Atlas 15K II

ENTERPRISE
ULTRA320 SCSI HARD DRIVE 36GB 73GB 147GB Reliability and performance: the first 15K RPM enterprise hard disk drive designed to deliver 1.4 million hours MTTF* with leading IOPS performance The Atlas 15K II Strengths
The Atlas 15K II SCSI drive is one of the industrys fastest and reliable 15K RPM hard disk drive, with average seek times as fast as 3.0ms, maximum sustained data transfer rate of 98MB/sec, and designed for a 1.4 million hours MTTF* The Atlas 15K II drive can manage the most. demanding I/O-intensive and high-bandwidth applications with superior reliability and adaptability. Maxtors intelligent drive technology, MaxAdapt, allows the Atlas 15K II drive to adapt to various system designs and conditions for ease of integration and maintains optimal performance.
Maxtor MaxAdapt technology offers the following key features:
Intelligent MaxAdapt features
Adaptive Bias Estimation (ABE) Maintains consistent performance by adjusting the bias applied to the actuator Virtual Cache Lines (VCL) Allows dynamic assignment of cache segments Adaptive Active Filtering (AAF) Improves signal integrity Rotational Vibration Compensation (RVC)Monitors and corrects for external vibration
15,000 RPM ENTERPRISECLASS HARD DRIVE EXCEPTIONAL CAPACITY AND PERFORMANCE
Third-generation Ultra320 SCSI Maximum sustained data rate up to 98MB/sec 8MB cache (ECC protected) Seek time as fast as 3.0ms

EXTRAORDINARY VALUE

Ease of integration MaxAdapt

Performance

Anticipatory read and write streams Minimizing hardware latency Opportunistic pre-fetch Uses latency between commands to read ahead, improving performance Auto Read mode Improves random and multiple sequential stream performance
Up to 39% more IO/sec than the Atlas 10K V SCSI investment protection

SUPERIOR RELIABILITY

Designed for 1.4 million hours MTTF* 24x7 dependability Proven Atlas family architecture and firmware Shock Protection System Thermal sensor S.M.A.R.T. features RoHS compliant

Reliability

Improved S.M.A.R.T. and self-diagnostics State-of-the-art manufacturing test processes Multiple temperature tests
Best-fit enterprise applications
OLTP SAN environments NAS environments High-performance workstations 3-D animation Scientific data processing Economic modeling Data mining Web servers Audio/video Digital imaging
SPECIFICATIONS 36.7 Formatted Capacity (GB) Bytes per Sector Interface Interface Connectors Disk Drive Configuration Disks Heads Performance Specifications Seek Time Average Read/Write (ms) Full Stroke Read/Write (ms) Spindle Speed (RPM) Average Rotational Latency (ms) Transfer Rates SCSI Maximum Burst (MB/s) SCSI Maximum Host (MB/s) Maximum Sustained (MB/sec) Cache (MBytes) Reliability Specifications Contact Start Stop (CSS) Data Error Rate per Bits Read Recoverable Nonrecoverable Limited warranty (years) ORDER INFORMATION Model Number 8E036L0 8E036J0 8E073L0 8E073J0 8E147L0 8E147J0 RoHS Model Number 8K036L0 8K036J0 8K073L0 8K073J0 8K147L0 8K147J0 Capacity 36.7GB 36.7GB 73.5GB 73.5GB 147.1GB 147.1GB Interface Ultra320 Ultra320 Ultra320 Ultra320 Ultra320 Ultra320 Connector 68-pin Wide LVD 80-pin SCA-2 68-pin Wide LVD 80-pin SCA-2 68-pin Wide LVD 80-pin SCA-2 <10 per 1012 <1 per 50,8 3.0/3.4 3.1/3.5 0.3/0.5 8.0/9.0 15,3.4/3.8 Track-to-Track Read/Write (ms) 36.7 73.5 73.5 147.1 147.1 Environmental Specifications Operating Temperature (C) Maximum Case Temperature (C) Non-Condensing Humidity (%) Shock 2 ms (G) R/W Rotational Shock (rad/sec2) (rad/sec2) 3.6 Vibration 5-500 Hz (G) Rotational Vibration Altitude (feet) Acoustics, Idle (bels) Non-Operating Temperature (C) Non-Condensing Humidity (%) Shock 2 msec (G) Rotational Shock Altitude (feet) Power Specifications Voltage Requirements Idle Power (W) Physical Dimensions Width max (inches/mm) Length max (inches/mm) Height max (inches/mm) Weight max (lb/kg) 4.0 / 101.6 5.787 / 147.0 1.028 / 26.1 2.0 / 0.91 5V +/- 5% 6.8 8.4 12V +/-5% 13 (rad/sec2) Vibration 5-500 Hz (G) -40 to to 25,-1,000 to 40,to 5 to 95 63/30 7,000 1.-1,000 to 10,000 3.6 3.8 36.7 73.5 147.1

512, 516, 518, 520, 522, 524 Ultra320 SCSI (Backwards compatible with Ultra160, Ultra2, UltraSCSI) 68-pin WIDE; 80-pin SCA-2
For support or information, call us at 1-800-2Maxtor or visit us at www.maxtor.com
All Maxtor products are backed by our leading service and support staff. Service includes: Advance replacement in two business days 24-hour on-line troubleshooting tools and email Telephone support representatives available Monday-Friday during business hours (except holidays) To speak with a Maxtor product support representative in the U.S. and Canada, call 1-800-2MAXTOR, Mon.-Fri. from 7 a.m. to 4 p.m. (PST). In Europe, call +Mon.-Thur. from 8:30 a.m. to 5 p.m. (GMT) and Fri. 8:30 a.m. to 4 p.m. (GMT). In Australia, call +9369 3662. In Japan, call 005316-53616, and in Singapore, call 65-6852-0220 or 1-800-481-6788.
For purposes of measuring drive capacity, a megabyte (MB) means 1,000,000 bytes, a gigabyte (GB) means 1,000,000,000 bytes. Total accessible capacity varies depending on operating environment. Seek times are at nominal conditions and include settling. Specifications subject to change without notice. * MTTF refers to the estimated mean time to failure based upon a statistical sample. 2004 Maxtor Corporation. Maxtor, Atlas, MaxAdapt, What drives you and the Maxtor stylized logo are registered trademark of Maxtor Corporation. Shock Protection Systemis a trademark of Maxtor Corporation. Maxtor Corporation, 500 McCarthy Boulevard, Milpitas, CA, 95035. DS-Atlas15KII-8/04-CL.

6.8.2 Transfer Error Protection...6-4 6.8.3 Addressing Error Protection..6-4 6.8.4 Data Sector Reallocation Error Protection..6-4 6.8.5 Data Verification....6-5 6.9 TAGGED COMMAND QUEUING... 6-5 6.10 COMMAND REORDERING.. 6-5 6.11 BANDED RECORDING.... 6-5 6.12 DIAGNOSTICS.... 6-6 6.12.1 Power On Self Test (POST)..6-6 6.12.2 Periodic Self-Adjustments..6-6 6.12.3 Host Diagnostics...6-7 6.12.4 Log Sense...6-7 6.13 ERROR RECOVERY... 6-7 6.13.1 Seek Error Recovery...6-7 6.13.2 Data Read Errors...6-7 6.13.3 Reallocation of Bad Blocks...6-7 6.14 THE ULTRA160 and ULTRA320 LOW VOLTAGE DIFFERENTIAL (LVD) SCSI INTERFACE... 6-8 6.14.1 DIFFSENS...6-9 6.14.2 DIFFSENS SWITCHING...6-9 6.14.3 DT AND ST CLOCKING... 6-10 6.14.4 CYCLIC REDUNDANCY CHECKING.. 6-11 6.14.5 DOMAIN VALIDATION... 6-11 6.14.6 Adaptive Active Filter (AAF).. 6-11 6.14.7 Transmitter Pre-Compensation with Cutback.. 6-12 6.14.8 Free Running Clock (FRC).. 6-12 6.14.9 Skew Compensation.. 6-12 6.14.10 Information Unit Transfers.. 6-12 6.15 HOT PLUGGING/REMOVAL AND INSERTION OF SCSI DEVICES.. 6-12 6.15.1 System Considerations... 6-13 6.15.2 APPLICABLE SCSI PHYSICAL DOCUMENTS.. 6-14 Appendix A QUICK REFERENCE A.1 SCSI-2/SCSI-3 Equivalent Terminology.. A.2 SCSI Commands and Messages... A.3 SCSI Pages.... A.4 Sense Keys... A.5 Status Codes....

A-1 A-2 A-3 A-5 A-5

Appendix B NEGOTIATED RATE INFORMATION PAGE REFERENCE B.1 TRANSFER PERIOD FACTOR FIELD... B-1 B.2 Settings for the QAS, DT, and IU fields.. B-2

List of Figures

Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure
3-1 3-2 3-3 3-4 3-5 3-6 3-7 3-8 3-9 6-1
Dimensions for the Maxtor Atlas 15K II Hard Disk Drives.. 3-3 Drive Packing Assembly... 3-5 Jumper Locations on the 68-Pin Wide SCSI Drive PCB. 3-7 Pin Locations on SCA-2 Connector.. 3-11 J1 Interface Connector Configurations.. 3-14 Mounting Dimensions Maxtor Atlas 15K II Drives.. 3-22 Mounting Screw Clearance for Maxtor Atlas 15K II Drive. 3-23 Lengthwise Airflow Cooling... 3-24 Crosswise Airflow Cooling... 3-25 ST and DT CLocking... 6-11

List of Tables

Table 3-1 Table 3-2 Table 3-3 Table 3-4 Table 3-5 Table 3-6 Table 3-7 Table 4-1 Table 4-2 Table 4-3 Table 4-4 Table 4-5 Table 4-6 Table 4-7 Table 4-8 Table 4-9 Table 4-10 Table 5-1 Table 5-2 Table 5-3 Table 5-4 Table 5-5 Table 5-6 Table 5-7 Table 5-8 Table 5-9 Table 5-10 Table 5-11 Table 5-12 Table 5-13 Table 5-14 Table 5-15 Table 5-16 Table 5-17 Table 5-18 Table 5-19
SCSI ID Selection on Option Connector (68-Pin SCSI Connector Drives). 3-8 SCSI ID Pin Assignments (SCA-2 Connector Versions of the Disk Drive). 3-12 Spin Up on Power On Options... 3-13 68-Pin Wide LVD Pin Assignments... 3-16 80-Pin SCA-2 LVD Pin Assignments... 3-17 68-Pin Wide Single-Ended Pin Assignments... 3-19 80-Pin SCA-2 Single-Ended Pin Assignments.. 3-20 Specifications.... 4-1 Formatted Capacity... 4-3 Timing Specifications for Maxtor Atlas 15K II.. 4-4 Power Supply Voltages... 4-5 Power Reset Limits1.... 4-5 Power Dissipation in Various Modes (Low-Profile Drives). 4-6 Acoustical CharacteristicsSound Power per ISO 7779.. 4-7 Environmental Specifications... 4-8 Shock and Vibration Specifications.. 4-9 Error Rates... 4-11 Supported SCSI Commands.. 5-1 Typical (6-Byte) Command Descriptor Block-Data Format.. 5-3 Command Descriptor Block-Field Descriptions.. 5-4 Typical (6-Byte) Command Descriptor Block - Data Format.. 5-5 Command Descriptor Block Control Field-Field Descriptions. 5-6 Status Codes... 5-7 Length Fields.... 5-8 DATA-Phase Command Contents.. 5-9 CHANGE DEFINITION Command Descriptor Block-Data Format. 5-11 CHANGE DEFINITION-Field Descriptions.. 5-11 FORMAT UNIT Command Descriptor Block-Data Format.. 5-13 FORMAT UNIT Command-Field Descriptions.. 5-13 FORMAT UNIT Command Supported Options.. 5-14 FORMAT UNIT Parameter List-Data Format.. 5-14 FORMAT UNIT Defect List Header-Data Format.. 5-15 FORMAT UNIT Defect List Header-Field Descriptions. 5-15 FORMAT UNIT Defect Descriptor-Block Format.. 5-16 FORMAT UNIT Defect Descriptor- Physical Sector and Bytes From Index Format... 5-16 FORMAT UNIT Initialization Pattern Descriptor-Data Format.. 5-16

Table 5-20 Table 5-21 Table 5-22 Table 5-23 Table 5-24 Table 5-25 Table 5-26 Table 5-27 Table 5-28 Table 5-29 Table 5-30 Table 5-31 Table 5-32 Table 5-33 Table 5-34 Table 5-35 Table 5-36 Table 5-37 Table 5-38 Table 5-39 Table 5-40 Table 5-41 Table 5-42 Table 5-43 Table 5-44 Table 5-45 Table 5-46 Table 5-47 Table 5-48 Table 5-49 Table 5-50 Table 5-51 Table 5-52 Table 5-53 Table 5-54 Table 5-55 Table 5-56 Table 5-57 Table 5-58 Table 5-59 Table 5-60 Table 5-61 Table 5-62 Table 5-63 Table 5-64 Table 5-65
FORMAT UNIT Initialization Pattern Descriptor-Field Descriptions. FORMAT UNIT Initialization Pattern Type.. INQUIRY Command Descriptor Block-Data Format. INQUIRY Command Descriptor Block-Field Descriptions.. Standard Inquiry Data Page-Data Format... Standard Inquiry Data Page-Field Descriptions.. Supported Vital Product Data Pages-Data Format.. Vital Product Data-Page Codes... Unit Serial Number Page-Data Format.. Unit Serial Number Page-Field Descriptions... Implemented Operating Definition Page-Data Format. Implemented Operating Definition Page-Field Descriptions.. ASCII Implemented Operating Definition Page - Data Format. Device Identification Page - Data Format.. Device Identification Page - Field Description.. Command Support Data Page-Data Format.. Command Support Data Page-Field Descriptions.. Command Support Data Page Command or Operation Codes. LOG SELECT Command Descriptor Block-Data Format.. LOG SELECT Command Descriptor Block-Field Descriptions.. Disk Drive Log Pages... LOG SENSE Command Descriptor Block-Data Format.. LOG SENSE Command Descriptor Block-Field Descriptions. LOG SENSE Log Page Format-Data Format.. LOG SENSE Log Page Format-Field Descriptions.. Generic Log Parameter-Data Format.. Generic Log Parameter-Field Descriptions.. MODE SELECT (6) Command Descriptor Block-Data Format. MODE SELECT (6) Command Field Descriptions.. Initiator-Changeable Mode Pages.. Mode Page Types... Mode Parameter List-Data Format... Mode Parameter List-Field Descriptions... Mode Parameter Header (6-Byte)-Data Format... Mode Parameter Header- Field Descriptions... Mode Parameter Block Descriptor-Data Format.. Mode Parameter Block Descriptor-Field Descriptions.. Categories of Changeable Pages... Unit Attention Condition Page (Page 0).. Unit Attention Condition Page (Page 0).. Read-Write Error Recovery Page-(Page 1).. Read-Write Error Recovery Page - Field Descriptions.. DisconnectReconnect Page (Page 2)... Disconnect-Reconnect-Field Descriptions (Page 2). Verify Error Recovery Page-(Page 7).. Verify Error Recovery Page-Field Descriptions (Page 7)..
5-17 5-17 5-18 5-18 5-19 5-20 5-21 5-21 5-21 5-22 5-22 5-22 5-23 5-23 5-25 5-25 5-26 5-27 5-29 5-29 5-30 5-31 5-31 5-32 5-32 5-32 5-33 5-34 5-34 5-35 5-36 5-37 5-37 5-37 5-37 5-38 5-38 5-39 5-40 5-40 5-41 5-41 5-43 5-43 5-45 5-45

Electrostatic Discharge (ESD) Protection
Various electrical components within the disk drive are sensitive to static electricity and Electrostatic Discharge (ESD). Even a static buildup or discharge that is too slight to feel can be sufficient to destroy or degrade a component's operation. To minimize the possibility of ESD-related damage to the drive, we strongly recommend using both, a properly installed workstation anti-static mat and a properly installed ESD wrist strap. When correctly installed, these devices reduce the buildup of static electricity which might harm the drive. Observe the following precautions to avoid ESD-related problems: Use a properly installed anti-static pad on your work surface. Always use a properly fitted and grounded wrist strap or other suitable ESD protection when handling the drive and observe proper ESD grounding techniques. Hold the drive only by its sides. Do not touch any components on the PCBA. Leave the drive in its anti-static bag until you are ready to install it in the system. Place the drive on a properly grounded anti-static work surface pad when it is out of its protective anti-static bag. Do not use the bag as a substitute for the work surface anti-static pad. The outside of the bag may not have the same anti-static properties as the inside. It could actually increase the possibility of ESD problems. Do not use any test equipment to check components on the electronics module. There are no user-serviceable components on the drive.

SPACE REQUIREMENTS

The Maxtor Atlas 15K II hard disk drive is shipped without a faceplate and comes in the following SCSI interface configurations: 68-pin Wide SCSI 80-pin SCA-2 SCSI Figure 3-1 shows the mechanical dimensions of each of the drives.
Figure 3-1 Dimensions for the Maxtor Atlas 15K II Hard Disk Drives

Shock Feet

Maxtor Atlas 15K II hard disk drives are outfitted with plastic shock feet on the bottom edge of the base casting, near the corners, beneath the side mounting holes (translucent), and near the corners of the top cover next to the screws (black). The shock feet give an additional level of isolation to prevent the head and disk damage that occasionally occurs during unpacking, staging, and installation. The shock feet attenuate the short-pulse shocks that occur when placing the drive on a hard surface.To provide optimal protection the shock feet are designed to exceed the form factor when uncompressed.

UNPACKING INSTRUCTIONS

CAUTION: The maximum limits for physical shock can be exceeded if the drive is not handled properly. Special care should be taken not to bump or drop the drive. 1. Open the shipping container and remove the packing assembly that contains the drive. 2. Remove the drive from the packing assembly. CAUTION: During shipment and handling, the antistatic electrostatic discharge (ESD) bag prevents electronic component damage due to electrostatic discharge. To avoid accidental damage to the drive, do not use a sharp instrument to open the ESD bag. Save the packing materials for possible future use. 3. When you are ready to install the drive, remove it from the ESD bag.

INTERFACE CONNECTOR (J1)

The configuration of J1 is different for the 68-pin and 80-pin SCSI variations. Figure 3-5 shows the various connector styles.
Figure 3-5 J1 Interface Connector Configurations
68-Pin Wide SCSI Connector - LVD

Table 3-4

SIGNAL NAME CONNECTOR CONTACT NUMBER
68-Pin Wide LVD Pin Assignments
CABLE CONDUCTOR NUMBER CONNECTOR CONTACT NUMBER SIGNAL NAME
+DB (12) +DB (13) +DB (14) +DB (15) +DB (P1) +DB (0) +DB (1) +DB (2) +DB (3) +DB (4) +DB (5) +DB (6) +DB (7) +DB (P) GROUND DIFFSENS TERMPWR TERMPWR RESERVED GROUND +ATN GROUND +BSY +ACK +RST +MSG +SEL +C/D +REQ +I/O +DB (8) +DB (9) +DB (10) +DB (11)
DB (12) DB (13) DB (14) DB (15) DB (P1) DB (0) DB (1) DB (2) DB (3) DB (4) DB (5) DB (6) DB (7) DB (P) GROUND GROUND TERMPWR TERMPWR RESERVED GROUND ATN GROUND BSY ACK RST MSG SEL C/D REQ I/O DB (8) DB (9) DB (10) DB (11)
Note: The conductor number refers to the conductor position when using 0.635 mm (0.025 inch) centerline flat ribbon cable. Other cable types may be used to implement equivalent contact assignments.
3.5.1.1 68-Pin Wide Mating Connector
The cable plug connector that mates with the 68-pin hard disk drive connector is the AMP AMPLIMITE.050 Series III, part number 749925-5.
80-Pin SCA-2 SCSI Connector - LVD

Table 3-5

80-PIN CONNECTOR CONTACT AND SIGNAL NAME
80-Pin SCA-2 LVD Pin Assignments
CABLE CONDUCTOR NUMBER 80-PIN CONNECTOR CONTACT AND SIGNAL NAME
12V CHARGE 12V 12V 12 V OPT 3.3 VOLTS OPT 3.3 VOLTS -DB (11) -DB (10) -DB (9) -DB (8) -I/O -REQ --C/D SEL -MSG -RST -ACK -BSY -ATN -DB (P0) -DB (7) -DB (6) -DB (5) -DB (4) -DB (3)

NOT APPLICABLE

12V GROUND 12V GROUND 12V GROUND MATED 1 OPT 3.3V CHARGE DIFFSNS +DB (11) +DB (10) +DB (9) +DB (8) +I/O +REQ +C/D +SEL +MSG +RST +ACK +BSY +ATN +DB (P0) +DB (7) +DB (6) +DB (5) +DB (4) +DB (3)
DB (2) DB (1) DB (0) DB (P1) DB (15) DB (14) DB (13) DB (12) 5V 5V 5V CHARGE SPINDLE SYNC RMT START SCSI ID (0) SCSI ID (2) 3.5.2.1
+DB (2) +DB (1) +DB (0) +DB (P1) +DB (15) +DB (14) +DB (13) +DB (12) MATED 2 5V GROUND 5V GROUND ACTIVE LED OUT DLYD_START SCSI ID (1) SCSI ID (3)
80-Pin SCA-2 Mating Connectors
The compatible 80-pin mating connectors are: Right-angle receptacle, AMP CHAMP,.050 Series I, part number 787535-1. Vertical receptacle, AMP CHAMP.050 Series I, part number 787311-2. Vertical receptacle, BERG part number 71780-001.

FmtData

CmpLst

Defect List Format

Vendor-Specific Interleave
Five Forms of FORMAT UNIT Commands
Five different forms of the FORMAT UNIT command are supported through different combinations of the Format Data bit, Complete List bit, and the information in the Defect List Length field. (Refer to FORMAT UNIT Defect Header List for a description of the Defect List Length field.) The different command forms give the initiator control over the contents of the Grown Defect List. Table 5-13 describes the contents of the Grown Defect List after FORMAT UNIT command execution.
Table 5-13 FORMAT UNIT Command Supported Options

FmtDat

Defect List Length

N/A 0 0

Contents of Grown Defect List after FORMAT UNIT Command Execution
All sectors found to be bad, but not listed in the Primary Defect List or Grown Defect List, are added to the Grown Defect List. Same as above. The existing Grown Defect List is discarded. A new Grown Defect List is generated, containing All sectors found to be bad, but not listed in the Primary Defect List. The Grown Defect List contains: Original Grown Defect List All sectors found to be bad, but not listed in the Primary Defect List. A list of sectors supplied by the initiator. These sectors are passed in defect descriptors in the data-out buffer transfer that occurs as part of the FORMAT UNIT command. Same as the case above, except that the current Grown Defect List is discarded before formatting begins.
FORMAT UNIT Parameter List
The FORMAT UNIT Parameter List (Table 5-14) consists of a Defect List Header (Table 5-15) followed by one or more Defect Descriptors. Descriptors are either four bytes or eight bytes in length.
Table 5-14 FORMAT UNIT Parameter List-Data Format Bit Byte
Defect List Header Initialization Pattern Descriptor (if any) Defect Descriptors (if any)

5.8.2.1

FORMAT UNIT Defect List Header
The FORMAT UNIT Defect List Header (Table 5-15) provides several optional format control bits to give the initiator more control over the defect lists. Table 5-16 provides descriptions of the data fields in the Defect List Header
Table 5-15.FORMAT UNIT Defect List Header-Data Format Bit Byte

2-3 FOV DPRY DCRT

Reserved STPF
Table 5-16 FORMAT UNIT Defect List Header-Field Descriptions Name

FOV DPRY DCRT

The Format Options Valid bit indicates that the remaining option bits in this byte are valid. If this bit is not set, the remaining bits, except Immed, are ignored. The Disable Primary bit specifies whether the drive should ignore the Primary Defect List during the format to identify defective areas on the media. The drive's default is 0, indicating that the drive should replace sectors found in the Primary Defect List during the format. The Disable Certification bit specifies whether the drive should certify the media during the format. The drives default is 0, indicating that the drive should certify the media. The Stop Format bit specifies the error conditions under which the drive is to terminate the format. The state of the bit is ignored and the drive acts as if this bit is set. If either the Grown Defect List or the Primary Defect List is not found, the format operation terminates with a CHECK CONDITION status and a sense key of MEDIUM ERROR The Initialization Pattern bit signals the drive that the FORMAT UNIT Parameter List contains initialization pattern information. An Initialization Pattern bit of 1 indicates that an initialization pattern descriptor (Table 5-17) is included in the FORMAT UNIT parameter list immediately following the Defect List Header. An Initialization Pattern bit of 0 indicates that an initialization pattern descriptor (Table 5-17) is not included in the FORMAT UNIT parameter list, and the drive will use its default initialization pattern of all zeros. The Disable Saving Parameter prohibits the drive from preserving MODE SELECT parameters received while the Format is in progress. This bit is ignored, and the drive acts as if it were set. When the Immediate bit is 0, the drive returns a status after the format operation completes. If the disconnect privilege is granted in the accompanying IDENTIFY message, the drive disconnects from the bus after fetching and validating the Command Descriptor Block and the FORMAT UNIT Parameter List. If the disconnect privilege is not granted, the drive stays connected to the bus during the entire FORMAT UNIT command. When the Immediate bit is set to 1, the drive returns a status after the Command Descriptor Block and FORMAT UNIT Parameter List have been fetched and validated. In this case, the drive ignores the IDENTIFY message's disconnect privilege bit and remains connected to the bus while it validates the Command Descriptor Block and Parameter List. It returns the completion status before it disconnects. The drive then proceeds with the format. Vendor-specific. Must be 0. Gives the length in bytes of the Defect Descriptors that follow. A value of 0 is valid and means that no Defect Descriptors follow. A CHECK CONDITION status is returned if this value is not a multiple of the defect descriptor size. The Defect List Length is equal to four times the number of defect descriptors if Block format is used, or eight times if Bytes From Index format or Physical Sector format is used.

2-6 7-Reserved

Operation Code (55h) PF Reserved Parameter List Length Control Reserved SP
Table 5-86 Mode Parameter Header (10-Byte)-Data Format Bit Byte

0-3 4-5 6-7

(MSB) Mode Data Length (LSB) Medium Type Device-Specific Parameter Reserved (MSB) Block Descriptor Length (LSB)
Table 5-87 Mode Parameter Block Descriptor-Data Format Bit Byte
(MSB) Number of Blocks (LSB) Density Code (MSB) Block Length (LSB)
Table 5-88 Mode Parameter Block Descriptor-Field Descriptions Data Field
Number of Blocks Density Code Block Length
The value in this field specifies the number of logical blocks on the disk to which the density code and block length fields apply. A value of 0 in this field indicates that all of the remaining logical blocks on the disk have the medium characteristics specified. The Density Code field is unique for each device type. Direct access devices value is zero. The value in this field specifies the length in bytes of each logical block described by the block descriptor.
MODE SENSE (6) Command (1Ah)
The MODE SENSE (6) command allows the initiator to read the drive's sector descriptor and mode pages. The MODE SENSE Command Descriptor Sector is shown in Table 5-89 and the data fields are described in Table 5-90. The Mode Parameter Header is shown in Table 5-91 and the data fields are described in Table 5-92.
Table 5-89 MODE SENSE (6) Command Descriptor Block-Data Format Bit Byte

PC Reserved

Operation Code (1Ah) DBD Page Code Sub Page Code Allocation Length Control Reserved
Table 5-90 MODE SENSE Command-Field Descriptions Data Field
Disable Block Descriptor. When the value is 0, the drive returns its Block Descriptor followed by the requested page(s). When the value is 1, the drive returns only the requested pages. Page Control. Determines which type of page values are returned. These values are:
00b Current 01b Changeable 10b Default 11b Saved The Page Code field determines the pages to be returned to the initiator in the command's data-in buffer transfer. Refer to Table 5-49 and Table 5-93 for a complete list of the Mode Pages (changeable and read-only). A Page Code of 3Fh causes all supported pages to be returned. The Sub Page Code field determines the sub pages to be returned to the initiator in the command's data-in buffer transfer. A Sub Page Code of 0xFF causes all supported sub pages to be returned.

Sub Page Code

Table 5-91 Mode Parameter Header (6 Byte)-Data Format Bit Byte
Table 5-92 Mode Parameter Header and Block Descriptor-Field Descriptions Data Field

This field indicates how much space has been reserved for the returned parameter list (Read Keys or Read Reservations parameters). The actual length of the parameter data is indicated in the parameter data field for those parameters. If the Allocation Length is not sufficient to contain the entire list of parameters, the first portion of the list that does fit is returned. If it is determined that the remainder of the list is required, the client should send a new PERSISTENT RESERVATION IN command with an Allocation Length field large enough to contain the entire list of parameters.
The data format and field descriptions for the Read Keys Parameters are described in Table 5-104 and Table 5-105 respectively.
Table 5-104 Read Keys Parameters-Data Format Byte

(MSB) Generation

(LSB) 4-7 (MSB) Additional Length (n 7) (LSB) (Reservation Key List Follows in Bytes 8 n) 8 - 15 n 7 to n (MSB) First Reservation Key (LSB) (MSB) Last Reservation Key (LSB)
Table 5-105 Read Keys Parameters-Field Descriptions Data Field Description
The value in this field is a 32-bit counter in the device server that is incremented each time a PERSISTENT RESERVATION OUT command requests a Register, Clear, Pre-empt, or Pre-empt and Clear operation. Note that PERSISTENT RESERVATION IN commands do not increment the counter, nor do PERSISTENT RESERVATION OUT commands that perform a Reserve or Release service action, or by a PERSISTENT RESERVATION OUT command that is not done due to an error or a reservation conflict. The value in the Generation field is set to 0 as part of the power on or reset processes. The value in the Generation field allows the application client that examines the value to verify that the configuration of the initiators attached to a logical unit has not been modified by another application client without any notification of the application client doing the examination. Additional Length This field contains the count of the number of bytes that are in the Reservation Key list (bytes 8 n). Note that this field contains the number of bytes in the reservation key list regardless of the value prescribed by the Allocation Length field in the commands CDB. Each of the Reservation Keys appear as items in a list as bytes 8 through n. Each entry reflects an 8-byte reservation key registered with the device server via the PERSISTENT RESERVATION OUT, Reserve, Pre-empt, Pre-empt and Clear, or Register service actions. Each key can be examined by the application client for correlation with a set of initiators and SCSI ports

Exclusive Access

READS: Shared; any application client on any initiator may execute commands that perform transfers from the disk to the initiator.
WRITE Exclusive, Registrants Only
WRITES: Exclusive; any command from an initiator that has not previously performed a Register service action with the device server that attempts a transfer to the disk results in a reservation conflict ADDITIONAL RESERVATIONS: Allowed; any initiator may reserve the logical unit, extents, or elements as long as the persistent reservations to not conflict with any reservations already known to the device server. READS: Exclusive; any command from an initiator that has not previously performed a Register service action with the device server that attempts a transfer from the disk results in a reservation conflict.
Exclusive Access, Registrants Only
WRITES: Exclusive; any command from an initiator that has not previously performed a Register service action with the device server that attempts a transfer to the disk results in a reservation conflict ADDITIONAL RESERVATIONS: Allowed; any initiator may reserve the logical unit, extents, or elements as long as the persistent reservations to not conflict with any reservations already known to the device server.

7h - Fh

Table 5-111 represents graphically all possible combinations of Persistent Reservations Being Attempted when Persistent Reservations are already held for each of the types of persistent reservations.
Table 5-111 When Do Conflicts Between Existing Reservations and New Reservations Exist? Persistent Reservation Already Held
Persistent Reservation Being Attempted LU READ Shared EX READ Shared LU EX NN NN LU WRITE Exclusive EX LU READ Exclusive EX LU Exclusive Access * EX LU Shared Access * EX LU WRITE Exclusive RO EX LU Exclusive Access RO EX YY YO YY YO YY YO NN NN YY YO WRITE Exclusive LU EX YY YO YY YO YY YO YY YO YY YO YY YO READ Exclusive LU EX YY YO YY YO YY YO YY YO YY YO YY YO Exclusive Access * LU EX YY YO YY YO YY YO YY YO YY YO YY YO Shared Access * LU EX NN NN YY YO YY YO YY YO NN NN YY OO NN NN WRITE Exclusive RO LU EX YY YO YY YO YY YO YY YO YY YO YO NN NN Exclusive Access RO LU EX YY YO YY YO YY YO YY YO YY
KEY: LU = Logical Unit Scope EX = Extent or Element Scope RO = Registrants Only * = Conflicts with all reservation requests from other initiators.
N = No Conflict Y = Conflict O = Conflict occurs if extent or element overlaps with existing extent or element reservation.
PERSISTENT RESERVATION OUT Command (5Fh)
The PERSISTENT RESERVATION OUT command is a 10-byte command used to reserve a logical unit or an extent within a logical unit for the exclusive or shared use by an initiator. The command is used in conjunction with the PERSISTENT RESERVATION IN command; it is not used with the RESERVE and RELEASE commands. Persistent reservations conflict with reservations made via the RESERVE command. Initiators that perform PERSISTENT RESERVATION OUT actions are identified by a reservation key assigned by the application client. The client may use the PERSISTENT RESERVATION IN command to identify which other initiators within a system hold conflicting or invalid persistent reservations and use the PERSISTENT RESERVATION OUT command to preempt those reservations if necessary. Note that since persistent reservations are not reset by the TARGET RESET task management function or other global actions, they can be used to enact device sharing among multiple initiators. The PERSISTENT RESERVATION OUT and PERSISTENT RESERVATION IN commands provide the means for resolving contentions in multiple-initiator systems with multiple port target. By using the reservation key to identify persistent reservations, it is possible to determine which ports hold conflicting persistent reservations and to take over such reservations from failing or greedy initiators. illustrates the format of the PERSISTENT RESERVATION OUT command. Table 5-113 explains the data fields of the command.

Extent

Element

3h Fh Type

The value of the Type field specifies the characteristics of the persistent reservation being established for all data blocks within the extent or within the logical unit. Refer to Table 5-115 for the applicable Type codes and their meanings Fields contained in the PERSISTENT RESERVATION OUT parameter list specify the reservation keys and extent information required to perform a persistent reservation service action. The parameter list is 24 bytes in length; the Parameter List Length field contains 24 (18h) bytes.
Table 5-114 provides detailed descriptions of each of the PERSISTENT RESERVATION OUT commands seven possible service actions (Service Action codes appear in bits of Byte 1).
Table 5-114 PERSISTENT RESERVATION OUT Commands Service Action Descriptions Code Name Description
When the command executes a Register service action, it registers a reservation key with a device server without generating a reservation. The device server holds these reservation keys from each initiator that performs a PERSISTENT RESERVATION OUT command with a Register service action until the key is changed by a new PERSISTENT RESERVATION OUT command with Register service action from the same initiator, or until the initiator registration is removed by:
Powering down the logical unit, if the last Activate Persist Through Power Loss (APTPL; see Figure 5-55 and Table 5-61) received by the device server was 0; Performing a Clear service action;

00h Register

Performing a Pre-empt service action; Performing a Pre-empt and Clear service action; or Performing a Register service action from the same initiator with the value of the service action reservation key set to 0.
When a reservation key has not yet been established or when the reservation key has been removed, a reservation key of 0 is used when the initiator performs a PERSISTENT RESERVATION OUT with the Register service action. When the reservation has been removed, no information is reported for the initiator in the Read Keys service action of the resulting PERSISTENT RESERVATION IN command. A PERSISTENT RESERVATION OUT command with Reserve service action creates a persistent reservation with a specified Scope and Type. Persistent reservations are not superseded by a new persistent reservation from any initiator except by the execution of a PERSISTENT RESERVATION OUT command that specifies either a Pre-empt or Pre-empt and Clear service action. New persistent reservations that do not conflict with an existing persistent reservation execute normally. Persistent reservations of logical units or extents having the same Type value are permitted if no conflicting persistent reservations are held by another initiator. When these types of overlapping reservations are released, each of the extent reservations and logical unit reservations are removed with a separate Release service action. Multiple identical reservations from the same initiator are all released simultaneously via a single Release service action that matches the reservations.

Skip Mask Length

REASSIGN BLOCKS Command (07h)
The REASSIGN BLOCKS Command reassigns one or more logical blocks to a spare location(s) set aside for this purpose and records the reassignments in the Grown Defect List. No command-specific Control Descriptor Block fields affect the processing performed for this command. The list of Logical Block Addresses to reassign is supplied in the REASSIGN BLOCKS Defect List passed in the command's data-out buffer transfer. The Defect List consists of a Defect List Header (Table 5-146) followed by zero or more Defect Descriptors (Table 5-148). Table 5-147 contains the field description for the Defect List Header. If multiple Logical Block Addresses are specified, they are listed in ascending order. The REASSIGN BLOCKS Command Descriptor Block is shown in Table 5-145.
Table 5-145 REASSIGN BLOCKS Command Descriptor Block-Data Format Bit Byte

0 1-4 5

Operation Code (07h) Reserved Control
Table 5-146 REASSIGN BLOCKS Defect List Header-Data Format Bit Byte

0-1 2-3

Table 5-147 REASSIGN BLOCKS Defect List Header -Field Description Field
The contents of this field indirectly indicates the number of defective Logical Block Addresses passed in the Defect List. The value must be four times the number of Defect Descriptors supplied. Zero (0) is a valid value and makes the command a NOP.
Table 5-148 REASSIGN BLOCKS Defect Descriptor-Data Format Bit Byte
Defect Logical Block Address
The drive returns a CHECK CONDITION status if it cannot complete the command. If this happens because the drive ran out of spare sectors, the drive sets the sense key to MEDIUM ERROR and the additional sense code to NO DEFECT SPARE LOCATION AVAILABLE. The sense key and additional sense key are set appropriately for any other cause. The Logical Block Address of the first defect descriptor not reassigned is returned in the command-specific information field of the sense data. If all defects were reassigned, but some other error occurred, then FFFFFFFFh is posted in the command-specific field.
RECEIVE DIAGNOSTIC RESULTS Command (1Ch)
The RECEIVE DIAGNOSTIC RESULTS command fetches the results of the last SEND DIAGNOSTIC command sent to the drive. The drives support the diagnostic pages listed in Table 5-149.

0 1-10 11

Operation Code (A0h) Reserved (MSB) Allocation Length (LSB) Reserved Control
Table 5-164 REPORT LUNS Command Descriptor Block-Field Description Field
If the Allocation Length is not sufficient to contain the logical unit number values for all configured logical units, the device server still reports as many logical number values as will fit in the Allocation Length. The format of the report of configured logical units is shown in Table 5-165.
Table 5-165 LUN Reporting Parameter List -Data Format Bit Byte
0-3 4-- 15 n-7 n (MSB) LUN (first LUN) (LSB) LUN (last LUN, if more than one)
(MSB) LUN List Length (n 7) (LSB) Reserved
The LUN List Length field contains the length in bytes of the LUN list that can be transferred. The LUN list length equals the number of logical unit numbers reported multiplied by eight. If the allocation length in the CDB is too small to allow transfer of information about all of the logical units configured, the LUN list length value is not adjusted to reflect the truncation. The data format for the LUN Reporting Parameter List is described in Table 5-165.
REQUEST SENSE Command (03h)
The REQUEST SENSE command causes the drive to transfer detailed sense data to the initiator. The drive maintains sense data on an individual initiator basis. The data format and field descriptions are shown in Table 5-166 and Table 5-167 respectively.
Table 5-166 REQUEST SENSE Command Descriptor Block-Data Format Bit Byte
Operation Code (03h) Reserved Allocation Length Control
Table 5-167 REQUEST SENSE Command Descriptor Block-Field Description Field
The drives can return a maximum of 18 bytes of sense data. The initiator should set this field to the maximum value to receive all the sense data.

5.32.1

Sense Data Availability
Sense data is available if the initiator is responsible for an in-progress FORMAT UNIT command (Format Progress Indication is available in the Format Progress Indication Bytes). The drive returns a sense key of NO SENSE and an additional sense code of NO ADDITIONAL SENSE INFORMATION if it has no sense data available for the initiator. For Parallel SCSI disk drives, sense data is available if the previous command from the initiator terminated with a CHECK CONDITION status, or if the previous command from the initiator ended with an unexpected BUS FREE error.REQUEST Clearing Sense Data Sense data is cleared from the drive's memory by one of the following conditions: After being returned from a REQUEST SENSE command. As soon as any other command is received from the initiator except INQUIRY.

5.32.2

SelfTest

If the PF bit is 0 and the SelfTest bit is 1, the drive runs its self-test routine. GOOD status is returned to the initiator if the self-test passes. CHECK CONDITION status is returned along with any available sense data if the self-test fails. This field is ignored if the PF bit is 1.
DevOnl UnitOfl Parameter List Length
Any value in this field is ignored by the drive. Any value in this field is ignored by the drive. When the Parameter List Length field is non-zero, the value specifies the length (in bytes) of the parameter list that is to be transferred from the initiator to the drive during the command's dataout buffer transfer. This field is only valid when the PF bit = 1 and must be 0 when the PF bit is 0.

5.38.1

Supported Diagnostic Page List
Table 5-193 shows the Supported Diagnostic Page List that can be supplied with a SEND DIAGNOSTIC command to request that the Supported Diagnostic Page List be returned after the next RECEIVE DIAGNOSTIC RESULTS command.
Table 5-193 Supported Diagnostic Page List-Data Format Bit Byte
Page Code (00h) Reserved Page Length (00h)

5.38.2

Translate Address Page
Table 5-194 shows the Translate Address Page that can be supplied with a SEND DIAGNOSTIC command to request that the Translate Address Page be returned after the next RECEIVE DIAGNOSTIC RESULTS command.
Table 5-194 Translate Address Page-Data Format Bit Byte
2-5 6-- 13 Reserved Reserved Address to Translate (LBA Format) Reserved
Page Code (40h) Reserved Page Length (0Ah) Supplied Format 000b (LBA) Translate Format 101b (Physical Sector)
SET DEVICE IDENTIFIER Command (A4h)
The SET DEVICE IDENTIFIER command requests that the device identifier information in the logical unit be set to the value send via the SET DEVICE IDENTIFIER commands parameter list. The data format and field descriptions are shown in Table 5-195 and Table 5-196 respectively. Upon successful completion of a SET DEVICE IDENTIFIER command, a Unit Attention is generated for all initiators except the one that issued the command.
Table 5-195 SET DEVICE IDENTIFIER Command Descriptor Block-Data Format Bit Byte
Reserved Reserved (MSB) Parameter List Length (LSB) Reserved Control
Operation Code (A4h) Service Action (06h)
Table 5-196 SET DEVICE IDENTIFIER Command Descriptor Block-Field Descriptions Field
Must = 06h. Any other value forces Check Condition, Illegal Request. This field specifies the length, in bytes, of the Identifier to be transferred from the application client to the device server. The maximum value for this field is 64 bytes. A parameter list length of 0 indicates that no data will be transferred, and that subsequent REPORT DEVICE IDENTIFIER commands will return an Identifier length of 0. If the parameter list length exceeds 64 bytes, then the drive returns a Check Condition status with the sense key set to Illegal Request, and an additional sense code of Invalid Field in CDB.

DATA INTEGRITY AND SECURITY
The disk drives use a combination of parity checking, error detection coding (EDC), error correction coding (ECC), and checkpointing to protect stored data from media errors, transfer or addressing errors, or errors introduced during block reallocation.

Media Error Protection

To ensure that data read is the same as data written, the drive computes and appends an Error Correction Code (ECC) to each block of data stored. The drive uses a 352bit Reed Solomon code with a 4:1 interleave, which can correct up to 20 bytes in each block. The drive can also correct up to 2 bytes per interleave (up to 8 per block) in hardware (on-the-fly), with no loss in throughput.
Transfer Error Protection
An end-to-end error detection code (EDC) protects data from any errors introduced by internal buses, the disk controller chip, the data cache, or the SCSI interface. An EDC is calculated and added to each data block as the data arrives from the SCSI bus (after SCSI bus parity is checked). The EDC is stored with the data and protected by the block ECC for added security. On reading or writing, the EDC is checked as the data is transferred between buffer RAM and the media or the SCSI bus.
Addressing Error Protection
Each data block on the media is identified and located by a servo spoke address. The spoke address consists of a two-byte word. Each spoke has multiple copies of the least significant bytes of the address. The disk hardware requires that a majority of the copies agree and that the result agrees with the expected head, track, and spoke number, before it will read or write the data. To further protect against addressing errors, the logical address (LBA) of the data is added to the EDC of each block. If data is written to the wrong block and subsequently read, or read from the wrong block, the error will be flagged. The hardware does not allow a blind read of a data block; the firmware must request specific data blocks. Even if the head selection hardware malfunctions, it is not possible for the drive to return data from the wrong head.
Data Sector Reallocation Error Protection
In any SCSI disk drive, bad blocks may be reallocated. However, a power failure or unrecoverable data could threaten data integrity during a block reallocation. The reallocation and defect list storage algorithms prevent a reallocation from being lost due to a power failure. Once a reallocation starts, the information about the data block to be moved is stored on the media. As the reallocation progresses, checkpoint
information about the reallocation (such as the new destination and the data in transit) is recorded. If the reallocation is interrupted, the process can automatically continue from the last checkpoint without data loss. If a block reallocation is performed on unrecoverable data (rare), a Bad Data Mark is set in the new location. Thus, there is no chance of undetected bad data being generated from a reallocation.