptions and Definitions
The SCSI message codes, descriptions, and directions are given in the following
table. Each of these SCSI messages is supported by the Seagate DDS-4 drive.
Code Description Direction*
00h
02h
03h
04h
05h
06h
07h
08h
0Ah
0Bh
0Ch
80h
C0h
01h**
* Direction: In = Drive to host; Out = Host to drive.
** Supports only two extended messages: Synchronous Data Transfer Request
and Wide Data Transfer Request.
Command Complete In
Save Data Pointer In
Restore data pointers In
Disconnect In
Initiator Detected Error Out
Abort Out
Message Reject In/Out
No Operation Out
Linked Command Complete In
Linked Command Complete with Flag In
Bus Device Reset Out
Identify (when sent by host, disables
Disconnect/Reconnect)
Identify (enable Disconnect/Reconnect) Out
Extended Message In/Out
In/Out
SCSI Interface Description Page 19
The following table lists SCSI messages and their definitions.
Hex
Code Message Definition
00h
02h
04h
05h
06h
07h
08h
09h
0Ah
Command
Complete
Save Data
Pointer
Disconnect This message is sent from the drive to inform an Initiator
Initiator
Detected
Error
Abort This message is sent from an Initiator to clear the present
Message
Reject
No
Operation
Parity Error This message is sent from the host to indicate that
Linked
Command
Complete
This message is sent from the drive to inform an Initiator
that execution of a command terminated and that valid
status was sent to the Initiator. After successfully sending
this message, the drive goes to the Bus Free Phase by
releasing BSY.
NOTE: The command may have been executed
successfully or unsuccessfully as indicated in the status.
This message is sent from the drive to direct the Initiator to
save a copy of the present active data pointer for the drive.
that the present physical path is going to be broken (DAT
drive plans to disconnect by releasing BSY) and, a later
reconnect is required to complete the current operation. If
the Initiator detects the Bus Free Phase, other than as the
result of a Reset condition, without first receiving a
Disconnect or Command Complete message, the Initiator
considers this as a catastrophic error condition. The
Disconnect message does not cause the Initiator to save
the data pointer. If Disconnect messages are used to break
a long data transfer into two or more shorter transfers, then
a Save Data Pointer message is issued by the DAT drive
before each Disconnect.
This message is sent from an Initiator to inform the DAT
drive that an error, such as a parity error, occurred.
drive operation. All pending data and status for the issuing
Initiator are cleared, and the drive goes back to Bus Free
Phase. No status or ending message is sent for the
operation. Any previously set modes are not changed.
This message is sent from the DAT drive to indicate the
last message it received was inappropriate or was not
implemented. The DAT drive sends Message Reject and
then goes to the Message-In phase prior to requesting
additional message bytes from the Initiator. This sequence
provides an interlock so the Initiator can determine which
message was rejected.
This message is sent from an Initiator in response to a DAT
drive request for a message, when the Initiator does not
currently have any other valid message to send.
incorrect parity was detected. See the following table.
This message is sent from the DAT drive to inform an
Initiator that execution of a linked command is completed
and that status was sent. The Initiator can then set the
pointers to the initial state for the next linked command.
Page 20 DDS-4 SCSI Interface Manual, Rev. B
Hex
Code Message Definition
0Bh
0Ch
Identify
Linked
Command
complete
with Flag
Bus Device
Reset
(80h
disable
Disconnect
/
Reconnect)
(C0h
enable
Disconnect
/
Reconnect)
This message is sent from the DAT drive to inform an
Initiator that execution of a linked command with the Flag
bit set to one, is completed and that status was sent. The
Initiator then sets the current pointers to the initial state of
the next linked command. Typically, this message is used
to cause an interrupt in the Initiator between two linked
commands.
This message is sent from an Initiator to direct the drive to
clear all current commands; it forces the drive to an initial
state with no operations pending for any Initiator. On
recognizing this message, the drive moves to the Bus Free
Phase. All modes are reset to the default state.
These messages are sent by either the Initiator or DAT
drive to establish the physical path between them. The
physical path connection indicates that both the Initiator
and DAT drive have message passing capability. Bit 7 is
set to one to distinguish these messages from other
messages. Bit 6 is set to one by the Initiator indicating the
Initiator has the ability to accommodate disconnection and
reconnection. Bits 5 through 3 are reserved (set to zero).
Bits 2 through 0 specify the logical unit number that must
be 0 for the DAT drive. When Identify is sent from the drive
to an Initiator during reconnection, an implied Restore
Pointers message is performed by the Initiator prior to
completion of this message.
The following table describes the operation for each of the possible cases of parity
error.
State or Phase Description
Bus Free State
Arbitration Phase
Selection Phase
Selection, Message
Out Phase (Identify
Message)
Reselection,
Message In Phase
(Identify Message)
The drive does not detect nor react to parity errors on the
SCSI bus while the drive is in a bus free state.
The drive does not detect nor react to parity errors on the
SCSI bus while arbitration is being performed.
The drive does not detect nor react to parity errors on the
SCSI bus while the drive is being selected.
If the drive detects a parity error while the host is sending
an Identify message, the drive will attempt one retry then
go to Bus Free.
If the drive is attempting to reconnect to the host and the
host asserts ATN because it detected an error, the drive
• Switches the host to the Message Out Phase.
• Waits for the host to send a 09 (Parity Error Message)
The drive then performs the retry option by:
• Switching the host to the Message In Phase.
• Resending the Identify Message
SCSI Interface Description Page 21
State or Phase Description
Command Phase
If the drive detects a parity error while the host is
transferring a CDB, the drive
• Terminates the transfer.
• Switches the host to the Status Phase and sends a
Check Condition.
• Switches the host to the Message In Phase and sends a
Command Complete.
• Sets the Sense Key = B and ASC/ASCQ = 47-00 (SCSI2 only).
Data In Phase
Data Out Phase
If the host detects a parity error while data is being
transferred from the drive and asserts ATN, the drive
• Terminates the transfer of data.
• Switches the host to the Message Out Phase.
• Waits for the host to send an 05 (Host Detected Error).
• Switches the host to the Status Phase and sends a
Check Condition.
• Switches the host to the Message In Phase and sends a
Command Complete.
• Sets the Sense Key = B and ASC/ASCQ = 48-00 (SCSI2 only).
If the drive detects a parity error while the host is
transferring data, the drive
• Terminates the transfer of data.
• Switches the host to the Status Phase and sends a
Check Condition.
• Switches the host to the Message In Phase and sends
Command Complete.
• Sets the Sense Key = B and ASC/ASCQ = 47-00 (SCSI2 only).
Status Phase
Message In Phase
If the host is in the Status phase and detects an error in the
status byte and asserts ATN, the drive:
• Switches the host to the Message Out Phase and waits
for the host to send 05 (Initiator Detected Error).
• Switches the host to the Status Phase and sends a
Check Condition.
• Target sends restore pointers and resends Status.
If the host is in the Message In Phase and detects an error
on a message bytes an asserts ATN, the drive:
• Switches the host to the Message Out Phase.
• Waits for the host to send 09 (Parity Error Message).
• Switches the host to the Message In Phase and resends
the message.
Page 22 DDS-4 SCSI Interface Manual, Rev. B
State or Phase Description
Message Out
Phase
If the host is in the Message Out Phase and sends a
message bytes and the drive detects a parity error, the
drive:
• attempts one retry then goes to Bus Free.
An Initiator that accommodates disconnect/reconnect can indicate this capability to
the DAT drive during the Selection phase by asserting both its own Initiator SCSI ID
bit as well as the DAT SCSI ID bit (allows the DAT drive to know with which Initiator
to reconnect). The Initiator must also assert ATN before exiting the Selection phase
(prior to releasing SEL) and send an Identify message out of C0h to the DAT drive.
This sequence causes the drive to enter the Message-Out phase when the
Selection phase completes.
The first message sent by the host after the Selection phase is an Identify message.
Under normal conditions, the first message sent by the DAT drive after a
Reselection phase is also Identify. Under certain exceptional conditions, the host
may send the Abort message or the Bus Device Reset message instead of Identify
as the first message.
SCSI Interface Description Page 23
SCSI Status Code Descriptions and Definitions
The status code format is shown in the following table. A status byte is sent from the
DAT drive to the Initiator during the Status phase at the termination of each
command unless the command is cleared by an Abort message, by a Bus Device
Reset message, or by a Reset condition.
BITS 7 6 5 4 3 2 1
Content 0 0 0 Status Code 0
The DAT drive uses the 4-bit status codes shown in the following table.
Bits
4 3 2 1 0 Definition
0 0 0 0 X Good Status
0 0 0 1 X Check Condition
0 1 0 0 X Busy
1 0 0 0 X Intermediate Status
1 1 0 0 X Reservation Conflict
Descriptions of the five supported SCSI status codes are given in the following table.
Message
Busy 4h 08h This status (DAT drive busy) is returned when the drive is
Check
Condition
Good
Status
Intermedi
ate Status
Reservation
Conflict
Status
Code
1h 02h Any error, exception, or abnormal condition that causes the
0h 10h This status indicates that the DAT drive successfully completed
8h 10h This status is returned for every command in a series of linked
Ch 18h This status is returned when a device attempts to access a DAT
Byte 0
Status
Phase
Definition
unable to accept a command from the Initiator. (For example, it
is returned during an intermediate disconnect or after an
immediate command has completed and the request function is
not done.) The normal Initiator recovery action is to issue the
command at a later time.
sense data to be set, causes a Check Condition status. A
REQUEST SENSE command should be issued following a
Check Condition status, to determine the nature of the condition.
the command.
commands except the last command unless an error, exception,
or abnormal condition causes either Check Condition or
Reservation Conflict status to be set. If this status is not
returned, the chain of linked commands is broken, and no
further commands in the series are executed.
drive when it is reserved for access to another device.
Page 24 DDS-4 SCSI Interface Manual, Rev. B
Attention Condition
The Attention Condition allows an Initiator to inform a Target that the Initiator has a
message to send. The DAT drive may read this message at its convenience by
performing a Message-Out phase. The Initiator creates the Attention Condition by
asserting the ATN signal at any time except during the Arbitration or Bus Free
phases. The DAT drive checks to see if ATN is set at every phase change. If ATN is
set, the drive goes into the Message-Out phase. The Initiator may de-assert the
ATN signal at any time. Normally, the Initiator de-asserts ATN during or before the
last REQ/ACK handshake of the Message-Out phase. The Attention signal must be
present prior to a phase change to allow the DAT drive time to respond with a
Message-Out phase at the phase change.
Reset Condition
The Reset Condition takes precedence over all phases and conditions and is used
to immediately terminate operation and clear all SCSI devices from the bus. Any
SCSI device can create a reset condition by asserting a RST signal for a minimum
reset hold time of 25 microseconds. The DAT drive never asserts reset. During the
Reset Condition, all SCSI devices release all SCSI signals (except RST) within a
Clear Reset Delay (800 nSec) of the transition of RST to low. The Bus Free phase
always follows the Reset Condition.
When a reset is issued to the DAT drive, the SCSI bus clears all uncompleted
commands, releases all SCSI device reservations, sets the DAT drive to default
modes, and returns to the Bus Free phase.
Unit Attention Condition
The Unit Attention condition in the DAT drive typically results from the following
conditions:
• A Reset was previously issued to the DAT drive.
• The DAT drive has just been powered on.
• The cartridge was removed when the tape is positioned away from BOT.
• The cartridge was removed when the tape is positioned at BOT following a
LOAD command.
• A cartridge has been inserted since the previous bus reset or power-on.
• A log exception condition occurred.
• The mode pages have been changed
The Unit Attention Condition persists for an Initiator until the Initiator issues a
command other than Inquiry for which the DAT drive returns with a Check Condition
Status. If the next command from that Initiator following the Check Condition Status
is Request Sense, then the unit attention sense key is returned.
If the Inquiry Command is received from an Initiator with a pending Unit Attention
Condition before the DAT drive reports Check Condition Status, the DAT drive
performs the Inquiry Command and does not clear the Unit Attention Condition.
SCSI Interface Description Page 25
If the Request Sense Command is received from an Initiator with a pending Unit
Attention Condition before the DAT drive reports Check Condition Status, the DAT
drive reports unit attention sense key and clears the Unit Attention Condition for that
Initiator.
Buffered Mode
Buffered Mode allows the most efficient operation of the DAT drive. The drive
defaults to Buffered mode. In this mode, the drive signals Command Complete when
all requested data for a WRITE command has been transferred from the host to the
DAT buffer. This mode provides data to maintain operation while the host readies a
new WRITE Command.
If an error occurs in writing data to the tape after the DAT drive signals Command
Complete, an error status is sent on the next Command issued.
Immediate Function
For Initiators that do not support the disconnect feature, the Immediate bit provides
a means of releasing the bus while the drive is busy completing a function such as
repositioning the tape. If a command is sent by the Initiator after a previous
Immediate Command was accepted, the drive continues the Immediate Function it
is currently performing and returns a Busy Status for the new command.
An immediate bit of zero means that the status is returned to the Initiator when the
operation is completed. (For example, the status is returned when the tape has been
repositioned.) An Immediate bit of one means that the status is returned to the
Initiator as soon as the function is started.
Residual Length Function
When performing a WRITE command, the drive returns a Good Status and
Command Completion Message when the last byte requested by the command is
placed in the Data Buffer, rather than when it is written onto tape. If an error occurs
while data is being written onto tape, the drive calculates the Residual Length and
places this value in the information bytes of the Sense Data Block. Also, the
Residual Length functions for other commands, such as READ and SPACE.
Residual Length is calculated by: RL = TL - AL
Where :
AL (Actual Length) = Blocks transferred from the host to the DAT drive across
TL (Transfer Length) = The Transfer Length from bytes 2-4 of the WRITE
RL (Residual Length) = The amount of blocks or bytes not written to tape.
the SCSI bus.
command (Request Transfer Length).
Page 26 DDS-4 SCSI Interface Manual, Rev. B
Disconnect/Reconnect Function
When the drive is performing a task not requiring communication with the Initiator or
when the DAT drive determines that a relatively long time has passed with no bus
activity, it disconnects from the SCSI bus. Examples are:
• When rewinding the tape.
• When writing to the tape and the buffer is full.
• When reading from the tape and the buffer is empty.
• When spacing, locating, or generally performing any tape motion when data
cannot be transferred on the SCSI bus.
During the time the Target is disconnected for one of these functions, the bus is free
for use by other devices. Both disconnect and reconnect are initiated by the Target.
When the Initiator first selects the drive, it sends an Identify Message indicating that
it is allowing the drive to disconnect and reconnect and to be capable of supporting
messages other than Command Complete (and the host set its own ID during
selection). To disconnect from the bus, the DAT drive performs the following
procedure:
1. The drive can send a Save Data Pointers Message if the disconnection function
was a data transfer.
2. The drive sends a Disconnect Message indicating it is going to disconnect.
3. The drive disconnects from the bus by de-asserting BSY and releasing control
of all bus signals.
The bus is now free for an Initiator to select any device on the bus, including the
drive that initiated the disconnect. The drive will respond to selection by another
Initiator.
If the DAT drive is selected while disconnected, it only allows the following actions:
• If the command is from a different initiator or is from the same initiator but to a
different LUN, the DAT drive accepts the command and immediately returns
Busy Status.
• Immediately following the selection, the Initiator may send the Identify, No Op,
Abort, or Bus Device Reset messages to the drive.
• If the command is from the same initiator to the same LUN, the current
command terminates with a Check Condition and an Abort Sense Key.
When the disconnected drive is ready to reconnect with the Initiator, it does the
following.
1. It monitors the bus waiting for a Bus Free Phase to occur. When a Bus Free
Phase is sensed, the DAT drive arbitrates for the bus.
2. If it wins arbitration, the DAT drive then attempts to reselect the Initiator. If the
Initiator fails to respond in 250 mSec, the drive drops all bus signals and allows
the bus to again enter the Bus Free Phase. The drive then repeats the
attempted Arbitration.
SCSI Interface Description Page 27
3. When the DAT drive has successfully reselected the Initiator, it sends an
Identify Message to reestablish the path between the drive and the Initiator.
This message is always 80h because the DAT drive is initiating the reselection
and is always LUN 0. A Restore Pointers Message is implied when the DAT
drive sends an Identify Message to the Initiator. The Initiator responds
accordingly.
SCSI Memory Address Pointers
SCSI provides for two sets of three pointers within each Initiator. When a physical
path is established with a host, and this path can accommodate disconnection and
reconnection, the host must ensure that its Current Pointers for the path are equal to
the Saved Pointers in the DAT drive. An implied Restore Pointers Operation occurs
in the host as a result of a connect or reconnect.
Current Data Pointers
Current data pointers, also known as Active Pointers, are used to represent the
state of the interface and point to the next Command, Status or Data byte to be
transferred between the memory of the Initiator and the Target. Each Initiator may
have only one set of Current Pointers. The Current Pointers are used by the Target
currently connected to the Initiator.
Saved Data Pointers
Whether or not a currently active device is currently connected, it has one set of
Saved Data Pointers. This set includes Command, Status and Data Pointers that
point to the Command Descriptor Block, Status Area and Data Area, respectively,
for that device.
The Saved Pointer continues to point to the start of the Data Area until the Target
reconnects to the Initiator. In response to the implied Save Data Pointer message,
the Initiator replaces the Current (active) Data Pointer with the value of the Saved
Data Pointer.
Early Warning Function
Early Warning on the DAT drive is a logical warning given when ten megabytes of
storage space remain on the tape. The position is calculated by the drive. When this
physical position is reached on a tape, the following occurs.
1. Data transfers from the host are terminated at the next block boundary.
2. All data remaining in the drive buffer is written to the tape if the Synchronize at
EW bit is set.
3. The command completes with a Check Condition and a 40h Sense data
meaning EOM and no Sense Key.
4. Subsequent WRITE commands write data and complete with check condition
with EOM Status and No Sense Key until the physical tape end is encountered.
Page 28 DDS-4 SCSI Interface Manual, Rev. B
Error Reporting
Soft Errors
Soft errors are generally tape-quality related and occur more frequently during write
operations than during read operations. Soft errors indicate repeated attempts by
the drive to read or write data on the tape. Some soft errors are normal, but an
increase in the usual count can indicate deteriorating tape quality. If the soft error
count remains higher than normal, clean the read/write heads. If this procedure does
not clear the problem, change to a new tape cartridge.
Hard Errors
If a hard error (unrecoverable error) occurs during operation, the drive terminates
operation immediately and returns a Check Condition. The Initiator should cease
any further read or write functions and issue a Request Sense Command to
determine the type of error.
When the drive detects a write error, it attempts to rewrite the data up to 127 times.
After the 127th attempt, the error is considered unrecoverable and the operation
terminates. In that case, the appropriate LED on the drive front panel flashes rapidly.
When a hard error is encountered, replace the tape with a new cartridge and repeat
the function or clean the heads with a cleaning cartridge.
Variable and Fixed Mode Recording
The DAT drive can write either fixed or variable block sizes. The recording mode is
determined by the Fixed bit in the SCSI WRITE and READ commands.
If the Fixed bit is set, the MODE SELECT command sets the size of the next block
or multiple blocks to be written with the next WRITE command. When a WRITE
command is issued with the Fixed bit set, the current block size is implemented. The
transfer length specifies the number of blocks to be written with this size. If a WRITE
command is issued with the Fixed bit set and the current block size set to 0, the
DAT drive returns a Check Condition with Illegal Request Sense Key. When writing
with the Fixed bit set, each WRITE command specifies the number of contiguous
blocks to be written of a fixed size, resulting in fixed-mode blocks.
If the Fixed bit is reset, then only one block can be written on the tape per SCSI
WRITE command CDB. In that case, the WRITE command CDB transfer length
specifies the size of the block to be written in bytes. With the Fixed bit reset, the
current block size specified with the last MODE SELECT command is ignored.
Setting the block size to 0 in the MODE SELECT page descriptor is not required.
Therefore, with the Fixed bit reset, each SCSI WRITE command may specify a
different byte count, resulting in variable-mode blocks.
The host may switch between fixed and variable mode recording. By issuing the
MODE SELECT command to specify different block sizes, blocks can be written to
the tape with different block sizes in the fixed mode. Also, the host may change the
block size after BOM, allowing on-the-fly block-size changes.
The READ command Fixed bit also specifies fixed or variable mode.
SCSI Interface Description Page 29
When reading in variable mode, the host must know the size of the block to be read
from the tape in advance in order to avoid causing the DAT drive to return a Check
Condition with Incorrect Length indicated in the Sense data (ILI). Also, the data
transfer may be truncated (cut off) when the recorded block does not match the
transfer length in variable mode or the current block size in fixed mode.
The SCSI-2 READ command includes a SILI bit to Suppress ILI Check conditions.
When the SILI bit is set, the host usually specifies the maximum block size before
reading so that the data blocks are not truncated, and no Check Conditions are
generated.
The SCSI READ BLOCK LIMITS command returns the minimum and maximum
block sizes that the DAT drive can support. The Block Limits data is not modified to
reflect the current mode of writing—fixed or variable. The Block Limits returned data
is not modified to reflect the current block size for the next fixed-mode WRITE. The
MODE SENSE command is used for that purpose.
Autoloader Operation
General SCSI Information
The Scorpion 240 is, in SCSI-3 terminology, a tape drive with an independent
medium changer: SCSI Stream Device, i.e., tape, commands are addressed to
Logical Unit Number 0 and SCSI Medium Changer (SMC) commands to LUN 1.
SCSI Primary commands (SPC) may be addressed to either LUN.
The Standard Inquiry Data of LUN 0 reports a Peripheral Device Type of 01h
(sequential access device) and a MCHNGR bit of 0. LUN 1 reports a Peripheral
Device Type of 08h (medium changer device).
(By contrast, Scorpion 240 is not a SCSI-3 attached medium changer device. These
devices have only logical unit 0, and all commands are addressed to that LUN. Their
Inquiry data has a Peripheral Device type of 01h and MCHNGR bit of 1.)
Sense data is maintained for each Initiator-Target nexus; in other words, while the
Request Sense command may be addressed to either LUN, the same data is
reported for each LUN.
SCSI Medium Changer Commands
Element Addresses
SMC commands control the process of moving cassettes (volumes) between the
magazine and the drive. They refer to the components of the Autoloader by element
addresses. (Element addresses are not to be confused with SCSI ID or LUN.)
• The medium transport element (MTE) is the elevator, which actually moves the
cassettes; its element address is 0. The MTE may not be either the source or
destination when a cassette is moved.
• The data transfer element (DTE) is the DAT drive itself. Its address is 1.
• The storage elements (SE) are the slots in the magazine. Slot 1 has element
address 2, slot 2 has address 3, etc. through address 7 for slot 6.
Page 30 DDS-4 SCSI Interface Manual, Rev. B
Move Medium Command
The SMC Move Medium command moves a cassette from one address to another
by specifying a Source Address and a Destination Address. Scorpion 240 requires
that one of the addresses be 1, the DTE (drive). This command cannot be used to
move a cassette from one magazine slot directly to another. However, when a
cassette is moved from the DTE, it is not required that it be placed back in the slot
from which it came; the Source Storage Element Address will then indicate the last
slot it occupied.
Exchange Medium Command
The SMC Exchange Medium command is used to swap the tape in the DTE for
another one. The command specifies a Source Address, First Destination Address,
and Second Destination Address. Scorpion 240 requires that the Source and
Second Destination addresses be storage elements (magazine slots) and the First
Destination Address must be the DTE. The command moves the cassette in the
DTE to the Second Destination Address and the cassette in the Source Address to
the DTE.
Both of the above commands check that there is a cassette in the source
address(es) and no cassette in the destination address before attempting to execute
the command. If an address is invalid or an element is inappropriately empty or full,
the autoloader mechanism will not move.
Furthermore, if one of the above commands is invoked with a cassette in the DTE,
all buffered data will be written to the tape and the tape positioned to PBOT before it
is removed.
Read Element Status Command
The current position of cassettes in the Autoloader can be obtained by a SCSI
initiator by issuing a Read Element Status command. During all movements of
cassettes inside the Autoloader, the previous storage element address of the
cassette in the magazine is tracked, making it unnecessary for this command ever
to scan the magazine or otherwise move the mechanism. This address is reported in
the Source Storage Element Address of the element descriptor; the address is valid
when the SVALID bit is one.
Initialize Element Status Command
The host computer can direct the Autoloader to scan its magazine by issuing an
Initialize Element Status command. If there is a cassette in the DTE, it will be
returned to its magazine slot before the scan begins, and it will remain in the slot
after the scan.
SCSI Interface Description Page 31
General Operation
• The Reserve and Release commands may be addressed to either LUN, but
affect both LUs as a unit. In other words, if an initiator has reserved either LU,
then both LUs will reply with a status of Reservation Conflict to most SCSI
commands from other initiators. This is departure from the SCSI standard, but it
prevents the situation of each LU being reserved by a different initiator.
• Busy status affects the Autoloader as one unit. LUN 0 is busy when LUN 1 is
busy, and vice versa.
• The Ready and Not Ready conditions are reported separately for the two LUs.
For example, if the loader contains a magazine, but no cassette has yet been
loaded into the drive, then LUN 1 will be ready and LUN 0 will not. Whenever
the drive contains a cassette, both LUs will be ready.
• Reset conditions affect the Autoloader as one unit. A Bus Device Reset or SCSI
Bus Reset initializes both LUN 0 and LUN 1.
• An Unload command to LUN 0 ejects the cassette from the drive and stores it in
the magazine slot in which it last resided. If the Autoload feature is enabled, the
next cassette in sequence is then loaded. An Unload command to LUN will
return a cassette in the drive (if any) to its original magazine slot and will then
eject the magazine.
• Following is a list of Request Sense sense key/additional sense code/qualifier
values that are particular to the medium changer. A complete list appears in the
Request Sense command discussion in Chapter 3.
Sense Code Title
05/21/01 Invalid Element Address
05/3B/0D Medium Destination Element Full
05/3B/0E Medium Source Element Empty
02/3A/00 Medium Not Present (Reported if there is no magazine in the
06/28/00 Not Ready To Ready Transition (Reported after magazine is
Loading and Initialization
The Scorpion 240 autoloader uses a six-slot magazine that is keyed to prevent
incorrect insertion of cassettes into the magazine and incorrect insertion of the
magazine into the loader. The slots are numbered 1 through 6, with 1, 2, and 3 on
one side of the magazine and 4, 5, and 6 on the other. When a magazine is inserted
into the loader, the door folds down and out of the way and the magazine should be
pushed in gently until the loader mechanism begins to draw the magazine in; the
door must be allowed to close.
loader or if the magazine is empty and the drive contains no
cassettes)
inserted or if cassettes are loaded or changed by front panel
command. Cassette movement by SCSI command will not
result in this condition.
Page 32 DDS-4 SCSI Interface Manual, Rev. B
When the magazine has been completely moved into the loader, the process of
initialization begins, in which the magazine is scanned to determine which slots
contain cassettes. In fact, initialization is performed at three times:
• Insertion of a magazine into a drive that is already powered up
• Power up with a magazine already inside
• Receipt of an Initialize Element Status SCSI command
Front Panel Operation
The front panel contains an LCD screen that displays the autoloader status. There
are also three pushbuttons, Eject, Select, and Enter, which are used to perform
various operations:
• Display drive firmware version and SCSI ID
• Select language for LCD messages
• Adjust LCD contrast
• Enable and disable Autoload mode
• Load a specific cassette
• Eject the magazine
• Lock and unlock the door
The menu can only be accessed when there is no magazine in the drive. Loading
cassettes and ejecting the magazine can be performed only when there is a
magazine in the drive.
To access the menu, first eject the magazine, if any, and then press and hold the
Select button until the menu appears on the LCD; this will take about five seconds.
Then press the Select button momentarily to scroll among the menu items. To
choose one, press the Enter button. To return to the regular display, momentarily
press the Eject button.
Cassette Selection
When the Select and Enter buttons on the Autoloader are used to select a cassette
in the magazine, it is loaded into the drive. If there is already a cassette in the drive,
it will first be placed back in the slot from which it came. If the drive is executing a
SCSI command, such as a Read or Write, that and all pending commands will be
terminated with Not Ready / Operator Medium Removal Request (02/5A/01) status.
Automatic Loading Mode
The front panel can be used to enable or disable automatic loading mode. When
enabled, certain events cause the next cassette in the magazine to be loaded
without either a SCSI or front panel command:
• Ejection of a cleaning cassette after a successful cleaning. (Cleaning failure due
to an expired cleaning tape will not autoload the next tape.)
• Ejection of a data cassette by a SCSI Unload command to LUN 0.
• Insertion of a magazine. After the magazine is scanned, the first cassette will be
loaded.
SCSI Interface Description Page 33
The following events will NEVER cause automatic loading:
• Ejection of a cleaning cassette after cleaning failed.
• Ejection of a data cassette by a SCSI Move Medium command.
• Ejecting the last cassette in the magazine (i.e., the cassette in the highest-
numbered occupied slot).
Magazine Ejection
The Eject button on the Autoloader can be used in two modes: regular and
emergency eject. When the button is pressed for less than five seconds, a regular
eject begins. The current SCSI command is completed and buffered data is written
to the tape. Pending commands are terminated with Not Ready / Operator Medium
Removal Request (02/5A/01) status, the cassette is returned to its slot in the
magazine, and the magazine is ejected.
If the Eject button is held for more than five seconds, emergency eject begins. All
commands are terminated within 35 seconds, the cassette in the drive is placed in
the magazine, and the magazine is ejected. If there is no cassette in the drive,
magazine ejection will occur immediately. Pending commands are terminated with
Not Ready / Operator Medium Removal Request (02/5A/01) status.
Caution: If the eject button is pressed for an emergency eject while a write
Stray Tape Recovery
If the magazine is forcibly removed from the loader while it is powered off, a stray
tape may be left in either the drive or the elevator. In this case, the recovery
procedure is to insert an empty magazine into the Loader. The Loader will attempt to
scan the magazine, place the stray tape into the lowest-numbered empty slot, and
eject the magazine.
operation is in progress, the tape may not be appendable.
Page 34 DDS-4 SCSI Interface Manual, Rev. B
X
X
X
X
X X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X X
X X
X X
Commands
3
Introduction
This chapter describes the SCSI commands for the Seagate DDS-4 and DAT 72
tape drives.
Command Reference List
This chapter describes the implementation of the ANSI SCSI command
specifications. The following table lists the commands numerically by opcode.
Commands specific for the AutoLoader are noted.
Code SCSI-2 Autoloader Command
01h
02h
03h
05h
07h
08h
0Ah
0Ch
10h
11h
12h
13h
15h
16h
17h
19h
1Ah
1Bh
1CH
1Dh
1Eh
2Bh
34h
3Bh
3Ch
4Ch
4Dh
44h REPORT DENSITY SUPPORT
A5h
A6h
B8H
REWIND
REQUEST BLOCK ADDRESS
REQUEST SENSE
READ BLOCK LIMITS
INITIALIZE ELEMENT STATUS
READ
WRITE
SEEK BLOCK
WRITE FILEMARKS
SPACE
INQUIRY
VERIFY
MODE SELECT
RESERVE UNIT
RELEASE UNIT
ERASE
MODE SENSE
LOAD/UNLOAD
RECEIVE DIAGNOSTIC RESULTS
SEND DIAGNOSTIC
PREVENT/ALLOW MEDIUM REMOVAL
LOCATE
READ POSITION
WRITE DATA BUFFER
READ DATA BUFFER
LOG SELECT
LOG SENSE
MOVE MEDIUM
EXCHANGE MEDIUM
READ ELEMENT STATUS
Commands Page 35
Conventions
The commands in this chapter are listed in alphabetical order. Each command is
described; its Command Descriptor Block (CDB) illustrated; and the Completion
Status is given. Bits and fields defined in the ANSI SCSI-2 document that are not
used by the drive are not described in this document. Rather, those bits and fields
are shown as 0, 1, or X (ignores), as appropriate. Bits and fields that are supported
by the drive are described.
Command Descriptor Blocks (CDBs)
A host makes request of the DAT drives by sending a command descriptor block
(CDB); some commands also require a parameter list. If the CDB or the parameter
list contains an invalid parameter, the drive terminates the command without altering
the medium.
Command Descriptor Block Formats
The SCSI-2 six-byte CDBs are arranged in the format shown in the following table.
Byte
0 Group Code | Command Code
1 LUN | Command Dependent
2 Command Dependent
3 Command Dependent
4 Command Dependent
5 vendor unique 0 0 0 0 Flag Link
Bits
7 6 5 4 3 2 1 0
The SCSI-2 ten-byte CDBs are arranged in the format shown in the following table.
Byte
0 Group Code | Command Code
1 LUN | Command Dependent
2 Command Dependent
3 Command Dependent
4 Command Dependent
5 Command Dependent
6 Command Dependent
7 Command Dependent
8 Command Dependent
9 Vendor Unique 0 0 0 0 Flag Link
Bits
7 6 5 4 3 2 1 0
Page 36 DDS-4 SCSI Interface Manual, Rev. B
Command Descriptor Block Field Descriptions
The command descriptor block fields are described in the following table.
Field Name Byte Bits Description
Group Code 0 5-7 This field indicates the SCSI command groups used
Command Code * 0 0-4 This field indicates which of 32 command codes is sent.
NOTE: The Group Code and Command Code are used
together to indicate the operation code.
Logical Unit 1 5-7 The LUN must be set to zero.
Command
Dependent
Vendor Unique 5 6-7 These bits are reserved. In the following command sections,
Flag and Link bits 5 0-1 These bits are used by all commands and described in the
1
2-4
0-4
0-7
See specific command.
these bits are shown by Xs.
following subsection, Flag and Link Bit Descriptions.
Flag and Link Bit Descriptions
The Link bit enables a chain of commands to execute sequentially without the
initiator reselecting the target. By setting the Link bit, the Initiator desires an
automatic link to the next command on successful completion of the current
command.
The Flag bit allows the initiator to track milestones in the linked command process
by requesting a linked Command Complete (with flag) message from the drive.
If the CDB has the Link bit set, the drive completes the command and takes the
following actions:
1. Goes to the Status phase and posts Intermediate Good status.
2. Goes to the Message-In phase and posts one of two possible messages
according to the setting of the Flag bit.
3. Proceeds to the Command Out phase to request the next CDB.
Command linking continues as long as the Link bit is set in successive CDBs.
Linking stops when the drive detects an error and posts a Busy or Check Condition.
The Flag and Link bits are used by all commands. These bits are defined in the
following table.
If the
Link bit
is . . .
1 0 The Target sends an Intermediate Good Status, then a Linked
1 1 The Target returns Immediate Status and then sends a Linked
0 0 The Target performs normal termination and goes to Bus Free.
0 1 Invalid, the Target returns Illegal Request.
and the
Flag bit
is . . .
the result is . . .
Command Complete message, and then requests the next command.
Command Complete (with flag) message, then requests the next
command.
Commands: Erase (19h) Page 37
ERASE (19h)
The ERASE command creates an EOD (end-of-recorded data) marker at the current
position of the tape when the command is issued, thus rendering the remainder of
the tape unreadable.
ERASE Command Descriptor Block
The following table shows the layout of the CDB.
Byte
0 0 0 0 1 1 0 0 1
1 0 0 0 0 0 0 IMMED Long
2 0 0 0 0 0 0 0 0
3 0 0 0 0 0 0 0 0
4 0 0 0 0 0 0 0 0
5 vendor unique* 0 0 0 0 Flag Link
* This vendor-unique field is reserved and must be set to zero.
Bits
7 6 5 4 3 2 1 0
Command Descriptor Block Field Description
The following table defines the fields in the CDB.
Field Name Bytes Bits Description
IMMED 1 1 If the IMMED (Immediate) Bit is 1, status is to be returned as
Long 1 0 If the Long bit is 0 or 1, a short ERASE is performed, and EOD
soon as the ERASE command is accepted and verified. If the
IMMED bit is 0, status is not returned until the tape is erased.
is recorded at the current tape position.
Completion Status
The Completion Status for the ERASE command is shown in the following table.
Code Message Description
00h
02h
Good Status
Check
Condition
• The drive remains in any previously set modes.
• The drive is ready to perform any appropriate command.
Extended Sense Byte 02h
Code Message Description
02h Not Ready No cartridge is inserted in the drive.
04h Hardware Error Parity error on SCSI bus or drive
hardware failure detected.
05h Illegal Request
06h Unit Attention
this command.
07h Write Protect The cartridge in the drive is write protected.
• The CDB contains an invalid bit.
• Both IMMED bit and Link bit are set to 1.
• Cartridge was changed prior to accepting
• Drive was reset prior to this command.
Page 38 DDS-4 SCSI Interface Manual, Rev. B
EXCHANGE MEDIUM (AutoLoader Only) (A6h)
The EXCHANGE MEDIUM command provides a means to exchange the cartridge
in the source element with the cartridge located in the drive. This command provides
the initiator a single-command alternative to two MOVE MEDIUM commands when
removing a cartridge from the DAT drive and inserting another cartridge into the
DAT drive.
EXCHANGE MEDIUM Command Descriptor Block
The following table shows the layout of the CDB.
Byte
0 1 0 1 0 0 1 1 0
1 LUN 0 0 0 0 0
2 0 0 0 0 0 0 0 0
3 0 0 0 0 0 0 0 0
4 0 0 0 0 0 0 0 0
5 Source Address
6 0 0 0 0 0 0 0 0
7 0 0 0 0 0 0 0 1
8 0 0 0 0 0 0 0 0
9 Source Destination
10 0 0 0 0 0 0 0 0
11 vendor unique* 0 0 0 0 Flag Link
Bits
7 6 5 4 3 2 1 0
* This vendor-unique field is reserved and must be set to zero.
Command Descriptor Block Field Descriptions
The source address field specifies the location from which the cartridge is taken; the
destination address field specifies the location to which the cartridge is moved.
The destination address must represent an empty storage element; the cartridge in
the drive is first moved to this location. The source address must represent a full
storage element; this cartridge is then moved to the drive.
The following table shows the addresses of the various elements.
Source/
Destination
Address
0 0
1 Drive
2 Magazine slot 1
3 Magazine slot 2
4 Magazine slot 3
5 Magazine slot 4
6 Magazine slot 5
7 Magazine slot 6
Source/
Destination Element
Commands: Exchange Medium (A6h) Page 39
Completion Status
The following table lists the completion status for the EXCHANGE MEDIUM
command.
Code Message Description
00h Good
Status
02h Check
Condition
• The cartridges have been moved to the specified destinations.
the cartridge in the drive is positioned at BOT.
• The drive is ready for another command.
Extended Sense Byte 02h
Code Message Description
02h Not Ready No magazine is present.
04h Hardware Error Parity error on SCSI bus or drive
hardware failure detected.
05h Illegal Request
.
the same.
06h Unit Attention
this command.
• The source element is empty.
• The drive is empty.
• The destination element is full
• The source and destination address are
• Cartridge was changed prior to accepting
• Drive was reset prior to this command.
Page 40 DDS-4 SCSI Interface Manual, Rev. B
INITIALIZE ELEMENT STATUS (AutoLoader Only) (07h)
The INITIALIZE ELEMENT STATUS command causes the AutoLoader to eject the
cartridge that is in the DAT drive (if any) and return it to the slot in the magazine.
Then the magazine is initialized, checking each element for cartridges or any other
status relevant to that element. This command performs the same magazine
initialization as is performed by mounting a magazine and powering-up the drive
except that no cartridge is inserted into the drive.
INITIALIZE ELEMENT STATUS Command Descriptor Block
The following table shows the layout of the CDB.
Byte
0 0 1 0 0 0 1 1 1
1 LUN 0 0 0 0 0
2 0 0 0 0 0 0 0 0
3 0 0 0 0 0 0 0 0
4 0 0 0 0 0 0 0 0
5 vendor unique* 0 0 0 0 Flag Link
Bits
7 6 5 4 3 2 1 0
* This vendor-unique field is reserved and must be set to zero.
Completion Status
The following table lists the completion status for the INITIALIZE ELEMENT
STATUS command.
Code Message Description
00h Good Status • The magazine has been initialized and positioned.
02h Check Condition Extended Sense Byte 02h:
• The drive is ready to accept another command.
Code Message Description
02h Not Ready No magazine is present.
04h Hardware Error Parity error on SCSI bus or drive
hardware failure detected.
05h Illegal Request The CDB contains an invalid bit.
06h Unit Attention
accepting this command.
• The cartridge was changed prior to
• The drive was previously reset.
Commands: Inquiry (12h) Page 41
Byte
INQUIRY (12h)
The INQUIRY command requests that the drive return parameter information. If an
INQUIRY command is received from an Initiator with a pending Unit Attention
Condition (before the drive reports Check Condition status), the drive performs the
INQUIRY command and does not clear the Unit Attention Condition.
INQUIRY Command Descriptor Block
The following table shows the layout of the CDB.
Bits
0 0 0 0 1 0 0 1 0
1 0 0 0 0 0 0 0 EVPD
2 Page Code
3 0 0 0 0 0 0 0 0
4 Allocation Length
5 vendor unique* 0 0 0 0 Flag Link
7 6 5 4 3 2 1 0
* This vendor-unique field is reserved and must be set to zero.
Command Descriptor Block Field Description
The following table defines the fields in the CDB.
Field Name Bytes Bits Description
EVPD 1 0 An Enable Vita Product Data (EVPD) bit of 1 specifies that
the drive return vital product data specified in the Page Code
field. An EVPD bit of 0 specifies that the drive return the
standard INQUIRY data.
Page Code 2 The Page Code field specifies which page of vital product
data information the target returns.
Allocation
Length
4 This field specifies the number of bytes that the Initiator
allocated for the response data. The drive transfers the
number of bytes specified up to a maximum of 36. An
Allocation Length of zero indicates that no data can be
transferred. This length is not considered an error.
Page 42 DDS-4 SCSI Interface Manual, Rev. B
Standard INQUIRY Data Format
The following table shows the standard data format for the INQUIRY command.
Byte
0 Peripheral Device Type
1 1 0 0 0 0 0 0 0
2 0 0 0 0 0 | ANSI Version (2)
3 0 0 0 0 0 0 1 0
4 Additional Length (1Fh)
5 0 0 0 0 0 0 0 0
6 0 0 0 0 0 0 0 Addr16
7 0 0 1 1 1 0 0 0
8 - 15 MSB—Vendor Identification
16 - 27 Product Name*
28 - 31 Product Revision Level*
32 - 35 Firmware Version*
Bits
7 6 5 4 3 2 1 0
Vendor Identification—LSB
* Indicates Seagate-specific field definitions. In the SCSI standard, bytes 16-31 are
grouped as “Product Identification”, and bytes 32-35 are captioned “Product
Revision Level”.
Standard INQUIRY Data Format Field Descriptions
The following table describes the data format fields.
Field Name Bytes Bits Description
Peripheral Device
Type
ANSI Version 2 0-2 This field contains 2 in compliance with ANSI standard SCSI-
Additional Length 4 This field indicates that 31 bytes of additional INQUIRY
Addr16 6 0 The Addr16 field contains 1, indicating that the device supports
Vendor
Identification
Product Name 16-27 This field contains 12 bytes of ASCII data: “DAT” followed by
Product Revision
Level
Firmware Version 32-35 This field contains the firmware version.
0 0-7 This field is set to 01h to indicate a sequential access device.
This field is set to 08h to indicate a medium changer device if
the drive is an Autoloader and the LUN=1.
Value 7Fh indicates that a logical unit is not present. This
value is returned in the case where an invalid LUN was
contained in the last Identify message sent by the Initiator.
2. The field may contain 3 for drives that are ANSI 2/3
compliant.
command parameters follow—beginning in Byte 05h. This
value is not changed if the Allocation Length in the CDB is too
small or too large to accommodate the entire response.
16-bit wide SCSI addresses, (SCSI-3)
8-15 This field contains 8 bytes of ASCII data: SEAGATE followed
by 1 space.
four spaces followed by “DAT72”
28-31 This field contains 4 bytes of ASCII data: "-xxx" the xxx is the
3-digit firmware revision part number; for example, "-001".
Note – See product manual # xxxxxxxxxxxx for details of alternate Inquiry data for
Vendor ID and Product Name.
Commands: Inquiry (12h) Page 43
The VPD pages supported are as follows:
Byte Description
0 Peripheral Device Type
1 Page Code (00h)
2 Reserved
3 Page Length (05h)
4 Supported Pages (00h)
5 Firmware Designation Page (03h)
6 Drive Serial Number Page (80h)
7 Controller Firmware Revision Page (C0h)
8 Drive Firmware Revision Page (C1h)
The Firmware Designation Page (03h) is defined as follows:
Byte Description
0 Peripheral Device type
1 Page Code (03h)
2 Reserved
3 Page Length (1Ch)
4-11 Reserved
12-15 REV Level
16-31 Reserved
The Controller Firmware Revision Page (C0h) is defined as follows:
Byte Description
0 Peripheral Device Type
1 Page Code (C0h)
2 Reserved
3 Page Length (13h)
4 ASCII Length (12h)
5-8 REV Level
9-18 Build Date (format yyyy/mm/dd)
19-22 OEM Model Number
The Drive Firmware Revision Page (C1h) is defined as follows:
Byte Description
0 Peripheral Device Type
1 Page Code (C1h)
2 Reserved
3 Page Length (04h)
4 ASCII Length (03h)
5-7 Drive Firmware Revision Level
Page 44 DDS-4 SCSI Interface Manual, Rev. B
INQUIRY Drive Serial Number Data Format Page
The following table describes the drive serial number, which is sent from the drive to
the host as data formatted in a parameter list.
Byte
0 Peripheral Device Type
1 Page code (80h)
2 Reserved
3 Page Length (7)
4 Drive Serial Number
5 Drive Serial Number
6 Drive Serial Number
7 Drive Serial Number
8 Drive Serial Number
9 Drive Serial Number
10 Drive Serial Number
Bits
7 6 5 4 3 2 1 0
Drive Serial Number Data Format Field Descriptions
The following table defines the fields in the Drive Serial Number Data Format.
Field Name Bytes Bits Description
Peripheral
Device Type
Page Code 1 0-7 The Page Code field is set to 80h to indicate that the
Page Length 3 0-7 This field specifies the length of the Drive Serial
Drive's Serial
Number
Completion Status
The Completion Status for the INQUIRY command is shown in the following table.
Code Message Description
00h Good Status
02h Check Condition Extended Sense Byte 02h:
0 0-7 This field is set to 01h to indicate a sequential
access device.
data following contain the Drive Serial Number.
Number.
4-10 0-7 This field contains the drive’s alphanumeric serial
number.
• The tape is not moved; the current position is maintained.
• The drive remains in any previously set modes.
• The drive is ready to perform any appropriate command.
Code Message Description
04h Hardware Error Parity error on SCSI bus or drive
hardware failure detected.
05h Illegal Request The CDB contains an invalid bit.
Commands: Load/Unload (1Bh) Page 45
LOAD/UNLOAD (1Bh)
The LOAD/UNLOAD command requests that the DAT drive enable or disable the
logical unit for further operations.
When a LOAD command is executed, the tape is positioned to LBOT and made
ready for the next command.
When an UNLOAD command is executed, the tape positions to BOT, and the
cartridge is ejected without an error condition. The UNLOAD command ejects the
cartridge even if a PREVENT MEDIA REMOVAL command is in effect.
For the Autoloader, an UNLOAD command ejects the cartridge from the drive into
the storage element positioned in front of the drive and then inserts the next
cartridge in sequence.
LOAD/UNLOAD Command Descriptor Block
The following table shows the layout of the CDB.
Byte
0 0 0 0 1 1 0 1 1
1 0 0 0 0 0 0 0 IMMED
2 0 0 0 0 0 0 0 0
3 0 0 0 0 0 0 0 0
4 0 0 0 0 0 0 RT LOAD
5 vendor unique* 0 0 0 0 Flag Link
Bits
7 6 5 4 3 2 1 0
* This vendor-unique field is reserved and must be set to zero.
Command Descriptor Block Field Description
The following table defines the fields in the CDB.
Field
Name Bytes Bits Description
IMMED
bit
RT bit 4 1 Retension bit. No retension is necessary for DDS/DAT cartridges. For
LOAD
bit
1 0 If the IMMED bit is 1, status is returned as soon as the operation is
initiated. If the IMMED bit is 0, status is not returned until the tape is
positioned.
software compatibility this field can be either 0 or 1.
4 0 If the LOAD bit is 1, the tape moves to BOT.
If the LOAD bit is 0, the tape repositions to BOT. The tape is then
ejected. Following an UNLOAD, the drive accepts and executes any
valid non-media access commands.
Page 46 DDS-4 SCSI Interface Manual, Rev. B
Completion Status
The Completion Status for the LOAD/UNLOAD command is shown in the following
table.
Code Message Description
00h Good Status
Load = 1 (Load)
Load = 0
(Unload)
IMMED = 1
02h Check Condition Extended Sense Byte 02h:
• The tape is positioned to the BOT of the current partition.
• The drive remains in any previously set modes.
• The drive is ready to perform any appropriate command.
• The tape is position to BOT and ejected.
• The drive remains in any previously set modes.
• The drive is ready to perform any appropriate command.
Good Status is returned once the command is transferred and verified.
The requested action may still be in progress when another command is
issued. Prior to returning status all buffered data is transferred to tape.
Code Message Description
02h Not Ready No cartridge is inserted in the drive.
04h Hardware Error Parity error on SCSI bus or drive
hardware failure detected.
05h Illegal Request
06h Unit Attention The cartridge was changed prior to BOT
or at BOT following a LOAD command.
• Both IMMED and Link bits are set to 1.
• The CDB contains an invalid bit.
Commands: Locate (2Bh) Page 47
LOCATE (2Bh)
The LOCATE command is used to position the tape to the specified block address in
the specified partition. Before the locate operation is performed, all buffered data,
filemarks, and setmarks are transferred to tape. On completion, the logical position
is before the specified location.
LOCATE Command Descriptor Block
The following table shows the layout of the CDB.
Byte
0 0 0 1 0 1 0 1 1
1 0 0 0 0 0 BT CP IMMED
2 0 0 0 0 0 0 0 0
3 MSB—Block Address
4 Block Address
5 Block Address
6 Block Address—LSB
7 0 0 0 0 0 0 0 0
8 Partition
9 vendor unique* 0 0 0 0 Flag Link
Bits
7 6 5 4 3 2 1 0
* This vendor-unique field is reserved and must be set to zero.
Command Descriptor Block Field Description
The following table defines the fields in the CDB.
Field
Name Bytes Bits Description
CP bit 1 1 If the Change Partition bit is 1, the drive changes to the partition whose
partition number is specified in the Partition field. If this bit is 0, the
Partition field is ignored.
IMMED
bit
BT bit 1 2 If the Block Address Type bit is 1, the Block Address field is the total
Block
Address
Partition 8 This field is valid only when the CP bit is 1. In that case, the field
1 0 If the IMMED bit is 1, status is returned as soon as the operation is
initiated. If the IMMED bit is 0, status is not returned until the tape is
positioned.
number of logical blocks from the beginning of partition. All filemarks
and setmarks are ignored when searching the tape. If this bit is 0, the
Block Address field is the total number of logical blocks, filemarks, and
setmarks from the beginning of partition. Filemarks and setmarks are
counted when searching the tape.
3-6 These fields specify the logical block address of the block to which the
tape is to be positioned. Block 0 is the first block in a partition.
specifies the partition to select—partition 0 (00h) or partition 1 (01h).
This selection should agree with the current tape format, either singlepartition or two-partition. Partition 1 is the first physical partition on a
dual-partition tape.
Page 48 DDS-4 SCSI Interface Manual, Rev. B
LOCATE and the BT Bit
Assume that the data was written on the tape as shown in the following table.
Block 0 1 2 3 4 5 6 7 8 9 10 11 12 13
BOT D S S S S S D D D F D D D EOD
Position A B C D
If the Block Address Type (BT) bit is set in the CDB when the LOCATE command is
issued, all the setmarks and/or filemarks are logically grouped with its nearest Data
Block on its BOT side to form one logical block. The following table shows the
logical blocks when the BT bit is set. In the table, the data block (block number 0)
and five setmarks (block numbers 1-5) are grouped together to form one logical
block. the filemark (in block number 9) would be grouped with the data block (block
number 8) to form another logical block.
Block 0 0 0 0 0 0 1 2 3 3 4 5 6 7
BOT D S S S S S D D D F D D D EOD
If the BT bit is 0 in the CDB when the LOCATE command is issued, then each
filemark and setmark on the tape is counted like the logical block while searching
the target. See the table that follows.
Block 0 1 2 3 4 5 6 7 8 9 10 11 12 13
BOT D S S S S S D D D F D D D EOD
Referring to the first table, the table below shows the tape position after each
LOCATE command is complete.
LOCATE Block Address From BT bit Tape Position
1 BOT ON C
1 EOD ON C
1 BOT OFF A
1 EPD OFF A
4 BOT ON D
4 EOD ON D
4 BOT OFF B
4 EOD OFF B
Commands: Locate (2Bh) Page 49
Completion Status
The Completion Status for the LOCATE command is shown in the following table.
Code Message Description
00h Good
Status
02h Check
Condition
• The tape is positioned to the logical block address specified.
• The drive remains in any previously set mode.
• The drive is ready to perform any appropriate command.
Extended Sense Byte 02h:
Code Message Description
02h Not Ready No cartridge is inserted in the drive.
04h Hardware Error Parity error on SCSI bus or drive
hardware failure detected.
05h Illegal Request
installed.
06h Unit Attention
accepting this command.
08h Blank Check The specified logical block address was not
found on tape. Tape is positioned at EOD.
That is, it is positioned after the last filemark
or set mark of the current partition. The next
WRITE command appends data to the tape.
• An illegal partition number is requested.
• The CDB contains an invalid bit.
• CP bit is set when a single-partition tape is
• The cartridge was changed prior to
• The drive was reset prior to this command.
Page 50 DDS-4 SCSI Interface Manual, Rev. B
LOG SELECT (4Ch)
The LOG SELECT and LOG SENSE commands are used in conjunction with each
other. These commands allow the host to obtain and control statistical information
about cartridge usage and error rates. This information consists of counts related to
particular events that occur. Count thresholds can be established which
automatically generate log exception conditions.
The LOG SELECT command is used to preset counts of pages 2 and 3 to a specific
value or to reset the counts if the maximum count is reached. The log sense data is
also reset by a SCSI Bus Reset or a Bus Device Reset with the exception of the
page 30 and 31 data. That data cannot be reset by the initiator.
Counts of pages 2 and 3 can be read by the LOG SENSE command before a reset
and then restored after the reset by issuing the LOG SELECT command.
The log select data is transferred to the drive by specifying the log select page or
pages to initialize. Multiple pages can be affected by a single log select CDB.
However, the LOG SENSE command can only request one page at a time.
The log select parameter data is transferred to the drive to initialize page data. The
parameter data is organized using structures called log parameters. A log parameter
is made up of a parameter header that contains a parameter code that identifies the
log parameter to be initialized.
The drive determines the pages used by reading log sense page 0. Any request to
initialize an unused log page causes a Check Condition. The request sense data is
set to Illegal Request, Invalid Field in the CDB.
The log parameter codes within each log page used by the drive are determined by
reading each particular log sense page. Any attempt to specify an unused
parameter causes a Check Condition. The sense data is set to Illegal Request,
Invalid Field in the parameter list.
LOG SELECT Command Descriptor Block
The following table shows the layout of the CDB.
Byte
0 0 1 0 0 1 1 0 0
1 0 0 0 0 0 0 PCR 0
2 PC 0 0 0 0 0 0
3 0 0 0 0 0 0 0 0
4 0 0 0 0 0 0 0 0
5 0 0 0 0 0 0 0 0
6 0 0 0 0 0 0 0 0
7 MSB—Parameter List Length
8 Parameter List Length—LSB
9 vendor unique* 0 0 0 0 Flag Link
Bits
7 6 5 4 3 2 1 0
* This vendor-unique field is reserved and must be set to zero.
Commands: Log Select (4Ch) Page 51
Command Descriptor Block Field Description
The following table defines the fields in the CDB.
Field
Name
PCR bit 1 1 The Parameter Code Reset (PCR) bit is used to reset log parameters.
PC bits 2 6-7 The Page Control bits select one of four possible types of log
Parameter
List Length
Bytes Bits Description
If the bit is set (1) and the parameter list length is set to 0, all
accumulated values are set to 0. All threshold values are set to the
defaults.
Pages 0, 30, and 31 are not reset with this bit. When the PCR bit is set,
the PC bits in the CDB are ignored.
A PCR bit of 0 does not cause the parameters to be reset.
information. The settings of these bits are shown in a subsequent table.
7-8 The Parameter List Length field specifies the length (in bytes) of the
parameter list to be transferred during the Data Out phase. A 0 length
indicates that no parameter data is to be transferred to the target.
Errors Detected in the Command Descriptor Block
The following conditions constitute errors that may be detected by the drive in
relation to the CDB. These errors cause a Check Condition. The request sense data
is set to Illegal Request, Invalid Field in the CDB.
• If the PCR bit is set and the parameter list length is not 0.
• A parameter list length that would cause a parameter within a valid page to be
truncated or otherwise incompletely initialized.
• The settings of the PC bits select the default threshold or default accumulated
values and the parameter list length is nonzero.
Use of the PC Bits
The Page Control bits (byte 2, bits 6-7) define the type of parameter values that are
to be sent to the drive. The following table lists the Page Control bit values for the
LOG SELECT CDB.
Bit 7 Bit 6 Type Requested Counts
0 0 0 Threshold values
0 1 1 Accumulated values
1 0 2 Default threshold values
1 1 3 Default accumulated values
USE OF DEFAULT VALUES TO SET CURRENT VALUES
The threshold values can be loaded with the default threshold values by setting the
PC bits to specify the default threshold values. The Parameter List Length must be
set to 0.
Likewise, the accumulated values can be loaded with the default accumulated
values by setting the PC bits to specify the default accumulated values. The
Parameter List Length must be set to 0.
Page 52 DDS-4 SCSI Interface Manual, Rev. B
The drive presets the values as follows:
• The default accumulated values are set to all 0s.
• The default threshold values are set to all 1s.
The default values cannot be changed. If the host issues a LOG SELECT command
with the PC bit specifying default values when the parameter list length is nonzero, a
Check Condition is returned. The sense data is set to Illegal Request, Invalid Field in
the CDB.
The Parameter Control bytes of each parameter are preserved when the LOG
SELECT command is issued with the PC bits specifying either of the default values.
SETTING ACCUMULATED VALUES DIRECTLY
The host can set the accumulated values of pages 2 and 3 to any value by
specifying the accumulated values in the PC bits and transferring the log parameters
containing the new values and the Parameter Control bytes.
This ability to manually set the accumulated values allows the original counts
previously read with the LOG SENSE command to be restored. Thus, the drive can
continue to accumulate counts after a Bus Reset, Bus Device Reset, or a manual
power cycle.
SETTING THE THRESHOLD VALUES DIRECTLY
The host can set the threshold values of pages 2 and 3 to any value by specifying
the threshold values in the PC bits and transferring the log parameters containing
the new values and the Parameter Control bytes.
The threshold value is used in conjunction with the Log Parameter Control byte of
the log data and the RLEC bit of the Mode Select/Sense Control Mode page to
determine if, and when, a log exception Check Condition should be generated when
an accumulated count is incremented.
MAXIMUM COUNTS
Threshold values are not the same as the maximum count. The maximum size of
each counter is according to the size of the counter in bytes. When the maximum
count is reached, the DU bit in the Parameter Control byte for the counter is set to 1
for the Log Sense data indicating that the parameter is no longer to be updated
because the maximum count has been reached.
When a counter reaches the maximum, the counter is no longer incremented. If a
maximum value for the parameter size is reached, the count does NOT roll over and
continue counting. When any counter of a page reaches maximum, all counters of
that page are no longer incremented. A LOG SELECT command can be issued with
the PC bits specifying default accumulated values to reset the accumulated values
and to allow counters to continue as normal. The DU bit of the affected parameter is
also reset to zero.
LOG SELECT Parameter Data
The host should issue a LOG SENSE command to initialize host software that
allows
Commands: Log Select (4Ch) Page 53
• Correct determination of the pages the drive uses.
• Determination of the parameter codes and length of each parameter.
The drive does not maintain a separate set of log parameters for each initiator.
Therefore, a LOG SELECT command affects all initiators.
Note: The subsequent section discusses the LOG SENSE command. The
log page codes and the log parameter codes are described in that section.
Parameter data is made up of one or more page descriptors. A page descriptor is
made up of a four-byte page header and one or more log parameters. A log
parameter is made up of a four-byte parameter header and one or more associated
parameter data bytes.
The parameter header contains a two-byte parameter code to identify the
parameter, a Parameter Control byte, and a parameter length byte.
When a counter reaches its maximum value, a log exception Check Condition is
generated if the RLEC bit is set in the Mode Select Control Mode page. The sense
data is set to Recovered Error key (01), Log Counter at Maximum qualifier, 5B/02.
Bytes 15 - 17 of the Request Sense data specify the page, MSB and LSB of the
Parameter Code respectively, which caused the Log Exception. The counters are
defined in a following section.
The drive uses pages 0, 2, 3, 30, and 31. However, the LOG SELECT command
can only select pages 2 and 3. The following table lists the LOG SELECT page
codes.
Page Page Information
2 Write error counts 0 0 0 0 1 0
3 Read error counts 0 0 0 0 1 1
Bits
5 4 3 2 1 0
Pages 30 and 31 are vendor unique and refer to the DDS Tape Log and Tape
Capacity, respectively. Page 0, the list of supported pages, is available only to the
LOG SENSE command.
Although the counts in page 30 are similar to those in pages 2 and 3, the page 2
and 3 counts are incremented separately from those in page 30. The accumulated
counts in pages 2 and 3 accumulate across cassettes after the counts are initialized
by the LOG SELECT command or since the last reset or power cycle. Thus, the
counts in those two pages can be used to accumulate events in a different way than
the counts in the log sense page 30.
If multiple log pages are sent, they must be sent in ascending order. All log
parameters of a particular page must be sent to the drive in ascending order. Not all
parameter codes need to be sent with the parameter data during the Data Out
phase. In that way, selected parameters of a particular page can be initialized to any
value desired.
The following conditions constitute errors that cause the drive to return a Check
Condition and to set sense data to Illegal Request, Invalid Parameter Data.
Page 54 DDS-4 SCSI Interface Manual, Rev. B
• If any page headers are received with unsupported page codes. (Table 4-30
lists the pages that can be set by the LOG SELECT command.)
• An incorrect log page length is specified in the page header.
• An illegal parameter code within a valid log page.
• Valid log pages are not sent in ascending order.
• Parameter codes of a supported page are not sent in ascending order.
The following table shows the log page header format.
Byte
0 0 0 Page Code
1 0 0 0 0 0 0 0 0
2 0 0 0 0 0 0 0 0
3 Page Length (n)
Bits
7 6 5 4 3 2 1 0
The Page Code field indicates which page is being sent to the drive.
The Page Length field indicates the length of log parameters (in bytes) that follow
the page header. If the LOG SELECT CDB parameter list length specified is too
small to transfer complete log parameters for the specified page, an Illegal Request,
Invalid Field in the CDB, is returned.
The following table shows a typical log parameter format.
Byte
0 MSB—Parameter Code
1 Parameter Code—LSB
2 DU 1 0 ETC TMC 0 0
3 Parameter Length (n-3)
4 Parameter Value
n Parameter Value
Bits
7 6 5 4 3 2 1 0
The Parameter Code field identifies which log parameter is being sent to the drive.
The parameter codes used for the page 2 (Write Error) and page 3 (Read Error) are
listed in subsequent tables.
All of the bits in byte 2 of the Log Parameter are collectively referred to as the
Parameter Control byte. The Parameter Control byte specifies counter controls.
The host specifies the Parameter Control byte to control
• Whether or not a counter is enabled.
• Whether or not a Check Condition is generated when a counter is incremented,
and if so, how the Check Condition is generated.
Each Log Parameter contains only one control byte. This control byte is shared
between the threshold and accumulated parameters. If a LOG SELECT command is
Commands: Log Select (4Ch) Page 55
issued to specify the control bytes for the threshold parameter, the accumulated
Parameter Control byte is affected. The LOG SENSE data reflects the current
setting of the control byte for the parameter.
The Parameter Control byte bits used by the drive are described in the following
table.
Bit Name Bytes Bits Description
Disable
Update (DU)
Enable
Threshold
Comparison
(ETC)
Threshold Met
Criteria Field
(TMC)
2 7 When set (1), this bit indicates that the host does not want this
counter to be updated.
2 4 When set (1), this bit enables the generation of a Check
Condition when an accumulated counter is incremented and
the new value meets the threshold criteria (and MODE
SELECT/SENSE Control Mode Page RLEC bit is set). See the
following heading, Use of the TMC Field and the ETC Bit.
2 3-2 This field indicates when a Check condition should be
generated when the ETC bit is set and a counter is updated
(and MODE SELECT/SENSE Control Mode Page RLEC bit is
set). See the following heading, Use of the TMC Field and the
ETC Bit.
The Parameter Length byte is the byte count of the parameter value that is being
sent. This length must be equal to the length returned by the LOG SENSE
command.
The Parameter Value bytes indicate the value of the selected parameter type
indicated in the CDB Parameter Code field. The initiator must select either the
accumulated or threshold values by the PC bits of the LOG SELECT CDB to send
Parameter Value bytes.
USE OF THE TMC FIELD AND THE ETC BIT
The Threshold Met Criteria (TMC) field specifies how and when a log exception is to
be reported to the host. The Enable Threshold Comparison (ETC) bit enables or
disables the generation of a Check Condition for log exceptions of individual log
parameters. The Report Log Exception Condition (RLEC) bit of the MODE
SELECT/SENSE Control Mode Page, when set, allows the generation of a Check
Condition for log exceptions for each log parameter whose ETC bit is set.
An RLEC bit of zero globally disables all log exception Check Conditions regardless
of the ETC bit value of each log parameter.
When the ETC bit is set (1) (and the RLEC bit is set in the MODE SELECT/SENSE
Control Mode Page), a Check Condition is generated each time the accumulated
value is incremented and the threshold condition is met as described in the following
paragraphs.
When a counter is incremented by the drive, a check is made to determine if the
threshold comparison is enabled (according to the ETC bit) for the counter being
incremented. If comparisons are enabled, the new accumulated value is compared
with the threshold value according to the Threshold Met Criteria for the counter.
When the criteria are met (and the RLEC bit is set in the MODE SELECT/SENSE
Control Mode Page), a log exception Check Condition is returned for the next
command. The sense key is set to Unit Attention (06); the additional sense code is
Page 56 DDS-4 SCSI Interface Manual, Rev. B
set to Log Exception; and the additional sense code qualifier is set to Threshold
Condition Met, 5B/01.
If the counter is disabled because the DU bit is set (1), no threshold conditions are
compared because the counter will not be incremented.
After the log exception is signaled by the Check Condition, the host may issue the
LOG SENSE command to determine which page and which counter (by parameter
code) met the criteria or may read the sense key specific field in the REQUEST
SENSE data.
The criteria for generating a Check Condition are specified by the TMC field in bits 2
and 3 of byte 2 of the parameter header. The possible settings and resulting
operation are as follows.
TMC Field Bits
3 2 Generate a Log Exception Check Condition
0 0 Whenever the accumulated value is incremented.
0 1 When the accumulated value equals the threshold.
1 0 Whenever the accumulated value does not equal the
threshold.
1 1 Whenever the accumulated value is greater than the
threshold value. This setting is the default value
Commands: Log Sense (4Dh) Page 57
LOG SENSE (4Dh)
The LOG SELECT and LOG SENSE commands are used in conjunction with each
other. These commands allow the host to obtain and control statistical information.
This information consists of counts related to particular events that occur.
Some tabulated events (page 3) refer to errors that have occurred since the Minicartridge drive was powered on or last cleared by a LOG SELECT command and
that have accumulated across several backup or restore operations on several
cartridges.
The vendor unique page 30 contains statistical information that is DDS specific. This
DDS information is contained in the Tape Log area on each tape or partition. This
protected reference data is tabulated and maintained for the particular tape currently
being used.
For pages 2 and 3, the LOG SELECT command is used to preset counts to a
specific value or to reset the counts if the maximum count is reached. The log sense
data is reset on a SCSI Bus Reset or a Bus Device Reset.
Log sense data is obtained by specifying a log page in the Page code field of the
LOG SENSE CDB. Only one page of data can be requested by the host of each
LOG SENSE command. The pages used by the drive are found by reading Log
Page 0. Any request for unused pages causes a Check Condition. The sense data
are then set to Illegal Request, Invalid Field in the CDB.
LOG SENSE Command Descriptor Block
The following table shows the layout of the CDB.
Byte
0 0 1 0 0 1 1 0 1
1 0 0 0 0 0 0 0 0
2 PC | Page Code
3 0 0 0 0 0 0 0 0
4 0 0 0 0 0 0 0 0
5 MSB—Parameter Pointer
6 Parameter Pointer—LSB
7 MSB—Allocation Length
8 Allocation Length—LSB
9 vendor unique* 0 0 0 0 Flag Link
* This vendor-unique field is reserved and must be set to zero.
Bits
7 6 5 4 3 2 1 0
Page 58 DDS-4 SCSI Interface Manual, Rev. B
Command Descriptor Block Field Description
The following table defines the fields in the CDB.
Field
Name
PC bits 2 6-7 These Page Control bits select one of four possible types of log
Page
Code
Parameter
Pointer
Allocation
Length
Bytes Bits Description
information. These bit settings are shown in Using Page Control Bits.
2 0-5 This field indicates the requested page code.
5-6 7-0 This two-byte field specifies the beginning parameter code to be
returned to the initiator. All remaining parameter codes are returned
in ascending order.
This field is invalid of page code 0. If the Parameter Pointer field is
nonzero when page 0 is requested, a Check Condition is returned.
The sense data are set to Illegal Request, Invalid Field in CDB.
7-8 7-0 The Allocation Length field specifies the maximum amount of
memory space (in bytes) that the initiator has reserved for log sense
data.
The drive returns the amount of bytes contained in the requested
page or the requested Allocation Length, whichever is less. If the
Allocation Length is greater than the actual page length, the transfer
is truncated.
The data for only one log sense page can be transferred with any one CDB. Log
sense data is organized by pages. The parameter data for each page is preceded
by a four-byte page header. The page header specifies the page code returned and
the length of that page in bytes.
If the Parameter Pointer field is used to specify a starting parameter code, a page
header is still returned.
Following the page header are 0 or more log parameters. Each log parameter is a
special data structure that contains several description bytes and the parameter
value itself. (Subsequent subsections describe the log parameter fields.)
The following table details the log sense page code field.
Bits
5 4 3 2 1 0 Page Page Information
0 0 0 0 0 0 0 List of available pages
0 0 0 0 1 0 2 Write error counts
0 0 0 0 1 1 3 Read error counts
1 0 1 1 1 0 2E Tape Alert Flags
1 1 0 0 0 0 30 DDS tape log (vendor unique)
1 1 0 0 0 1 31 Tape capacity (vendor unique)
1 1 1 0 0 1 39 Data compression transfer log
Commands: Log Sense (4Dh) Page 59
Using Page Control Bits
The Page Control (PC) bits specify the type of counts that the initiator is requesting.
For page 0, the PC bits are ignored. For other pages, the PC bits specify the type of
parameter values. The following table lists the Page Control bit values.
Bits
7 6 Type
0 0 0 Threshold values
0 1 1 Accumulated values
1 0 2 Default threshold values
1 1 3 Default accumulated values
Threshold Values
The type 0 indicates the current threshold values. If threshold values are requested,
the drive returns the maximum count that each parameter code can attain (the
default) or the last threshold values set by the last LOG SELECT command.
Threshold values for pages 2 and 3 can by changed by the initiator through the LOG
SELECT command. The drive does not change these values during normal
operation.
Requested Counts
Accumulated Values
Type 1 (accumulated vales) is the most common page control bit setting. This type
requests the current counts for the page. These counts reflect the current count of
events since the last power-on cycle, SCSI bus Reset, Bus Device Reset, or the last
LOG SELECT command that cleared or set parameter values.
Note: For pages 30 and 31, the parameter values cannot be changed and
are never reset. These values represent historical data regarding tape
usage and cannot be changed by the LOG SELECT command. Page 30
data is written to the tape log when the tape is ejected either manually or
through the SCSI UNLOAD command.
Accumulated values are incremented by the drive as an event occurs. These page 2
and 3 values can be updated or initialized by the initiator with the LOG SELECT
command. Also, the default accumulated values (all zeros) can be loaded with the
accumulated values through the LOG SELECT command.
Because the drive maintains the accumulated values in volatile memory, the values
may be lost if a power cycle occurs. The values for page 30 are saved on the
cassette itself and are preserved. However, if a power cycle occurs before the
cassette is ejected, the current page 30 data is lost.
The tape log values cannot be updated if the cassette is write protected.
Default Threshold Values
The type 2 page control bit setting requests default threshold values. This request
generally returns the maximum count that each parameter code can attain. For
example, a two-byte field returns two bytes of all 1s; a three-byte field returns three
bytes of all 1s. These default values cannot be changed by the LOG SELECT
Page 60 DDS-4 SCSI Interface Manual, Rev. B
command. The default threshold values are loaded into the threshold values when a
reset occurs or by a LOG SELECT command with the PCR bit set.
Default Accumulated Values
The type 3 page control bit setting specifies default accumulated values. The default
accumulated values are all zero and cannot be changed by the LOG SELECT
command. The default accumulated values are loaded into the accumulated values
when a reset occurs or when a LOG SELECT command is received with the PCR bit
set.
Using the Parameter Pointer Field
The Parameter Pointer field specifies the starting parameter code of the page that is
to be transferred.
If the Parameter Pointer field specifies a parameter code larger than the maximum
parameter code for that page, a Check Condition is generated. The sense data is
set to Illegal Request, Invalid Field in the CDB.
For example, if the page uses parameter codes 2 through 6 and the parameter
pointer field is set to 3, then the drive returns parameters 3 through 6. Likewise, if
the parameter pointer is set to 1, parameters 2 through 6 are returned.
If the Parameter Pointer field is set to 0, all parameters for that particular page are
returned up to the maximum number of bytes specified in the allocation length.
Regardless of the starting parameter code specified in the Parameter Pointer field,
the page header is always returned.
Log Sense Pages
Only one log page is transferred to the host with a single LOG SENSE CDB. Byte 2
of the CDB specifies the page to be transferred. Each log page begins with a fourbyte header, as shown in the following table.
Byte
0 0 0 | Page Code
1 0 0 0 0 0 0 0 0
2 0 0 0 0 0 0 0 0
3 Page Length (n)
The Page Code field indicates the page being returned. This data matches the page
code requested in byte 2 of the LOG SENSE CDB.
The Page Length field indicates the length of the page in bytes that follow the page
header. If the allocation length specified in the CDB is too small to transfer all of the
requested page, this value is not adjusted to reflect the truncation.
Bits
7 6 5 4 3 2 1 0
However, if the Parameter Pointer field of the CDB specifies a starting parameter
code other than zero, the page length is adjusted to indicate the number of bytes
that follow the page header.
Commands: Log Sense (4Dh) Page 61
The following subsections describe the log sense pages.
List of Available Pages (Page 0)
Page 0 indicates the log sense pages used by the drive. To determine the size of
each page and of each parameter in the page, the individual pages must be
requested.
Page 0 is unique in that the parameter data returned does not contain log parameter
headers. All other pages return a page header followed by zero or more variablelength log parameters.
Page 0 is valid only for the LOG SENSE command. When page 0 is requested, the
four-byte page header is returned followed by the page codes used—one byte for
each. the available page codes are returned in ascending order. The following table
shows the page 0 data format.
Byte HEX Code Description
0 00 Header, page 0
1 00 Header, reserved
2 00 Header, page length MSB
3 09 Header, page length LSB
4 00 Page 0
5 02 Page 2
6 03 Page 3
7 2E Page 2E
8 30 Page 30
9 31 Page 31
10 39 Page 39
11 3A Page 3A
12 3C Page 3C
Page Code Data Format for Pages
The parameter data returned consists of a four-byte page header followed by zero
or more log parameters as shown in the following illustration.
Page Header
First Log Parameter
.
.
.
Last Log Parameter
A log parameter is made up of
• A two-byte parameter code followed by
• A Parameter Control byte
• A Parameter Length byte
• One or more parameter bytes
Parameter Code Field
The following table presents the Parameter Code Field format.
Page 62 DDS-4 SCSI Interface Manual, Rev. B
Byte
0 MSB—Parameter Code
1 Parameter Code—LSB
2 Parameter Control
3 Parameter Length (n-3)
4 Parameter Value
N Parameter Value
Bits
7 6 5 4 3 2 1 0
The Parameter Code field identifies the log parameter being returned to the initiator.
If the Parameter Pointer field of the CDB is zero, the parameter code of the first log
parameter indicates the first parameter code supported by the drive for this log
page. Parameter codes are always returned in ascending order.
All of the bits in byte 2 of the log parameter are collectively referred to as the
Parameter Control byte. The bit descriptions as they are returned by the LOG
SENSE command are given in the following table.
Bit Name Bytes Bits Description
Disable Update (DU) 2 7 When set (1), this bit indicates that the particular
parameter is not to be updated by the drive. This bit is
set by the drive when the accumulated value reaches its
maximum. It is also returned set if the host set the bit for
the last LOG SELECT command. The default is zero.
For parameter types other than threshold and
accumulated values, this bit is always 0.
Disable Save (DS) 2 6 A DS bit of zero indicates that the target supports saving
for the specified log parameter. A DS bit of one
indicates that the target does not support saving for the
specified log parameter in response to Log Select
command.
Target Save Disable
(TSD)
Enable Threshold
Comparison (ETC)
Threshold Met Criteria
Field (TMC)
List Binary Information 2 1 Always 0
List Parameter (LP) 2 0 Always 0
2 5 Always 0
2 4 This bit is returned as set by the last LOG SELECT
command. The default is zero.
2 3-2 This field is returned as set by the last LOG SELECT
command. As the default, both bits of this field are
returned set.
The Parameter Length byte is the byte count of the parameter value that follows.
This length is not adjusted to match any truncation.
The Parameter Value bytes are the actual data requested according to the PC bits
of the CDB.
Write Error Counts (page 2)
Page 2 contains the Write Error counts, which are tabulations of the possible errors
(if any) encountered during writing to tape. The page 2 parameter codes for write
errors are listed in the following table.
Parameter Code Length Name
0002 2 Total rewrites
0003 3 Total errors corrected
Commands: Log Sense (4Dh) Page 63
0004 2 Not applicable (always returns 0)
0005 8 Total bytes processed (written)
0006 2 Total uncorrectable errors
8007 2 Rewrites since last read-type operation
The total rewrite count is incremented each time a frame sequence is rewritten on
tape. Rewrites are used to recover from media errors detected by the RAW (readafter-write) checking feature of the drive.
The total errors corrected count is the same as the total rewrite count.
The total bytes written count is incremented by the size of each group (in bytes) that
is successfully written to tape. This count includes any fill bytes needed to fill a
group before the group is written to tape.
However, this count does NOT include the count of additional bytes written during
rewrites. Also, it is greater than the count of write data bytes transferred to the drive.
The total bytes written count includes a total of
• All user data
• ECC frames
• Other structures in the group data (as defined by the DDS format)
The total uncorrectable errors count is a tabulation of the times the drive could not
correct a write error by any means (including the rewrite retry limit being exceeded).
The rewrites since last read-type operation count is similar to the total rewrite count
(0002) except the counter is set to 0 in two cases:
• When the operation is changed to a read-type operation.
• When a rewind operation is performed.
Read Error Counts (page 3)
Page 3 contains the Read Error counts, which are tabulations of any errors
encountered while reading the tape. these tabulations include possible errors from
media access commands.
The page 3 parameter codes for read errors are listed in the following table.
Parameter Code Length Name
0002 2 Total rereads
0003 3 Total errors corrected
0004 2 Total correctable ECC C3 errors
0005 8 Total bytes processed (read)
0006 2 Total uncorrectable errors
8007 2 Rereads since last write-type operation
The total reread count is incremented each time the tape is repositioned to reread a
frame from the tape.
Page 64 DDS-4 SCSI Interface Manual, Rev. B
The total errors corrected count is incremented each time the drive recovers from a
read error. This count includes the reread count and all C3 ECC errors that could be
corrected.
Note: This count does NOT include any frames read that were rewritten
when the tape was recorded.
The total correctable ECC error count is incremented each time the ECC process
corrects one to two erroneous tracks read from the tape. For performance reasons,
the ECC process corrects errors rather than retrying if the tracks in error are less
than three; otherwise, a reread is performed.
Note: This count is a count of the correctable C3 errors ONLY. It does NOT
include the C1 or C2 ECC errors normally encountered while reading DDSformatted cassettes.
The total bytes read count is incremented by the size of each group (in bytes) that is
successfully read from the tape. This count includes overhead bytes (as defined by
the DDS format specification) that were read from the group.
However, this count does NOT include any additional rewritten frames that were
read nor any rereads. Also, it is greater than the count of read data bytes
transferred.
The total uncorrectable errors count is a tabulation of the times the drive could not
correct a read error by any means (including the reread retry limit being exceeded).
The rereads since last write-type operation count is similar to the total reread count
(0002) except the counter is set to 0 in two cases:
• When the operation is changed to a write-type operation.
• When a rewind operation is performed.
Tape Alert Log Page (page 2Eh)
The Tape Alert Log page provides information about errors and status of the drive
and media in the form of a flag for the supported Tape Alert Flags. The DDS-4 and
DAT 72 drive autoloader conform to the Tape Alert v3.0 Standard. The Tape Alert
Log page can only be accessed by the Log Sense commands, there are no Log
Select actions defined in the Tape Alert specification. The log page is used in
conjunction with the Tape Alert Mode Page that controls the action taken by the
drive when a flag is set and sets various other control mechanisms of the Tape Alert
system. A flag is set whenever the condition for setting the flag exists. There are
three types of flags, Informational, Warning, and Critical. The type of flag depends
upon the severity of the condition that sets the flag. Flags are cleared on the
following conditions:
• At drive power on
• When the TapeAlert Log page is read
• When specified corrective action has been taken (such as using a cleaning
cartridge)
• On a SCSI bus reset or bus device reset message
• On Log Select reset (note that the recommended action on receiving Log Select
for the TapeAlert Log page is to reject the command with an error
Commands: Log Sense (4Dh) Page 65
Additional information regarding the Tape Alert system is found in the SCSI-3
Stream Commands (SSC) Standard.
Page 66 DDS-4 SCSI Interface Manual, Rev. B
The format of the Tape Alert Log Page is shown below:
Byte
0
1 0 0 0 0 0 0 0 0
2 MSB Page Length (140h)
3 LSB
5n-1 MSB – Parameter Code
5n LSB – Parameter Code
5n+1 DU(0) DS(1) TSD(0) ETC(0) TMC LBIN(0) LP(0)
5n+2 Parameter Length (1)
5n+3 Value of Flag (set when bit 0 = 1)
Bits
7 6 5 4 3 2 1 0
There are 64 Tape Alert flags for drives and autoloaders, (some are reserved). The
following table lists the flags supported by the drive:
No Flag Type Required Host Message Cause
1 Read Warning W The tape drive is having problems reading data. No data has been
lost, but there has been a reduction in the performance of the tape.
2 Write
Warning
3 Hard Error W The operation has stopped because an error has occurred while
4 Media C Your data is at risk:
5 Read Failure C The tape is damaged or the drive is faulty. Call the tape drive
6 Write Failure C The tape is from a faulty batch or the tape drive is faulty:
7 Media Life W The tape cartridge has reached the end of its calculated useful life:
8 Not Data Grade W The tape cartridge is not data-grade. Any data you back up to the
9 Write Protect C You are trying to write to a write-protected cartridge.
10 No Removal I You cannot eject the cartridge because the tape drive is in use.
11 Cleaning Media I The tape in the drive is a cleaning cartridge.
12 Unsupported
Format
13 Recoverable
Snapped Tape
W The tape drive is having problems writing data. No data has been
lost, but there has been a reduction in the capacity of the tape.
reading or writing data that the drive cannot correct.
1. Copy any data you require from this tape.
2. Do not use this tape again.
3. Restart the operation with a different tape.
supplier helpline.
1. Use a good tape to test the drive.
2. If the problem persists, call the tape drive supplier helpline.
1. Copy any data you need to another tape
2. Discard the old tape.
tape is at risk.
Replace the cartridge with a data-grade tape.
Remove the write-protection or use another tape.
Wait until the operation is complete before ejecting the cartridge.
If you want to back up or restore, insert a data-grade tape.
I You have tried to load a cartridge of a type that is not supported by
this drive.
C The operation has failed because the tape in the drive has snapped:
1. Discard the old tape.
2. Restart the operation with a different tape.
The drive is having
severe trouble reading
The drive is having
severe trouble writing
The drive had a hard read
or write error
Media can no longer be
written/read, or
performance is severely
degraded
The drive can no longer
read data from the tape
The drive can no longer
write data to the tape
The media has exceeded
its specified life
The drive has not been
able to read the MRS
stripes
Write command is
attempted to a write
protected tape
Manual or s/w unload
attempted when prevent
media removal on
Cleaning tape
encountered during
backup or restore
Attempted loaded of
unsupported tape format,
e.g. DDS1 or DDS2 tapes
Tape snapped/cut in the
drive where media can be
ejected
Commands: Log Sense (4Dh) Page 67
No Flag Type Required Host Message Cause
14 Unrecoverable
Snapped Tape
16 Forced Eject C The operation has failed because the tape cartridge was manually
17 Read Only
Format
19 Nearing Media
Life
20 Clean Now C The tape drive needs cleaning:
21 Clean Periodic W The tape drive is due for routine cleaning:
22 Expired
Cleaning Media
30 Hardware A C The tape drive has a hardware fault:
31 Hardware B C The tape drive has a hardware fault:
32 Interface W The tape drive has a problem with the host interface:
33 Eject Media C The operation has failed:
34 Download Fail W The firmware download has failed because you have tried to use the
C The operation has failed because the tape in the drive has snapped:
1. Do not attempt to extract the tape cartridge.
2. Call the tape drive supplier helpline.
ejected while the tape drive was actively writing or reading.
W You have loaded a cartridge of a type that is read-only in this drive.
The cartridge will appear as write-protected
I The tape cartridge is nearing the end of its calculated life. It is
recommended that you:
1. Use another tape cartridge for your next backup.
2. Store this tape cartridge in a safe place in case you need to
restore data from it.
1. If the operation has stopped, eject the tape and clean the drive
2. If the operation has not stopped, wait for it to finish and then
clean the drive.
Check the tape drive users manual for device specific cleaning
instructions.
1. Wait for the current operation to finish.
2. Then use a cleaning cartridge.
Check the tape drive users manual for device specific cleaning
instructions.
C The last cleaning cartridge used in the tape drive has worn out:
1. Discard the worn out cleaning cartridge.
2. Wait for the current operation to finish.
3. Then use a new cleaning cartridge.
1. Eject the tape or magazine.
2. Reset the drive.
3. Restart the operation.
1. Turn the tape drive off and then on again.
2. Restart the operation.
3. If the problem persists, call the tape drive supplier helpline.
Check the tape drive users manual for device specific instructions
on turning the device power on and off.
1. Check the cables and cable connections.
2. Restart the operation.
1. Eject the tape or magazine.
2. Insert the tape or magazine again.
3. Restart the operation.
incorrect firmware for this tape drive.
Obtain the correct firmware and try again.
Tape snapped/cut in the
drive where media cannot
be ejected
Manual or forced eject
while drive actively writing
or reading
Media loaded that is readonly format
Media may have
exceeded its specified
number of passes
The drive thinks it has a
head clog, or needs
cleaning
The drive is ready for a
periodic clean
The cleaning tape has
expired
The drive has a hardware
fault that requires reset to
recover.
The drive has a hardware
fault that is not read/write
related or requires a
power cycle to recover.
The drive has identified
an interfacing fault
Error recovery action
Firmware download failed
Page 68 DDS-4 SCSI Interface Manual, Rev. B
The following table lists the flags supported by the autoloader:
No. Flag Type Required Host Message Cause
40 Loader
Hardware A
41 Loader Stray
Tape
42 Loader
Hardware B
43 Loader Door C The operation has failed because the autoloader door is open:
44 Loader
Hardware C
45 Loader
Magazine
C The changer mechanism is having difficulty communicating with the
tape drive:
1. Turn the autoloader off then on.
2. Restart the operation.
3. If problem persists, call the tape drive supplier helpline.
C A tape has been left in the autoloader by a previous hardware fault:
1. Insert an empty magazine to clear the fault.
2. If the fault does not clear, turn the autoloader off and then on
again.
3. If the problem persists, call the tape drive supplier helpline.
W There is a problem with the autoloader mechanism. Loader mechanism has a
1. Clear any obstructions from the autoloader door.
2. Eject the magazine and then insert it again.
3. If the fault does not clear, turn the autoloader off and then on
again
4. If the problem persists, call the tape drive supplier helpline.
C The autoloader has a hardware fault:
1. Turn the autoloader off and then on again.
2. Restart the operation.
3. If the problem persists, call the tape drive supplier helpline.
Check the autoloader users manual for device specific instructions
on turning the device power on and off.
C The autoloader cannot operate without the magazine.
1. Insert the magazine into the autoloader
2. Restart the operation.
Loader mechanism is
having trouble
communicating with the
tape drive
Stray tape left in loader
after previous error
recovery
hardware fault
Tape changer door open
The loader mechanism has
a hardware fault that is not
mechanically related.
Loader magazine not
present
DDS Tape Log (page 30)
Page 30, the DDS Tape Log page, allows you to obtain the information last written
to the DDS Tape Log. Each time the cartridge is ejected the Tape Log is updated.
This log contains two types of counts:
• The Previous counts indicate events from the previous usage cycle of the
cassette. (A usage cycle is composed of a single cartridge insertion, use, and
ejection.)
• The Total counts accumulate events over the life of the cassette since the
cassette was first initialized by a format process.
When a cartridge is formatted, the total and previous counts are initialized to 0. The
previous counts in this page are updated when a previously formatted cassette is
inserted to indicate the counts from the previous usage of the tape. The previous
counts remain consistent through the usage cycle. Previous counts can be
inspected with the LOG SENSE command.
If no activity occurs during the usage cycle, then the previous counts are all set to 0
and the Load Count is incremented by 1.
Commands: Log Sense (4Dh) Page 69
When a previously used cartridge is inserted, the total counts are updated to the
values contained in the Tape Log of the inserted cassette. The total counts can be
read with the LOG SENSE command.
Also, page 30 contains a set of current counts. The current usage counts are not
part of the Tape Log. They are initialized to 0 when the cartridge is inserted. While
the tape is in use, the current usage counts are incremented according to the DDS
format definitions as an event occurs. These counts cannot be changed by the
initiator.
When the cartridge is ejected or the UNLOAD command is received, the current
usage counts are saved in the previous counts. The total counts are also updated to
indicate the new totals. This updated data is written to the tape in the Tape Log
area. The Tape Log data is not cleared by any reset and cannot be changed with
the LOG SELECT command.
Note: If power is lost before the update data is written to the tape, the Tape
Log area is not updated. Also, if the cassette is write protected, the Tape
Log area is not updated when the cassette is ejected. The drive does NOT
post a Check Condition for this case.
Each page 30 count is a positive binary value. The counts do not overflow. Once a
count reaches its maximum value, that maximum value is always returned. The
counts do not roll over at the maximum value.
A single-partition tape contains one Tape Log; a dual-partition tape contains two
Tape Logs. The appropriate Tape Log is returned as determined by the previously
selected partition.
The page 30 parameter codes are listed in the following table (* indicates that the
most significant 4 bits are 0).
Name Code Length Description
Current
Groups
Written
Current
Rewritten
Frames
Current
Groups Read
Current ECC
C3 Corrections
Previous
Groups
Written
Previous
Rewritten
Frames
Previous
Groups Read
0001 3* This count is the number of groups written to the partition since
the cartridge was inserted. Each group contains 126 KB of
data.
0002 2 This count is the number of rewritten frames within the
partitions since the cartridge was inserted. It is incremented by
one each time a series of frames is rewritten. Frames are
rewritten following an error detected by the RAW check. If the
data written during the rewrite is also found to be bad by the
RAW check, the series of frames is rewritten again, and this
count is incremented by 1 again.
0003 3* This count is the number of groups read from the cartridge
since it was inserted.
0004 2 This count is the number of times the drive used the C3 ECC
correction to recover data from the partition since the cartridge
was inserted.
0005 3* This count is the number of groups written to the partition
during the last cartridge usage cycle.
0006 2 This count is the number of frames rewritten in the partition
during the last cartridge usage cycle. It is greater than or equal
to the last current rewritten frames count prior to the Tape Log
being updated when the cartridge is ejected.
0007 3* This count is the number of groups read from the partition
during the last cartridge usage cycle.
Page 70 DDS-4 SCSI Interface Manual, Rev. B
Name Code Length Description
Previous ECC
C3
Corrections
Total Groups
Written
Total
Rewritten
Frames
Total Groups
Read
Total ECC C3
Corrections
Load Count 000D 2 This count is the number of times the tape has been loaded
*Most significant 4 bits are 0.
0008 2 This count is the number of times the drive used C3 ECC
correction to recover data from the partition during the last
cartridge usage cycle.
0009 4 This count is the number of groups written since the partition
was formatted. This count accumulates over the life of the
cartridge but is zeroed if the tape is formatted again.
000A 3 This count is the total number of times frames were rewritten
within the partition since the partition was formatted. It is
incremented by one each time a series of frames is rewritten
following an error detected by the RAW check. It accumulates
over the life of the cartridge but is zeroed if the tape is
formatted again.
000B 4 This count is the number of groups read from the partition
since it was formatted. It accumulates over the life of the
cartridge but is zeroed if the tape is formatted again.
000C 3 This count is the number of times the drive used C3 ECC
correction to recover data from the partition since the tape
was formatted. It accumulates over the life of the cartridge but
is zeroed if the tape is formatted again.
over the life of the tape since it was formatted. One load is the
same as a usage cycle (inserting, using, and subsequent
ejection of the cartridge). It accumulates over the life of the
cartridge but is zeroed if the tape is formatted again. In a dualpartition tape, only one load count is maintained.
Tape Capacity Page (page 31)
Page 31, the Tape Capacity page, allows you to obtain information about total and
remaining storage capacity for each partition.
Host software can use this page to determine cassette size and remaining capacity
while writing to the tape.
The Tape Capacity Page parameter codes area given in the following table.
Parameter Code Length Name
0001h 4 Remaining capacity, partition 0 (Kbytes)
0002h 4 Remaining capacity, partition 1 (Kbytes)
0003h 4 Maximum capacity, partition 0 (Kbytes)
0004h 4 Maximum capacity, partition 1 (Kbytes)
All capacities are estimates as to the maximum available user-data capacities. The
actual capacity may be slightly less because of rewrites. All values are in kilobytes.
Partition 1 is the first partition on a dual-partition tape. For single-partition tapes, only
partition 0 is used.
The remaining capacity reflects the current head position. Thus, a rewind will reset
the remaining capacity to the maximum capacity. The remaining capacity of the
partition that is not current will be the same as its maximum capacity. The maximum
capacity never changes for a particular formatted cartridge. To obtain capacity
estimates, multiply the values returned by 1,024.
Commands: Log Sense (4Dh) Page 71
Data Compression Transfer Log Page (Page 39)
Page 39, the Data Compression Transfer Log Page, allows you to monitor the
performance of data compression.
The Data Compression Page Parameter Codes are given in the following table:
Name Code Length Description
Number of
entities written
Number
Number of
entities read
Number of
records written
Number of
records read
Kilobytes to data
compression
Kilobytes from
data compression
Kilobytes to tape 0007h 8 The total number of kilobytes written to or read from the
Kilobytes from
tape
Logical entity size 0009h 8 The logical size of the last entity written to or read from
Physical entity
size
Uncompressed
entities
0001h 8 The total number of complete entities written to or read
from the tape since the last power-on or Clear Log
operation
0002h 8
0003h 8 The total number of records (both compressed and
uncompressed) written to or read from the tape since the
last power-on or Clear Log operation.
0004h 8
0005h 8 The total number of kilobytes written to or read from the
tape since the last power-on or Clear Log operation. When
reading this number may be greater than the number
transferred to the host due to read ahead.
0006h 8
tape since the last power-on or Clear Log operation
0008h 8
tape (size = number of records in entity X record size)
000Ah 8 The physical size of the last entity written to or read from
the tape. (size = entity header length X uncompressed
data length)
000Bh 8 The total number of times a non-compressed entity has
been encountered on the tape during a read operation
since the last power-on or Clear Log operation
Drive Configuration Information Page (Page 3A)
Parameter Code Length Name
02 4 Compression enabled set by mode select
03 4 Decompression enabled set by mode select
04 4 Block size used for writing
05 4 Current partition
06 4 Prevent (1) or allow (0) media removal
07 4 Cartridge write protected
08 4 Report setmarks
09 4 Data compression ratio
Drive Usage Information Page (Page 3C)
Parameter Code Length Name
02 8 Total rewrites
03 8 Total groups read
04 8 Total ecc c3 corrections
05 8 Total rereads
06 3 Total load count
Page 72 DDS-4 SCSI Interface Manual, Rev. B
Parameter Code Length Name
08 3 Power on time (minutes)
09 3 Cylinder on time (minutes)
0A 2 Cleaning Cartridge count
0B 2 Reserved
0C 2 Reserved
0D 2 Reserved
0E 2 Reserved
0F 2 Reserved
10 2 Reserved
11 1 Reserved
12 1 Reserved
Completion Status
The Completion Status for the LOG SENSE command is shown in the following
table.
Code Message Description
00h Good Status
02h Check
Condition
Extended Sense Byte 02h:
Code Message Description
02h Not Ready No cartridge is inserted in the drive.
04h Hardware Error Parity error on SCSI bus or drive
hardware failure detected.
05h Illegal Request
specifies an unsupported log page.
06h Unit
Attention this command
command.
• The CDB contains an invalid bit.
• The Page Code in the Log Page Header
• The parameter pointer is invalid.
• Cartridge was changed prior to accepting
• The drive was reset prior to this
Commands: Mode Select (15h) Page 73
MODE SELECT (15h)
The MODE SELECT command allows the host to assign device parameters.
After a power-on or SCSI reset condition, the drive sets its device parameters to the
default values. By issuing a MODE SELECT command, the host can change the
device parameters. The parameters are transferred to the drive as data formatted in
a parameter list.
Parameters assigned by the MODE SELECT command remain in effect until the
drive receives a subsequent MODE SELECT command or a reset. The MODE
SELECT parameters are not unique to the initiator that assigned the parameters.
The MODE SELECT command immediately checks for invalid parameters or invalid
combinations of parameters before executing. If an exception is found, the drive
returns a Check Condition, and the request sense data is set to Illegal Request.
You can issue a MODE SENSE command following a MODE SELECT command to
determine which parameters (if any) were rounded to the nearest supported value.
MODE SELECT Command Descriptor Block
The following table shows the layout of the CDB.
Byte
0 0 0 0 1 0 1 0 1
1 0 0 0 PF 0 0 0 0
2 0 0 0 0 0 0 0 0
3 0 0 0 0 0 0 0 0
4 Parameter List Length
5 vendor unique* 0 0 0 0 Flag Link
* This vendor-unique field is reserved and must be set to zero.
Bits
7 6 5 4 3 2 1 0
Page 74 DDS-4 SCSI Interface Manual, Rev. B
Command Descriptor Block Field Description
The following table defines the fields in the CDB.
Field Name
Bytes Bits Description
PF bit 1 4 The Page Format (PF) bit indicates whether the drive should
interpret the MODE SELECT parameters that follow the CDB and
the header as SCSI-1 or as SCSI-2 parameters. For the Scorpion
40 drives, the PF bit must be set to 1 (SCSI-2).
Parameter List
Length
4 This field specifies the number of bytes in the MODE SELECT
parameter list that are transferred from the host to the drive
during a Data-Out Phase. A Parameter List Length of zero
specifies that no data is to be transferred. This length is valid and
is not considered an error. A length in this field that truncates a
parameter list (as returned in a MODE SENSE command) causes
the drive to return a Check Condition, and the request sense data
is set to Illegal Request.
MODE SELECT Parameters
The parameter list for the MODE SELECT command contains
• A four-byte header, followed by
• Zero or one eight-byte Block Descriptor, followed by
• Zero or more variable-length pages.
The following table illustrates the structure of the mode select parameter list.
Byte
0
.
.
3
4
.
.
11
12
.
.
.
n
Bits
7 6 5 4 3 2 1 0
Parameter List Header
Block Descriptor
Mode Pages
The following table shows the Parameter List Header format.
Byte
0 0 0 0 0 0 0 0 0
1 0 0 0 0 0 0 0 0
2 0 BUFFERED MODE 0 0 0 0
3 Block Descriptor Length
Bits
7 6 5 4 3 2 1 0
Commands: Mode Select (15h) Page 75
Parameter List Header Field Descriptions
The following table defines the fields in the Parameter List Header.
Field Name
Bytes Bits Description
Buffered
Mode
Block
Descriptor
Length
2 6-4 This field sets the buffering mode of the drive
The default mode is 001, buffered, The drive may report good
status on write commands as soon as all the data specified in the
write command has been transferred to the drive’s buffer.
Mode 000 is unbuffered, The drive will not report good status on
write commands until the data is actually written on the tape.
3 This field indicates the number of bytes of block descriptor
information that follow the MODE SELECT header.
The block descriptor length may be set to 00h, which indicates that
no block descriptor bytes are included in the parameter list. This
selection is valid and is not considered an error.
If the block descriptor length is 08h, eight bytes of block descriptor
information are specified in the parameter list.
The mode pages can be sent immediately following the header
depending on the specified parameter list length.
Parameter List—Block Descriptor
The following table presents the Parameter List block descriptor.
Byte
0 Density Code
1 0 0 0 0 0 0 0 0
2 0 0 0 0 0 0 0 0
3 0 0 0 0 0 0 0 0
4 0 0 0 0 0 0 0 0
5 MSB—Block Length
6 Block Length
7 Block Length—LSB
Bits
7 6 5 4 3 2 1 0
Page 76 DDS-4 SCSI Interface Manual, Rev. B
Parameter List—Block Descriptor Field Descriptions
The following table defines the fields in the Parameter List block descriptor.
Field
Name Bytes Description
Density
Code
Block
Length
0 The following values define this field:
Hex Code Meaning
00 Default format (DDS)
7F No-op
25 DDS-3
26 DDS-4
47 DAT 72
The drive will accept any of these density codes as valid when sending a Block
Descriptor. The density code does not dictate the format written on the media, the
drive detects the media type and writes the correct format for that media.
Selecting a Density Code value not listed above, or not supported by the drive,
will result in a Check Condition. The sense data is set to Invalid field in parameter
list.
5-7 If this field is nonzero, it indicates the length of the fixed-length block to be read or
written when the Fixed bit is set in a READ or WRITE command. (It is acceptable
to set the Fixed bit to 0.). The default is 512-byte blocks, which may be changed
at any time by the host with a MODE SELECT command.
If this field is zero, variable-length block mode is specified, and the Fixed bit in the
READ or WRITE command must be 0. The Block Length in a READ or WRITE
command may be set in a range from 1 to 16 MB (as specified in the READ
BLOCK LIMITS command.)
Mode Page Format
The following table shows the generic Mode Page format.
Byte
0 0 0 | Page Code
1 Page Length
2
.
.
.
n
Bits
7 6 5 4 3 2 1 0
Mode Parameters
Commands: Mode Select (15h) Page 77
The following table explains the page layout fields
Field Name
Bytes Description
Page Code 0 The Page Code field identifies the format and parameters for this page as
follows:
Hex Code Page Name
02 Disconnect/reconnect
0A Control mode
0F Data compression control
10 Device configuration
11 Medium partition
1C Tape Alert Mode
1D Element address assignment
(Autoloader only)
1F Device capabilities
(Autoloader only)
Page Length 1 The Page Length field specifies the length (in bytes) of the mode
parameters that follow the Page Length field.
Mode
Parameters
2-n The mode parameters are described in the following subsections.
Disconnect/Reconnect Page (02h)
The Disconnect/Reconnect page allows the host to tune performance of the SCSI
bus. The following table illustrates this page.
Byte
0 0 0 | Page Code (02h)
1 Page Length (0Eh)
2 0 0 0 0 0 0 0 0
3 0 0 0 0 0 0 0 0
4 0 0 0 0 0 0 0 0
5 0 0 0 0 0 0 0 0
6 0 0 0 0 0 0 0 0
7 0 0 0 0 0 0 0 0
8 0 0 0 0 0 0 0 0
9 0 0 0 0 0 0 0 0
10 MSB—Maximum Burst Size
11 Maximum Burst Size—LSB
12 0 0 0 0 0 0 DTDC
13 0 0 0 0 0 0 0 0
14 0 0 0 0 0 0 0 0
15 0 0 0 0 0 0 0 0
Bits
7 6 5 4 3 2 1 0
Page 78 DDS-4 SCSI Interface Manual, Rev. B
Disconnect/Reconnect Page Field Descriptions
The following table describes the fields in the Disconnect/Reconnect page.
Field
Name Bytes Bits Description
Page
Code
Page
Length
Maximu
m Burst
Size
DTDC 12 0-1 The Data Transfer Disconnect Control (DTDC) field values are as
0 The Page Code field must be set to 02h to select this page.
1 The Page Length field must be set to 14 (0Eh), which indicates that 14
parameter bytes follow the Page Length byte.
10-11 The Maximum Burst Size field specifies the maximum amount of data
that the drive can transfer during a Data phase before disconnecting if
the initiator granted the disconnect capability.
This value is in increments of 512 bytes. That is, a value of one means
512 bytes; a value of two means 1024 bytes, and so forth.
A value of zero (0000h) indicates that no limit exists on the amount of
data transferred per connection. In that case, the drive does not have
to disconnect until all requested data is transferred.
On read operations, the drive disconnects when drive buffer is empty,
and the Allocation Length has not been satisfied. On write operations,
the drive disconnects when the drive buffer is full, and the Transfer
Length has not been exhausted.
follows:
00 Data transfer disconnect control is not used.
01 The target does not attempt to disconnect once
the data transfer of a command has been started
until all data the command is to transfer has been
completed.
10 Reserved.
11 The target does not attempt to disconnect once
the data transfer of command has been started
until the command is complete.
Control Mode Page (0Ah)
The Control Mode page allows the host to enable or disable the generation of a
Check Condition when log parameters whose ETC bits are set to 1 meet their
Threshold Condition. The layout of this page is shown in the following table.
Byte
0 0 0 | Page Code (0Ah)
1 Page Length (06h)
2 0 0 0 0 0 0 0 RLEC
3 0 0 0 0 0 0 0 DQUE
4 0 0 0 0 0 0 0 0
5 0 0 0 0 0 0 0 0
6 0 0 0 0 0 0 0 0
7 0 0 0 0 0 0 0 0
Bits
7 6 5 4 3 2 1 0
Commands: Mode Select (15h) Page 79
Control Mode Page Field Descriptions
The following table describes the fields in the Control Mode page.
Field
Name
Page Code 0 0-5 The Page Code field must be set to 0Ah to select this page.
Page
Length
Report Log
Exception
Condition
Disable
Queuing
Bytes Bits Description
1 The Page Length field must be set to 6 (06h), which indicates that 6
parameter bytes follow the Page Length byte..
2 0 A Report Log Exception Condition (RLEC) bit of 1 specifies that the
target reports log exception conditions.
3 0 The Disable Queuing bit must be 1 to signify that tagged queuing is
disabled.
Data Compression Control Page (0Fh)
The Data Compression (DC) Control page specifies whether or not data is
compressed during a WRITE command and whether or not data is decompressed
during a READ command. It also provides for error reporting and selection of the
compression algorithm.
The following table illustrates the DC Control page for drives with data compression.
Byte
0 0 0 0 0 1 1 1 1
1 Page Length (0Eh)
2 DCE DCC 0 0 0 0 0 0
3 DDE RED 0 0 0 0 0
4 Compression Algorithm
5 Compression Algorithm
6 Compression Algorithm
7 Compression Algorithm
8 Decompression Algorithm
9 Decompression Algorithm
10 Decompression Algorithm
11 Decompression Algorithm
12 0 0 0 0 0 0 0 0
13 0 0 0 0 0 0 0 0
14 0 0 0 0 0 0 0 0
15 0 0 0 0 0 0 0 0
Bits
7 6 5 4 3 2 1 0
Page 80 DDS-4 SCSI Interface Manual, Rev. B
DC Control Page Field Descriptions
The following table describes the fields in the DC Control page.
Field Name Bytes Bits Description
DCC 2 6 The DCC (Data Compression Capable) bit indicates that the drive is
capable of data compression. This bit is always 1. Note that data
control compression may be disabled by the switches on the drive.
DCE 2 7 The DCE (Data Compression Enable) bit controls data
compression. If the value is 1, the drive compresses data received
from the host during a WRITE command before it writes the data to
tape in the DDS-DC format.
If the value is 0, the drive does not compress data sent during a
WRITE command, and the host data is written to tape in the
uncompressed DDS format.
DDE 3 7 The DDE (Data Decompression Enable) bit is for control of data
decompression; however, for either value, the drive decompresses
data that has been compressed on the tape before it sends the data
to the host during a READ command.
RED 3 5-6 The RED field specifies when Check Conditions are reported to the
host when reading and the compressed format on the tape changes.
Only RED = 0 is supported. Other mode settings will be accepted,
but will function identically as RED = 0.
Compression
Algorithm
Decompressi
on Algorithm
4-7 The Compression Algorithm byte allows the host to specify the
algorithm that is to be used to compress data. If the drive does not
support the algorithm specified in the Compression Algorithm bytes, a
Check Condition is returned with the Sense Key set to Illegal Request.
8-11 The Decompression Algorithm byte allows the host to specify
the algorithm that is to be used to decompress data. If the
drive does not support the algorithm specified in the
Decompression Algorithm bytes, a Check Condition is returned
with the Sense Key set to Illegal Request.
The following table shows the DCE bit configuration and the supported algorithms.
An algorithm value of 01, which is the default, can be used to determine the
supported algorithm for the drive, DCLZ (20). The value of 20 is returned by the
corresponding MODE SENSE command.
DCE Algorithm Description
0 XX Compression is disabled.
1 00 Compression is disabled.
1 01 Compression is enabled using the default algorithm (DCLZ).
1 02-1F Illegal Request.
1 20 Compression is enabled using the DCLZ algorithm.
1 21-FF Illegal Request.
Device Configuration Page (10h)
The Device Configuration page specifies the appropriate sequential access device
configuration. The following table illustrates this page.
Byte
Bits
7 6 5 4 3 2 1 0
Commands: Mode Select (15h) Page 81
0 0 0 | Page Code (10h)
1 Page Length (0Eh)
2 0 CAP CAF DIS RAW EN C3 N-Group
3 Active Partition
4 0 0 0 0 0 0 0 0
5 0 0 0 0 0 0 0 0
6 MSB—Write Delay Time
7 Write Delay Time—LSB
8 0 BIS RSMK 0 0 0 0 REW
9 0 0 0 0 0 0 0 0
10 0 0 0 EEG SEW 0 0 0
11 0 0 0 0 0 0 0 0
12 0 0 0 0 0 0 0 0
13 0 0 0 0 0 0 0 0
14 0 0 0 0 0 0 0 0
15 0 0 0 0 0 0 0 0
Page 82 DDS-4 SCSI Interface Manual, Rev. B
Device Configuration Page Field Descriptions
The following table describes the fields in the Device Configuration page.
Field
Name
Page Code 0 0-5 The Page Code field must be set to 10h to select this page.
Page
Length
CAP bit 2 6 The CAP (Change Active Partition) bit is used to effect a partition
CAF bit 2 5 The CAF (Change Active Format) bit is used to modify the tape
DIS RAW 2 4 The DIS RAW (Disable Read-After-Write) bit enables and disables
EN C3 2 3 The EN C3 (Enable C3) bit enables and disables C3 ECC code
N-Group 2 0-3 The N-Group field specifies the number of copies of each tape group
Active
Partition
Write
Delay time
BIS bit 8 6 Not supported, must be set to 1
RSMK bit 8 5 The Report Setmarks (RSMK) bit determines whether or not the drive
Bytes Bits Description
1 The Page Length field must be set to 14 (0Eh), which indicates that
14 parameter bytes follow the Page Length byte.
change. When the CAP bit is 1, the drive switches to the partition
specified in the Active Partition field. Once the MODE SELECT
command completes, the logical position is the BOP of the new
partition.
If the CAP bit is 0, no partition change results
format. When the CAF bit is 1, bits 4-0 of byte may be changed.
the read-after-write capability. If the value is 1, read-after-write check
and rewrites are disabled. If the value is 0, read-after-write is enabled.
generation during writing. If the value is 1, C3 ECC code is generated
during writing; if the value is 0, C3 ECC code is not generated.
to record. Values greater than 0 can be used to increase reliability.
3 The Active Partition field is valid only when the CAP bit is 1. This field
specifies the partition number of the new partition to which the drive
switches when the MODE SELECT command completes.
The drive supports a maximum of two partitions. The valid values for
this field are 0 (00h) and 1 (01h). Partition 1 is the first partition on a
dual-partition tape.
6-7 For a WRITE command, the Write Delay Time field indicates to the
drive how long in 100 millisecond increments, to delay writing
buffered data to tape after the last WRITE command.
recognizes setmarks.
If the value is 1, the drive recognizes and reports setmarks during
appropriate read and space operations.
If the value is 0, the drive ignores setmarks. It skips any setmark it
finds during execution of read- or space-type commands.
Commands: Mode Select (15h) Page 83
Field
Name
REW bit 8 0 The default value for the REW bit is 0.
EEG 10 4 Not supported, must be set to 1
SEW bit 10 3 If the Synchronize at Early-Warning (SEW) bit is set to 1 (the default
Bytes Bits Description
The Report Early Warning (REW) bit determines whether or not the
drive reports an early-warning condition on a read operation. The
effect of this bit is different for read and write operations.
For read-type operations, if this bit is 0, the drive does not report the
early-warning condition. This setting is recommended for applications
where the intent is to read data from tape until the end-of-recordeddata (EOD) or end-of-partition (EOP) is reached.
If this bit is 1, the drive returns a Check Condition status with an Endof-Medium (EOM) bit of 1 when the logical early-warning position is
encountered during read-type operations.
For write-type operations, the drive always reports the Check
Condition status when the logical early-warning position is
encountered during a WRITE command. The intent of this action is to
warn the host that the EOP is approaching and that any additional
data will be written at the risk of an unexpected EOP.
If REW = 1 and SEW = 0, the drive returns a Check Condition status
with the sense key set to Volume Overflow when the early-warning
position is encountered during write operations.
value), it causes the drive to flush all buffered write data to tape when
the early-warning position is encountered during writing.
Medium Partition Page (11h)
The Medium Partition page specifies the number and size of partitions to be created
on the medium. The layout of this page is shown in the following table.
Byte
0 0 0 | Page Code (11h)
1 Page Length (06h)
2 Maximum Additional Partitions
3 Additional Partitions Defined
4 0 SDP IDP | PSUM | 0 0 0
5 Medium Format Recognition
6 0 0 0 0 0 0 0 0
7 0 0 0 0 0 0 0 0
8 MSB—Additional Partition Size
9 Additional Partition Size—LSB
Bits
7 6 5 4 3 2 1 0
Page 84 DDS-4 SCSI Interface Manual, Rev. B
Medium Partition Page Field Descriptions
The following table describes the fields in the Medium Partition page.
Field Name Bytes Bits Description
Page Code 0 0-5 The Page Code field must be set to 11h to select this page.
Page
Length
Maximum
Additional
Partitions
Additional
Partitions
Defined
SDP bit 4 6 If the SDP bit is set to 1 the drive automatically formats a partition 1
IDP bit 4 5 If set to 1, the Initiator Defined Partitions bit causes the drive to
PSUM 4 3-4 The partition Size Unit of Measure field defines the units in which the
Medium
Format
Recognition
Additional
Partition
Size
1 If no additional partition is to be defined or if going from a dual-
partition format to a single-partition format, the Page Length field
must be set to 6 (06h), which indicates that 6 parameter bytes follow
the Page Length byte.
If an additional partition is to be defined (dual-partition), this field is
set to 8 (08h).
2 7-0 The Maximum Additional Partitions field indicates the maximum
number of additional partitions supported by the drive that can exist
on a cartridge.
3 The Additional Partitions Defined field specifies the number of
partitions to add when formatting the tape. This bit may be set by the
initiator to 1 (dual partitions) or 0 (single partition), which specifies the
desired number of additional partitions to format on the tape when the
IDP bit = 1 or the SDP bit = 1.
with a capacity of approximately 100 MB. The following conditions
are also required:
• The Additional Partition byte =1. If the Additional Partition byte = 0,
the SDP bit is ignored.
• IDP bit = 0. The SDP and IDP bits cannot be set at the same time.
• If the SDP bit=1 then the Additional Partitions byte must also be 1.
CAUTION: Setting the SDP bit causes the tape to be reformatted to a
2 partition tape. All previous information recorded on the cassette is
destroyed.
partition the tape into 1 or 2 partitions as specified by the Additional
Partitions Defined field and the partition size descriptors.
CAUTION: Setting the IDP bit causes the tape to be reformatted to a
1- or 2-partition tape. All previous information recorded on the
cassette is destroyed.
partition size descriptors select the partition size. The valid values
are:
4 3 Unit
0 0 bytes
0 1 kilobytes
1 0 megabytes
5 7-0 Medium Format Recognition field must be set to 3 indicating the drive
is capable of recognizing both the format and the partitions.
8-9 The Partition Size field indicates the size of the additional partition
(partition 1) formatted on the tape when formatting a dual-partition
tape. The first physical partition is partition 1. The remainder of the
tape is partition 0.
Commands: Mode Select (15h) Page 85
Tape Alert Mode Page 1C
Bits
Byte 7 6 5 4 3 2 1 0
0 0 0 Page Code (1Ch)
1 Page Length (0Ah)
2 Perf Reserved DExcpt Test Reserved LogErr
3 Reserved MRIE
4 Interval Timer (MSB)
5 Interval Timer
6 Interval Timer
7 Interval Timer (LSB)
8 Report Count/Test Flag Number (MSB)
9 Report Count/Test Flag Number
10 Report Count/Test Flag Number
11 Report Count/Test Flag Number (LSB)
DExcpt (Default set to 1):
When this bit is set to zero the reporting method indicated by the MRIE field is used.
When this bit is set to one this indicates that the drive disables all information
exception operations, ignoring the MRIE field (In this mode the software must poll
the TapeAlert Log page). Thus to enable “check condition” mode set the bit to zero.
LogErr (Default set to 0):
When this bit is set to zero, this indicates that the logging of informational exception
is vendor specific, this setting is reserved in Scorpion 40. When this bit is set to
one, Scorpion 40 logs informational exception conditions.
Perf (Default set to 0):
When this bit is set to zero, this indicates that informational exception operations
that can cause delays are acceptable. When this bit is set to one, the drive shall not
cause delays while doing informational exception operations. In Scorpion 40 there is
no significant performance impact from Tape Alert operations.
Test (Default set to 0):
When this bit is set to zero, this indicates that the drive shall not generate any
false/test informational exception conditions.
When this bit is set to one and Test Flag Number is set to zero, the drive will
generate a false informational exception condition based on the MRIE field (the
Interval Timer field is ignored and the Report Count field is used as the Test Flag
Number). When a false information exception condition is posted, the TapeAlert
flags in the Log page are not modified. However, real informational exception
conditions have priority over false informational exception conditions.
The Test bit will be automatically cleared when the false information exception
condition is posted on the first command (excluding Inquiry and Request Sense) that
is received with no real information exception condition pending. The false
informational exception condition will then be reported in the method specified by
the MRIE value, except with the additional sense code set to 0x5DFF. If both the
Page 86 DDS-4 SCSI Interface Manual, Rev. B
Test and DExcpt bits are set to one and Test Flag Number set to zero, then the
MODE SELECT command will return Check Condition, with sense key set to Illegal
Request and extended sense set to Invalid Field in Parameter List.
When the Test bit is set to one and the Test Flag Number is set to valid non-zero
value, the drive will generate/clear a test informational exception condition. The
value of the Test bit returned by a Mode Sense command will remain at zero. The
test action is based on the Test Flag Number value:
1 to 64: This will set in the Log page the TapeAlert flag indicated by the Test Flag
Number. Once the TapeAlert flag is set it is processed normally based on the
DExcpt, MRIE, Interval Count and Report Count values.
-1 to -64: This will clear the TapeAlert flag indicated by the absolute value of the
Test Flag Number. Clearing the flag in this way is equivalent to performing the
specified corrective action for that flag, thus allowing a real information exception
condition to be set if the real error condition occurs for that flag.
32767 (0x7FFF): This will set in the Log page all of the TapeAlert flags that are
supported by the drive. Once the supported TapeAlert flags are set they are
processed normally based on the DExcpt, MRIE, Interval Count and Report Count
values.
MRIE (Default set to 3h):
This field indicates the method used by the drive to report informational exception
conditions. If the informational exception condition was generated by an event that
caused a real Check Condition to occur, then this real Check Condition will override
(i.e. be used instead of) the Check Condition defined in MRIE modes 1h to 5h. The
values defined for the MRIE field are:
0x0: No reporting of informational exception conditions. This method instructs
the tape drive to not report information exception conditions
0x1: Asynchronous Event Reporting. This method instructs the tape drive to
report information exception conditions by using the rules for asynchronous event
reporting as described in the SCSI-3 Architecture Model and the relevant protocol
standard. The sense key shall be set to RECOVERED ERROR and the additional
sense code shall indicate the cause of the information exception condition (which is
5D00h for a Tape Alert event).
0x2: Generate Unit Attention. This method instructs the tape drive to report
information exception conditions by returning a Check Condition status on the next
SCSI command (excluding Inquiry and Request Sense) after an informational
exception condition was detected. The sense key shall be set to UNIT ATTENTION
and the additional sense code shall indicate the cause of the information exception
condition (which is 5D00h for a Tape Alert event). The command that has the Check
Condition shall NOT be executed before the informational exception condition is
reported (and thus needs to be repeated).
0x3: Conditionally generate Recovered Error. This method instructs the tape
drive to report information exception conditions, if reporting of recovered errors is
allowed, by returning a Check Condition status on the next SCSI command
(excluding Inquiry and Request Sense) after an informational exception condition
was detected. The sense key shall be set to RECOVERED ERROR and the
additional sense code shall indicate the cause of the information exception condition
Commands: Mode Select (15h) Page 87
(which is 5D00h for a Tape Alert event). The command that has the Check
Condition shall complete without error before any exception condition may be
reported (and thus does NOT need to be repeated).
0x4: Unconditionally generate Recovered Error. This method instructs the tape
drive to report information exception conditions, by returning a Check Condition
status on the next SCSI command (excluding Inquiry and Request Sense) after an
informational exception condition was detected. The sense key shall be set to
RECOVERED ERROR and the additional sense code shall indicate the cause of the
information exception condition (which is 5D00h for a Tape Alert event). The
command that has the Check Condition shall complete without error before any
exception condition may be reported (and thus does NOT need to be repeated).
0x5: Generate No Sense. This method instructs the tape drive to report information
exception conditions by returning a Check Condition status on the next SCSI
command (excluding Inquiry and Request Sense) after an informational exception
condition was detected. The sense key shall be set to NO SENSE and the additional
sense code shall indicate the cause of the information exception condition (which is
5D00h for a Tape Alert event). The command that has the Check Condition shall
complete without error before any exception condition may be reported (and thus
does NOT need to be repeated).
0x6: Only report informational exception condition on request. This method
instructs tape drive to preserve information exception(s) information. To find out
about information exception conditions the application client polls the tape drive by
issuing an unsolicited REQUEST SENSE command. The sense key shall be set to
NO SENSE and the additional sense code shall indicate the cause of the
informational exception condition (which is 5D00h for a Tape Alert event).
In MRIE modes 1h to 6h, the additional sense code of 5D00h indicates that a Tape
Alert event has occurred on the device, and detailed information about this event is
stored in the Tape Alert Log page. Note that the MRIE modes do not affect the
logging of events in the Tape Alert Log page.
Interval Timer (Default set to 0):
When this field is set to zero, this indicates that the target shall only report the
informational exception condition one time. When this field is set to non-zero, the
value indicates the period in 100 millisecond increments for reporting that an
informational exception condition has occurred. The drive shall not report
informational exception conditions more frequently than the specified timer interval
and as soon as possible after the timer interval has elapsed. A value of
0xFFFFFFFF in the field indicates the timer interval is vendor specific and is a
reserved setting in the Scorpion 40 drive.
Page 88 DDS-4 SCSI Interface Manual, Rev. B
Completion Status
The Completion Status for the MODE SELECT command is shown in the following
table.
Code Message Description
00h Good Status • The drive is ready to perform any appropriate command.
• The defined mode is set and remains set until another MODE
SELECT or RESET command is issued.
• The tape position is not changed.
02h Check
Condition
The MODE SELECT command immediately checks the current page for invalid
parameters or invalid combinations of parameters before executing. If such an
exception is found, the drive returns a Check Condition status byte with an Illegal
Request sense key.
Extended Sense Byte 02h:
Code Message Description
02h Not Ready No cartridge is inserted in the drive.
05h Illegal Request The CDB or Parameter List contains
an invalid bit.
06h Unit Cartridge was changed prior to accepting
Attention this command or the drive was reset
prior to this command.
If the initiator does not set the Page Length field of the Mode Page to the value
indicated in the Mode Page definition (for example, 0Eh for the Device Configuration
Page, 01h for the Read/Write Error Recovery Page, and so forth), the drive
terminates the MODE SELECT command and returns a Check Condition status
bytes with an Illegal Request sense key. The Additional Sense Code and Additional
Sense Code Qualifier are set to Invalid Field in Parameter List.
Loading...