Tandberg Data SLR7-140 PROGRAMMER, SLR7, SLR50, SLR60, SLR75 User Manual

...

Download

TANDBERG SLR

Tandberg SLR7, SLR50,

SLR60, SLR75, SLR100, SLR140

SCSI INTERFACE

FUNCTIONAL SPECIFICATION

TANDBERG DATA ASA

P.O. Box 134 Kjelsås N-0411 OSLO, NORWAY Phone + 47 22 18 90 90 Telefax + 47 22 18 95 50

Part No. 43 07 43 – 13 October 2003

Related publications available from Tandberg Data ASA:

Part No. Title

43 04 44 - 11 Tandberg SLR7, SLR50, SLR60, SLR75, SLR100,

SLR140 Reference Manual

This publication may describe designs for which patents are granted or pending. By publishing this information, Tandberg Data ASA conveys no license under any patent or any other rights.

Every effort has been made to avoid errors in text and diagrams. However, Tandberg Data ASA assumes no responsibility for any errors, which may appear in this publication.

It is the policy of Tandberg Data ASA to improve products as new techniques and components become available. Tandberg Data ASA therefore reserves the right to change specifications at any time.

We would appreciate any comments on this publication.

i Tandberg SLR Product Line SCSI Functional Specifications

Table Of Contents

1. Introduction .....................................................................1

1.1. General .............................................................................................. 1

1.2. Overview............................................................................................ 2

1.3. Glossary............................................................................................. 3

1.4. Additional Reference Documentation................................................ 5

2. About Tape Streamers ....................................................1

2.1. Physical Elements ............................................................................. 1

2.2. Data Storage Characteristics............................................................. 1

2.3. Partitions Within a Volume ................................................................ 3

2.3.1. Partitioning a Volume................................................................. 4

2.3.2. Selecting a Partition ................................................................... 6

2.3.3. Using Initiator Defined Partitions ............................................... 8

2.3.4. Quick File Access .................................................................... 11

2.4. Logical Elements Within a Partition................................................. 14

2.5. Overwrite ......................................................................................... 15

2.6. Using Fixed and Variable Length Blocks......................................... 17

2.6.1. Variable and Fixed Length Blocks ........................................... 17

2.6.2. Writing...................................................................................... 17

2.6.3. Reading.................................................................................... 18

2.6.4. Illegal Length Conditions when Reading ................................. 18

2.7. Data Buffering.................................................................................. 19

2.7.1. Introduction .............................................................................. 19

2.7.2. Data Formatting ....................................................................... 21

2.7.3. Buffered Mode ......................................................................... 23

2.7.4. Read-Ahead............................................................................. 25

2.7.5. Underrun/Overrun .................................................................... 25

2.7.6. Buffer Thresholds..................................................................... 26

2.7.7. Disconnect/Re-connect............................................................ 27

2.7.8. Data Re-transfer ...................................................................... 28

2.7.9. Buffer Parity Errors .................................................................. 28

2.8. Data Compression........................................................................... 29

2.8.1. Background.............................................................................. 29

2.8.2. Controlling Data Compression................................................. 30

2.9. Optimizing Streaming Operation ..................................................... 33

2.9.1. Forced Streaming .................................................................... 33

2.9.2. AVC (Automatic Velocity Control)............................................ 33

2.9.3. Using both Forced Streaming and AVC................................... 34

2.9.4. Recommended Settings .......................................................... 35

2.10. Recorded Objects............................................................................ 35

2.11. TapeAlert ......................................................................................... 36

3. Logical Characteristics...................................................1

3.1. SCSI-bus Phases .............................................................................. 1

3.1.1. Bus Management Phases.......................................................... 1

3.1.2. Information Transfer Phases ..................................................... 2

3.2. SCSI-bus Phase Sequences ............................................................. 3

3.2.1. Legal Phase Sequences............................................................ 4

3.3. SCSI Pointers .................................................................................... 9

3.4. Unit Attention ................................................................................... 11

3.5. SCSI-bus Conditions ....................................................................... 12

3.5.1. Attention (ATN) ........................................................................ 12

3.5.2. Reset........................................................................................ 13

Tandberg Data Table of Contents

ii Tandberg SLR Product Line SCSI Functional Specifications

4. Commands.......................................................................1

4.1. The Command Descriptor Block........................................................ 1

4.2. Command Control Byte ..................................................................... 2

4.3. Reserved Fields................................................................................. 3

4.4. Command Set Summary ................................................................... 4

4.5. Command Sequencing...................................................................... 6

4.5.1. Normal Modes............................................................................ 6

4.6. Multiple Connections ......................................................................... 9

4.6.1. Background................................................................................ 9

4.6.2. Commands Received Serially, LUN Is 0.................................. 10

4.6.3. Commands Received Serially, LUN Not 0............................... 11

4.6.4. Concurrent Command, Same Initiator, LUN Is 0 ..................... 11

4.6.5. Concurrent Command, Any Initiator, LUN Not 0 ..................... 11

4.6.6. Concurrent Command, Different Initiator, LUN Is 0................. 11

5. Status Bytes.....................................................................1

6. Message System .............................................................1

6.1. Message In ........................................................................................ 2

6.2. Message Out ..................................................................................... 3

6.3. Extended Message............................................................................ 5

6.3.1. Wide Data Transfer Request ..................................................... 6

6.3.2. Synchronous Data Transfer Request Message......................... 8

6.4. Message Reject Message Handling................................................ 11

6.4.1. Message-In Phase ................................................................... 11

6.4.2. Message-Out Phase ................................................................ 11

6.5. Abort Message Handling ................................................................. 12

6.6. Unexpected Bus Free...................................................................... 14

7. General Exception Handling...........................................1

7.1. Error Codes ....................................................................................... 1

7.2. Error Conditions for All Commands................................................... 5

7.3. Deferred Errors.................................................................................. 6

7.4. Error Conditions for Media Access Commands ................................ 7

7.5. Power On Selftest (POST) Error Handling ........................................ 8

7.5.1. Commands Executed After POST Error .................................... 8

7.5.2. Commands Terminated After POST Error................................. 9

7.6. Bus Parity Error Handling .................................................................. 9

7.6.1. Errors Detected by the Drive ................................................... 10

7.6.2. Errors Detected by the Initiator ................................................ 11

7.7. Buffer Parity Error Handling............................................................. 11

7.8. Error Priority..................................................................................... 12

7.9. Suggested Error Recovery Action ................................................... 13

8. Erase ................................................................................1

8.1. Command Description....................................................................... 1

8.2. Command Descriptor Block............................................................... 1

8.3. Exception Handling............................................................................ 2

8.4. Phase Sequencing ............................................................................ 2

9. Inquiry ..............................................................................1

9.1. Command Description....................................................................... 1

9.2. Command Descriptor Block............................................................... 1

9.3. Parameter Lists ................................................................................. 3

9.3.1. Standard Inquiry Data................................................................ 3

Tandberg Data Table of Contents

iii Tandberg SLR Product Line SCSI Functional Specifications

9.3.2. Vital Product Data ...................................................................... 6

9.4. Exception Handling.......................................................................... 16

9.5. Phase Sequencing .......................................................................... 16

10. Load/Unload.....................................................................1

10.1. Command Description ....................................................................... 1

10.2. Command Descriptor Block............................................................... 2

10.3. Exception Handling............................................................................ 3

10.4. Phase Sequencing ............................................................................ 3

11. Locate...............................................................................1

11.1. Command Description ....................................................................... 1

11.2. Command Descriptor Block............................................................... 2

11.3. Exception Handling............................................................................ 3

11.4. Phase Sequencing ............................................................................ 3

12. Log Select ........................................................................1

12.1. Command Description ....................................................................... 1

12.2. Command Descriptor Block............................................................... 1

12.3. Parameter List ................................................................................... 3

12.3.1. General Parameter Description ................................................. 3

12.3.2. Log Page Headers ..................................................................... 3

12.3.3. Log Parameter Headers ............................................................ 3

12.3.4. Modifiable Parameter Values..................................................... 6

12.3.5. The TapeAlert Information Page................................................ 6

12.4. Exception Handling............................................................................ 7

12.5. Phase Sequencing ............................................................................ 7

13. Log Sense ........................................................................1

13.1. Command Description ....................................................................... 1

13.2. Command Descriptor Block............................................................... 1

13.3. Parameter List ................................................................................... 3

13.3.1. Supported Log Pages ................................................................ 3

13.3.2. Buffer Overrun/Underrun Counters Page .................................. 4

13.3.3. Write Error Counter Page .......................................................... 7

13.3.4. Read Error Counter Page ........................................................ 14

13.3.5. TapeAlert Page ........................................................................ 22

13.3.6. Data Block Counters Page ...................................................... 25

13.3.7. Remaining Capacity Page ....................................................... 29

13.3.8. Tape Mark Counters Page....................................................... 36

13.3.9. Head Cleaning Page................................................................ 39

13.3.10. Drive Page ............................................................................... 43

13.3.11. Servo Page ............................................................................. 46

13.3.12. Track Number Page................................................................. 53

13.3.13. Cartridge Usage Page ............................................................. 54

13.3.14. Compression Ratio Page ......................................................... 62

13.4. Exception Handling.......................................................................... 70

13.5. Phase Sequencing .......................................................................... 70

14. Mode Select .....................................................................1

14.1. Command Description ....................................................................... 1

14.2. Command Descriptor Block............................................................... 1

14.3. Parameter List ................................................................................... 2

14.3.1. Header List................................................................................. 3

14.3.2. Block Descriptor List .................................................................. 6

14.3.3. Read-Write Error Recovery Page ............................................ 11

Tandberg Data Table of Contents

iv Tandberg SLR Product Line SCSI Functional Specifications

14.3.4. Disconnect/Reconnect Page ................................................... 13

14.3.5. Control Mode Page .................................................................. 16

14.3.6. Data Compression Page.......................................................... 18

14.3.7. Device Configuration Page ...................................................... 20

14.3.8. Medium Partition Page (1) ....................................................... 23

14.3.9. Informational Exceptions Control Page ................................... 27

14.3.10. Miscellaneous Parameters Page ............................................. 29

14.3.11. User Page 0 ............................................................................. 35

14.3.12. User Page 1 ............................................................................. 36

14.3.13. Cartridge Manufacturer Page .................................................. 37

14.4. Exception Handling.......................................................................... 38

14.5. Phase Sequencing .......................................................................... 38

15. Mode Sense .....................................................................1

15.1. Command Description ....................................................................... 1

15.2. Command Descriptor Block............................................................... 2

15.3. Parameter List ................................................................................... 3

15.3.1. Header List................................................................................. 3

15.3.2. Block Descriptor List .................................................................. 5

15.3.3. Read-Write Error Recovery Page Descriptor ............................ 6

15.3.4. Disconnect/Reconnect Page Descriptor .................................... 6

15.3.5. Control Mode Page .................................................................... 7

15.3.6. Data Compression Page Descriptor .......................................... 7

15.3.7. Device Configuration Parameters Page Descriptor................... 8

15.3.8. Medium Partition Parameters Page Descriptor ......................... 8

15.3.9. TapeAlert Configuration Page ................................................... 9

15.3.10. Miscellaneous Parameters Page Descriptor ........................... 10

15.3.11. User Page 0 Page Descriptor .................................................. 10

15.3.12. User Page 1 Page Descriptor .................................................. 11

15.3.13. Cartridge Manufacturer Page .................................................. 12

15.4. Exception Handling.......................................................................... 13

15.5. Phase Sequencing .......................................................................... 13

16. Prevent/Allow Medium Removal ....................................1

16.1. Command Description ....................................................................... 1

16.2. Command Descriptor Block............................................................... 1

16.3. Exception Handling............................................................................ 2

16.4. Phase Sequencing ............................................................................ 2

17. Read .................................................................................1

17.1. Command Description ....................................................................... 1

17.2. Command Descriptor Block............................................................... 2

17.3. Exception Handling............................................................................ 3

17.3.1. General ...................................................................................... 3

17.3.2. No Data...................................................................................... 3

17.3.3. Tapemark Detected ................................................................... 3

17.3.4. Illegal Length.............................................................................. 4

17.3.5. End of Data ................................................................................ 6

17.3.6. End of Partition .......................................................................... 7

17.3.7. Non-Recoverable Read Error .................................................... 7

17.3.8. Illegal Termination...................................................................... 8

17.3.9. Read After Write ........................................................................ 8

17.4. Phase Sequencing ............................................................................ 9

18. Read Block Limits ...........................................................1

18.1. Command Description ....................................................................... 1

Tandberg Data Table of Contents

v Tandberg SLR Product Line SCSI Functional Specifications

18.2. Command Descriptor Block............................................................... 1

18.3. Parameter List ................................................................................... 1

18.4. Exception Handling............................................................................ 2

18.5. Phase Sequencing ............................................................................ 2

19. Read Buffer ......................................................................1

19.1. Command Description ....................................................................... 1

19.2. Command Descriptor Block............................................................... 1

19.3. Read Data Mode (2) .......................................................................... 3

19.3.1. Data Buffer (Buffer ID = 0) ......................................................... 3

19.3.2. Static RAM (Buffer ID = 1) ......................................................... 3

19.3.3. Media Statistics (Buffer ID = 2) .................................................. 3

19.3.4. Microcode Store (Buffer ID = 3)................................................. 3

19.3.5. EEPROM (Buffer ID = 4)............................................................ 3

19.3.6. External RAM Adapter (Buffer ID = 5) ....................................... 3

19.3.7. NVP DATA (Buffer ID = 6) ......................................................... 4

19.3.8. Tape Buffer Control Data (Buffer ID = 7) ................................... 4

19.3.9. Volume Directory (Buffer ID = 8)................................................ 4

19.3.10. Dbase Variables (Buffer ID = 9)................................................ 4

19.3.11. Header File Dates (Buffer ID = 10 )................................................. 4

19.3.12. Trace Buffer Control Blocks (Buffer ID = 11) .................................. 4

19.3.13. Complete Trace Buffer (Buffer ID = 12) .......................................... 4

19.4. Read Descriptor Mode (3) ................................................................. 5

19.5. Exception Handling............................................................................ 6

19.6. Phase Sequencing ............................................................................ 7

20. Read Position ..................................................................1

20.1. Command Description ....................................................................... 1

20.2. Command Descriptor Block............................................................... 2

20.3. Data Format....................................................................................... 3

20.4. Exception Handling............................................................................ 4

20.5. Phase Sequencing ............................................................................ 5

21. Receive Diagnostic Results............................................1

21.1. Command Description ....................................................................... 1

21.2. Command Descriptor Block............................................................... 1

21.3. Results From the SEND DIAGNOSTICS Command......................... 3

21.3.1. The Header Page....................................................................... 3

21.3.2. The Diagnostic Pages................................................................ 4

21.3.3. Supported Page Codes ............................................................. 5

21.4. Results From the Stand Alone Diagnostics Test............................... 6

21.4.1. The Stand Alone Diagnostic Result Page.................................. 6

21.5. Exception Handling............................................................................ 7

21.6. Phase Sequencing ............................................................................ 7

22. Release.............................................................................1

22.1. Command Description ....................................................................... 1

22.2. Command Descriptor Block............................................................... 1

22.3. Exception Handling............................................................................ 2

22.4. Phase Sequencing ............................................................................ 2

23. Request Sense.................................................................1

23.1. Command Description ....................................................................... 1

23.2. Command Descriptor Block............................................................... 1

23.3. Parameter List ................................................................................... 2

23.4. Sense Keys........................................................................................ 5

Tandberg Data Table of Contents

vi Tandberg SLR Product Line SCSI Functional Specifications

23.5. Additional Sense Code and Qualifier ................................................ 6

23.6. Exception Handling............................................................................ 8

23.7. Phase Sequencing ............................................................................ 8

24. Reserve ............................................................................1

24.1. Command Description ....................................................................... 1

24.2. Command Descriptor Block............................................................... 2

24.3. Exception Handling............................................................................ 2

24.4. Phase Sequencing ............................................................................ 2

25. Rewind..............................................................................1

25.1. Command Description ....................................................................... 1

25.2. Command Descriptor Block............................................................... 1

25.3. Exception Handling............................................................................ 1

25.4. Phase Sequencing ............................................................................ 2

26. Send Diagnostics ............................................................1

26.1. Command Description ....................................................................... 1

26.2. Command Descriptor Block............................................................... 1

26.3. Predefined Selftest Sequence 1........................................................ 3

26.4. Predefined Selftest Sequence 2........................................................ 4

26.5. Exception Handling............................................................................ 5

26.6. Phase Sequencing ............................................................................ 5

27. Space................................................................................1

27.1. Command Description ....................................................................... 1

27.2. Command Descriptor Block............................................................... 2

27.3. Using Fast Space .............................................................................. 3

27.4. Exception Handling............................................................................ 3

27.4.1. General ...................................................................................... 3

27.4.2. No Data...................................................................................... 4

27.4.3. Filemark Detected...................................................................... 4

27.4.4. Setmark Detected ...................................................................... 4

27.4.5. End of Data ............................................................................... 5

27.4.6. Beginning of Partition................................................................ 5

27.4.7. End of Partition ......................................................................... 5

27.4.8. Non-Recoverable Read Error During Space Forward ............... 6

27.4.9. Error Condition or Bad Block During Space Reverse................ 6

28.4.10. Space Forward After Write ........................................................ 7

27.5. Phase Sequencing ............................................................................ 7

28. Test Unit Ready ...............................................................1

28.1. Command Description ....................................................................... 1

28.2. Command Descriptor Block............................................................... 1

28.3. Exception Handling............................................................................ 1

28.4. Phase Sequencing ............................................................................ 2

29. Verify (not for SLR7 and SLR140) ..................................1

29.1. Command Description ....................................................................... 1

29.2. Command Descriptor Block............................................................... 2

29.3. Exception Handling............................................................................ 3

29.4. Phase Sequencing ............................................................................ 3

30. Write .................................................................................1

Tandberg Data Table of Contents

vii Tandberg SLR Product Line SCSI Functional Specifications

30.1. Command Description ....................................................................... 1

30.2. Command Descriptor Block............................................................... 2

30.3. Data Compression............................................................................. 2

30.4. Write from BOP.................................................................................. 2

30.5. Write from EOD, Append................................................................... 3

30.6. Writing Over Existing Data, Overwrite............................................... 3

30.7. Terminating Write Operations............................................................ 3

30.8. Exception Handling............................................................................ 4

30.8.1. General ...................................................................................... 4

30.8.2. Unsupported Block Length ........................................................ 4

30.8.3. Illegal Media Type...................................................................... 5

30.8.4 Illegal Overwrite ......................................................................... 6

30.8.5. Illegal Append Tape Format ...................................................... 6

30.8.6. Logical Early Warning................................................................ 6

30.8.7. End of Partition .......................................................................... 7

30.8.8. Non-Recoverable Write Error .................................................... 7

30.8.9. Append Error.............................................................................. 8

30.9. Phase Sequencing ............................................................................ 8

31. Write Buffer......................................................................1

31.1. Command Description ....................................................................... 1

31.2. Command Descriptor Block............................................................... 1

31.3. Write Combined Header and Data Mode (0)..................................... 2

31.4. Write Data Mode (2) .......................................................................... 3

31.5. Download Microcode Mode (4).......................................................... 3

31.6. Download Microcode and Save Mode (5) ......................................... 3

31.7. Download Microcode with Offsets and Save Mode (7) ..................... 4

31.8. Microcode Verification and Save ....................................................... 4

31.9. Microcode Data Format ..................................................................... 5

31.10. Exception Handling............................................................................ 6

31.11. Phase Sequencing ............................................................................ 7

32. Write Filemarks................................................................1

32.1. Command Description ....................................................................... 1

32.2. Command Descriptor Block............................................................... 1

32.3. Terminating Write Operations............................................................ 2

32.4. Write Filemarks from BOM ................................................................ 2

32.5. Exception Handling............................................................................ 2

32.5.1. General ...................................................................................... 2

32.5.2. Illegal Media Type...................................................................... 3

32.5.3. Illegal Append Tape Format ...................................................... 3

32.5.4. Logical Early Warning................................................................ 3

32.5.5. End of Partition .......................................................................... 4

32.5.6. Non-Recoverable Write Error .................................................... 4

32.5.7. Append Error.............................................................................. 5

32.6. Phase Sequencing ............................................................................ 5

Tandberg Data Table of Contents

viii Tandberg SLR Product Line SCSI Functional Specifications

Table Of Figures

Figure 2-1: Serpentine Recording ................................................................. 2

Figure 2-2: Logical Track Areas .................................................................... 2

Figure 2-3: Multiple Partition Track Layout (SLR32 example) ....................... 9

Figure 2-4: Overwrite After the 1st Logical Data Block ............................... 16

Figure 2-5: Overwrite of Last Filemark ........................................................ 16

Figure 2-6: The SLR50 Buffer System (Write Mode) .................................. 20

Figure 2-7: The SLR140, SLR100, SLR75, SLR60 and SLR7 Buffer System

(Write Mode) ......................................................................................... 21

Figure 2-8: Host Buffer Data flow During Write Operations ........................ 26

Figure 2-9: Host Buffer Data flow During Read Operations........................ 26

Figure 2-10: Packing Data in Compression Block Groups.......................... 29

Figure 3-1: Phase Sequencing...................................................................... 3

Figure 3-2: SCSI Pointers ........................................................................... 10

Figure 4-1: Connections, Some Examples.................................................. 10

Tandberg Data Table of Contents

ix Tandberg SLR Product Line SCSI Functional Specifications

Table Of Tables

Table 1-1: Capacity and medium types.......................................................... 1

Table 2-1: The Medium Partition Page(1) ..................................................... 4

Table 2-2: Minimum Partition Size ................................................................ 8

Table 2-3: Partitions Within A QFA Volume ................................................ 11

Table 2-4: Directory Track set Allocation .................................................... 11

Table 2-5: Data Compression Page Descriptor ........................................... 30

Table 2-6: Forced Streaming and AVC On Different Medium Types (to be

continued…).......................................................................................... 34

Table 3-1: SCSI-bus Phases......................................................................... 1

Table 4-1: Typical Six-byte Command Descriptor Block............................... 1

Table 4-2: Typical Ten-byte Command Descriptor Block .............................. 1

Table 4-3: The Command Control Byte ......................................................... 2

Table 4-4: SCSI Command Set...................................................................... 4

Table 4-5: Normal Mode Actions (to be continued...)..................................... 7

Table 4-6: Connections, LUN and Initiator ID Combinations ...................... 10

Table 5-1: The Status Set ............................................................................. 1

Table 6-1: The Message-In Set..................................................................... 2

Table 6-2: The Message-Out Set .................................................................. 3

Table 6-3: The IDENTIFY Message .............................................................. 4

Table 6-4: Extended Message Format.......................................................... 5

Table 6-5: Supported Extended Message Codes ......................................... 5

Table 6-6: Wide Data Transfer Request ....................................................... 8

Table 6-7: Synchronous Data Transfer Request........................................... 8

Table 6-8: Response to MESSAGE REJECT ............................................. 11

Table 6-9: ABORT TASK SET Message Handling ...................................... 13

Table 7-1: Error Codes (to be continued…) .................................................. 2

Table 7-2: Error Priority ............................................................................... 12

Table 8-1: ERASE Command Descriptor Block ............................................ 1

Table 9-1: INQUIRY Command Descriptor Block ......................................... 1

Table 9-2: INQUIRY Parameter List (to be continued...)............................... 3

Table 9-3: Summary of Supported VPD Pages ............................................ 6

Table 9-4: Unit Serial Number Page ............................................................. 7

Table 9-5: Implemented Operating Definitions Page .................................... 8

Table 9-6: ASCII Implemented Operating Definition Page ........................... 9

Table 9-7: Hardware Revision Levels Page ................................................ 10

Table 9-8: PROM Microcode Revision Level .............................................. 11

Table 9-9: Drive Manufacturing Date Page................................................. 13

Table 9-10: PROM Microcode Creation Date Page .................................... 14

Table 9-11: Drive Adjustment Date Page.................................................... 15

Table 10-1: LOAD/UNLOAD Command Descriptor Block ............................ 2

Table 10-2: LOAD/UNLOAD Operations....................................................... 2

Table 11-1: LOCATE Command Descriptor Block........................................ 2

Table 12-1: LOG SELECT Command Descriptor Block................................ 1

Table 12-2: Parameter Control Byte............................................................... 3

Table 13-1: LOG SENSE Command Descriptor Block ................................. 1

Table 13-2: Supported Log Pages Page....................................................... 3

Table 13-3: Buffer Overrun/Underrun Counters Page .................................. 4

Table 13-4: Underrun Log Parameters ......................................................... 4

Table 13-5: Overrun Log Parameters............................................................. 6

Table 13-6: Write Error Counter Page ........................................................... 7

Table 13-7: Rewrite Counter Log Parameters ............................................... 7

Table 13-8: Total Write Errors Log Parameters ............................................. 9

Table 13-9: Total Write Errors Corrected Log Parameters .......................... 10

Table 13-10: Total Times Errors Processed Log Parameters...................... 11

Table 13-11: Total Bytes Written Parameters .............................................. 12

Table 13-12: Total Uncorrected Write Errors Log Parameters .................... 13

Tandberg Data Table of Contents

x Tandberg SLR Product Line SCSI Functional Specifications

Table 13-13: Read Error Counter Page ....................................................... 14

Table 13-14: Reread Counter Log Parameters........................................... 15

Table 13-15: Total Read Error Parameters................................................. 16

Table 13-16: Total Read Error Corrected Log Parameters......................... 17

Table 13-17: ECC Correction Counter Log Parameters ............................. 18

Table 13-18: Total Bytes Read Counter Log Parameters........................... 19

Table 13-19: Total Uncorrected Read Errors Log Parameters .................... 20

Table 13-20: ECC Error Counter Log Parameters ...................................... 21

Table 13-21: TapeAlert Page ....................................................................... 22

Table 13-22: TapeAlert Information Log Parameters................................... 22

Table 13-23: Supported TapeAlert Flags (to be continued…) ..................... 23

Table 13-24: Data Block Counters Page..................................................... 25

Table 13-25: Logical Block Counter Log Parameters ................................. 26

Table 13-26: Write Media Blocks Counter Parameters............................... 27

Table 13-27: Read Media Block Counter Log Parameters ......................... 28

Table 13-28: Remaining Capacity Page ...................................................... 29

Table 13-29: Remaining Capacity Log Parameter code 01 ......................... 30

Table 13-30: Remaining Capacity Log Parameter code 02 ......................... 32

Table 13-31: Maximum Capacity Log Parameter code 03........................... 34

Table 13-32: Maximum Capacity Log Parameter code 04........................... 35

Table 13-33: Tape Mark Counters Page...................................................... 36

Table 13-34: Filemark Counter Log Parameters......................................... 36

Table 13-35: Setmark Counter Log Parameters ......................................... 38

Table 13-36: Head Cleaning Page ............................................................... 39

Table 13-37: Clean Head Log Parameters ................................................. 39

Table 13-38: Head Cleaning Time Log Parameters.................................... 40

Table 13-39: Cleaning Count Log Parameters............................................ 42

Table 13-40: Drive Page ............................................................................. 43

Table 13-41: Total Power On Time Parameters ......................................... 44

Table 13-42: Cartridge Load Counter Parameters...................................... 45

Table 13-43: Servo Page ............................................................................ 46

Table 13-44: Servo Lock Retry Log Parameters........................................ 47

Table 13-45: Servo Track Seek Log Parameters....................................... 48

Table 13-46: Write Servo Lock Lost Counter Log Parameters .................. 49

Table 13-47: Write Servo Dropout Log Parameters.................................... 50

Table 13-48: Read Servo Lock Lost Counter Log Parameters .................. 51

Table 13-49: Read Servo Dropout Log Parameters.................................... 52

Table 13-50: Track Number Page ............................................................... 53

Table 13-51: Cartridge Usage Page............................................................ 54

Table 13-52: Cartridge Serial Number Parameters..................................... 55

Table 13-53: Cartridge Load Counter Parameters...................................... 56

Table 13-54: Cartridge BOT Pass Counter Log Parameters ....................... 58

Table 13-55: Cartridge EOT Pass Counter Log Parameters ....................... 59

Table 13-56: Cartridge Write Pass Counter Log Parameters ...................... 60

Table 13-57: Cartridge Motion Time Log Parameters.................................. 61

Table 13-58: Compression Ratio Page ....................................................... 62

Table 13-59 Write Compression Ratio Log Parameter ................................ 63

Table 13-60: Read Decompression Log Parameter..................................... 64

Table 13-61: Write Compression Interval 1 Log Parameter......................... 65

Table 13-62: Write Compression Interval 2 Log Parameter......................... 66

Table 13-63: Write Compression Interval 3 Log Parameter......................... 67

Table 13-64: Write Compression Interval 4 Log Parameter......................... 68

Table 13-65: Write Compression Interval 5 Log Parameter......................... 69

Table 14-1: MODE SELECT Command Descriptor Block ............................ 1

Table 14-2: MODE SELECT Header List...................................................... 3

Table 14-3: Available Tape Speeds for SLR140............................................ 3

Table 14-4: Available Tape Speeds for SLR100............................................ 4

Table 14-5: Available Tape Speeds for SLR75.............................................. 4

Table 14-6: Available Tape Speeds for SLR60.............................................. 4

Tandberg Data Table of Contents

xi Tandberg SLR Product Line SCSI Functional Specifications

Table 14-7: Available Tape Speeds for SLR50............................................. 5

Table 14-8: Available Tape Speeds for SLR7................................................ 5

Table 14-9: MODE SELECT Block Descriptor List ....................................... 6

Table 14-10: Density codes, formats and suitable medium .......................... 6

Table 14-11: Type of Media Related to Tape Format ................................... 8

Table 14-12: Fixed Block Lengths................................................................. 9

Table 14-13: Legal Block Lengths (bytes)................................................... 10

Table 14-14: Read-Write Error Recovery Page Descriptor......................... 11

Table 14-15: Disconnect/Reconnect Page Descriptor ................................ 13

Table 14-16: Control Mode Page Descriptor............................................... 16

Table 14-17: Data Compression Page Descriptor ...................................... 18

Table 14-18: Device Configuration Page Descriptor................................... 20

Table 14-19: Medium Partitions Page......................................................... 23

Table 14-20: Maximum Additional Partitions................................................ 24

Table 14-21: Using the FDP, SDP and IDP bits.......................................... 25

Table 14-22: Informational Exceptions Control Page.................................. 27

Table 14-23: Miscellaneous Page Descriptor ............................................. 29

Table 14-24: BSYI Usage............................................................................. 31

Table 14-25: BSYA Usage ........................................................................... 31

Table 14-26: User Page 0 Page Descriptor ................................................ 35

Table 14-27: User Page 1 Page Descriptor ................................................ 36

Table 14-28: Cartridge Manufacturer Data Parameters.............................. 37

Table 15-1: MODE SENSE Command Descriptor Block .............................. 2

Table 15-2: MODE SENSE Header List........................................................ 3

Table 15-3: MODE SENSE Block Descriptor List ......................................... 5

Table 16-1: PREVENT/ALLOW MEDIUM REMOVAL Cdb ........................... 1

Table 17-1: READ Command Descriptor Block ............................................ 2

Table 17-2: Illegal Length Summary .............................................................. 6

Table 18-1: READ BLOCK LIMITS Command Descriptor Block .................. 1

Table 18-2: READ BLOCK LIMITS Data....................................................... 1

Table 19-1: READ BUFFER Command Descriptor Block............................. 1

Table 19-2: Read Buffer Modes ..................................................................... 2

Table 19-3: Read Buffer ID’s.......................................................................... 2

Table 19-4: READ BUFFER Descriptor List.................................................. 5

Table 19-5: Buffer Capacity............................................................................ 6

Table 20-1: READ POSITION Command Descriptor Block.......................... 2

Table 20-2: READ POSITION Data .............................................................. 3

Table 21-1: RECEIVE DIAGNOSTIC RESULTS Command Block................ 1

Table 21-2: The Header Page........................................................................ 3

Table 21-3: The Diagnostic Pages ................................................................. 4

Table 21-4: Supported Page Codes.............................................................. 5

Table 21-5: The Stand Alone Diagnostic Page ............................................. 6

Table 22-1: RELEASE Command Descriptor Block..................................... 1

Table 23-1: REQUEST SENSE Command Descriptor Block ....................... 1

Table 23-2: REQUEST SENSE Parameter List ............................................ 2

Table 23-3: Sense Key Specific Information, Invalid Fields ........................... 3

Table 23-4: Sense Keys ................................................................................. 5

Table 23-5: Additional Sense Code and Qualifier (to be continued…) ......... 6

Table 24-1: RESERVE Command Descriptor Block..................................... 2

Table 25-1: REWIND Command Descriptor Block........................................ 1

Table 26-1: SEND DIAGNOSTIC Command Descriptor Block..................... 1

Table 26-2: SEND DIAGNOSTICS Functions............................................... 2

Table 27-1: SPACE Command Descriptor Block .......................................... 2

Table 28-1: TEST UNIT READY Command Descriptor Block ...................... 1

Table 28-2: TEST UNIT READY Response................................................... 1

Table 29-1: VERIFY Command Descriptor Block ........................................ 2

Table 30-1: WRITE Command Descriptor Block........................................... 2

Table 30-2: Legal Media/Drive Combinations ............................................... 5

Table 31-1: WRITE BUFFER Command Descriptor Block ........................... 1

Tandberg Data Table of Contents

xii Tandberg SLR Product Line SCSI Functional Specifications

Table 31-2: Supported Modes....................................................................... 2

Table 31-3: Microcode Data Format.............................................................. 5

Table 32-1: WRITE FILEMARKS Command Descriptor Block ..................... 1

1-1 Tandberg SLR Product Line SCSI Functional Specifications

Introduction

1.1. General

This manual covers the SCSI Specifications for the Tandberg SLR Product Line Tape Drives. Currently there are six available models:

 Tandberg SLR140  Tandberg SLR100  Tandberg SLR75  Tandberg SLR60  Tandberg SLR50  Tandberg SLR7

This manual replaces Part no. 430743–12, titled, “ Tandberg SLR7, SLR50, SLR60, SLR75, SLR100. SCSI

Interface Functional Specification.”

The main differences between the five models are capacity, data transfer rate and the medium types supported. The following table shows the different capacities and read/write compatibility for the various drive models.

Medium Type1Capacity

SLR7 SLR50 SLR60 SLR75 SLR100 SLR140

SLRtape140 70/140 Gbyte N/A N/A N/A N/A N/A Read/Write

SLRtape100 50/100 Gbyte N/A N/A N/A N/A Read/Write Read/Write

SLRtape75 38/75 Gbyte N/A N/A Read/Write Read/Write Read/Write Read/Write

SLRtape60 30/60 Gbyte N/A N/A Read/Write Read/Write Read/Write Read/Write

SLRtape50 25/50 Gbyte N/A Read/Write Read/Write Read/Write Read/Write Read/Write

SLRtape40 20/40 Gbyte N/A N/A Read/Write Read/Write Read/Write Read/Write

SLRtape7 20/40 Gbyte Read/Write N/A Read Read Read Read

SLR32 16/32 Gbyte N/A Read/Write Read Read Read N/A

SLRtape24 12/24 Gbyte N/A Read/Write Read Read Read N/A

SLR5 4/8 Gbyte Read Read Read Read N/A N/A

DC9250 2.5/5 Gbyte N/A Read Read

Read N/A N/A

Table 1-1: Capacity and medium types

If a certain functional specification is valid for a particular drive, two methods are used to separate the different specifications throughout this manual:

The specific functional specification is marked out with horizontal lines, or

Footnotes are used to indicate for which drive the specification applies.

The specifications described in this publication are subject to change without notice.

All drives supports different variants of these medium types. See table 30-3 for details.

The highest number is with 2:1 data compression. The actual capacity on brand new or old and worn tapes may be lower.

On SLR60 tape drives with MAN. DATE 2802 (July 2002) or higher.

Tandberg Data Introduction

1-2 Tandberg SLR Product Line SCSI Functional Specifications

1.2. Overview

Chapter 2

Describes tape streamers in general and the Tandberg SLR Product Line Tape Drives in particular. The chapter also describes volume partitioning, the overwrite function, the usage of fixed and variable length data blocks, the data buffer system and data compression.

Chapter 3

Gives a description of the SCSI logical characteristics as implemented by the Tandberg SLR Product Line Tape Drives.

Chapter 4

Specifies the SCSI Command Descriptor Blocks (CDB) in general.

Chapter 5

Lists the Status Bytes implemented by the Tandberg SLR Product Line Tape Drives.

Chapter 6

Describes the SCSI Message system as implemented by the Tandberg SLR Product Line Tape Drives.

Chapter 7

Specifies the Tandberg SLR Product Line Tape Drives exception handling in general.

Chapters 8 - 32

Detailed specifications of the SCSI commands.

Tandberg Data Introduction

1-3 Tandberg SLR Product Line SCSI Functional Specifications

1.3. Glossary

BOM Beginning Of Medium. The extreme position along the medium

in the direction from the supply-reel, which can be accessed by the use of a REWIND command.

BOP

Beginning Of Partition. The position at the beginning of the

permissible recording region of a partition. If only one partition is defined, this position is equivalent to BOM (see above).

BOT

Beginning Of Tape. Physical marker on the tape marking the

start of the useful area of the tape (located at BOM, see above).

CDB Command Descriptor Block. The structure used to communicate

commands from an Initiator to a Target.

Compression Block Group

A group of compressed data recorded as one variable length block. The Compression Block Group either contains a number of host-defined logical fixed length blocks, or a complete or partial host-defined variable length logical block. The Compression Block Group also contains a Compression Header as its initial sequence of data.

Compression Header

A sequence of uncompressed data at the beginning of each Compression Block Group. The Compression Header contains specific information related to this Compression Block Group.

Disconnect

The action that occurs when a SCSI device releases control of the SCSI-bus, allowing it to go to the BUS FREE phase.

EOD End Of Data. A tape format specific end-of-data indication

on the current partition. The recording medium may be positioned at EOD by reading until the Drive signals an EOD exception or by issuing a SPACE command with a Space Code of 3 (Space to End-Of-Data).

EOM

End Of Medium. The extreme position along the medium in

the direction from the take-up-reel, which can be accessed by the device. This position may be accessed by the use of a LOAD/UNLOAD command with the EOT-bit set to one.

EOP End Of Partition. The position at the end of the permissible

recording area of a partition. If only one partition is defined, this position is equivalent to EOM (see above).

EOT

End Of Tape. Physical marker on the tape marking the end

of the useful area of the tape (located at EOM, see above).

Early Warning. Physical tape-mark near - but logically

before - EOP (independent of physical direction).

Field

A group of one or more contiguous bits. Fields containing only one bit are usually referred to as the XX bit instead of the XX field.

Initiator

SCSI-bus Device issuing SCSI commands to a SCSI Target.

LED

Light Emitting Diode. An indicator on the front of the

Drive.

LEW Logical Early Warning. Simulated EW marker on the last track

on each partition. LEW is moved some distance in front of the actual EW.

LSB

Least Significant Bit.

LUN

Logical Unit Number.

MLR

Multi Channel Linear Recording.

MSB

Most Significant Bit.

Tandberg Data Introduction

1-4 Tandberg SLR Product Line SCSI Functional Specifications

Overlength

The incorrect length condition that exists after executing a read group command, when the length of the actual block read exceeds the requested transfer length in the command descriptor block (CDB).

Page

Several commands use regular parameter structures that are referred to as pages. These pages are identified with a value known as a page code.

Parameter

A structure containing one or more fields.

Partition

The entire region of recording and reading paths in a volume or in a portion of a volume.

Reconnect

The act of re-establishing the physical Initiator/Target connection. A Target reconnects to an Initiator by issuing RESELECTION and MESSAGE IN phases after winning arbitration.

Reserved

The term used for bits, fields and code values that are set aside for future standardization.

SCSI

Small Computer Systems Interface. Industry standard computer

peripheral interface. Used to connect several devices via a common data and control bus.

SCSI address

The representation of the unique address (0-7) assigned to a SCSI device. This address would normally be assigned and set in the SCSI device during system initialization.

SCSI ID

The bit-significant representation of the SCSI address referring to one of the SCSI-bus data lines.

Signal assertion

The act of driving a signal to the true state.

Signal de-assertion

The act of driving a signal to the false state.

SLR

Scaleable Linear Recording.

Status

One byte of information sent from a Target to an Initiator upon completion of each command.

Third-party

When used in reference to RESERVE or RELEASE commands, third-party means a reservation made on behalf of another device.

Target

SCSI-bus Device receiving/executing SCSI commands.

TLA Three Letter Acronym. Track Set

A logical collection of N physical tracks which are written or read simultaneously. A track set can be viewed as a logical track that holds N times as much data as a physical track and can transfer data N times as fast as a physical track. A track set may consist of only a single track, i.e. N = 1.

Underlength

The incorrect length condition that exists after executing a read group command when the requested transfer length in the command descriptor block (CDB) exceeds the length of the actual block read.

Volume

A recording medium together with its physical carrier.

Tandberg Data Introduction

1-5 Tandberg SLR Product Line SCSI Functional Specifications

1.4. Additional Reference Documentation

[1] Tandberg Data ASA, “ Tandberg SLR Product Line

Reference Manual” , Revision 6, Publ. No. 90316, December 1999.

[2] American National Standards Institute, “ SCSI-2,

Enhanced Small Computer System Interface” , ANSI Standard X3.131-1994.

[3] American National Standards Institute, “ SCSI-3,

Primary Commands” , ANSI Working Draft X3T10/995D, Revision 11a, 28 March 1997.

[4] American National Standards Institute, “ SCSI-3

Stream Device Command Set” , ANSI Working Draft X3T10/997D, Revision 22, 1-Jan-2000.

[5] American National Standards Institute, “ SCSI-3

Interlocked Protocol” , ANSI Working Draft X3T10/856D, Revision 10, 31 Mar 1996.

[6] American National Standards Institute, “ SCSI-3

Parallel Interface” , ANSI Working Draft X3T10/855D, Revision 15a. American National Standards Institute, “ SCSI-3 Parallel Interface 2” , ANSI Working Draft X3T10/855D, Revision 20b.

[7] QIC, “ Development Standard For 1/4-inch

Cartridge Tape Drive SCSI-2 Interface” , QIC121, Revision P, 14-Dec-1995.

[8] QIC, “ Adaptive Lossless Data Compression” ,

QIC-154, Revision A, 10-Mar-1994.

[9] QIC, “ Serial Recorded Magnetic Tape Cartridge

For Information Interchange” , QIC-2GB-DC, Revision B, 10-Mar-1994.

[10] QIC, “ Serial Recorded Magnetic Tape Cartridge

For Information Interchange” , QIC-4GB-DC, Revision B, 20-Apr-1997.

[11] Tandberg Data ASA, “ Serial Recorded Magnetic

Tape Cartridge For Information Interchange” , SLR6 Tape Format, Revision A, 1998.

[12] QIC, “ Serial Recorded Magnetic Tape Cartridge

For Information Interchange” , QIC-5010-DC, Revision E, 15-Dec-1994.

[13] Tandberg Data ASA, “ Serial Recorded Magnetic

Tape Cartridge For Information Interchange” , MLR3 Tape Format, Revision A, 30-Aug-1995.

[14] Tandberg Data ASA, “ Advanced Linear Recording

Format 1” , ALRF-1, Revision B, 22-Apr-1998.

[15] QIC, “ Serial Recorded Magnetic Tape Cartridge

For Information Interchange” , QIC-CRF1, Revision J, 15 Jun 1995.

[16] Tandberg Data ASA, “ Advanced Linear Recording

Format 2” , ALRF-2, Revision 1. 24-Apr-2000.

[17] Tandberg Data ASA, “ Advanced Linear Recording

Format 6” , ALRF-6, Revision 2.00-Oct-2002

[18] American National Standards Institute, “ SCSI-3

Fast-20” , ANSI Working Draft X3T10/1071D, Revision 6.

Tandberg Data Introduction

1-6 Tandberg SLR Product Line SCSI Functional Specifications

This Page Intentionally Left Blank

2-1 Tandberg SLR Product Line SCSI Functional Specifications

About Tape Streamers

2.1. Physical Elements

Tape streamer devices optimize their use in storing or retrieving user data in a sequential manner. Since access is sequential, position changes typically take long time, when compared to direct-access devices like disks.

The recording medium used with the Drive consists of a flexible substrate coated with a magnetic material. The recording medium is wound onto two reels.

Both the supply reel and the take-up reel are encapsulated into a cartridge. Several standards exist, covering the construction of cartridges for interchange.

A complete unit composed of the recording medium and its physical carrier (the cartridge) is called a volume. In tape streamers like the Tandberg SLR Product Line Tape Drives, the volumes are removable.

When a volume is inserted it has the attribute of being loaded or unloaded. Loaded is the state when the streamer device is capable of executing commands that cause the medium to be moved (so-called media access commands). A volume is unloaded when the media access commands can not be executed (that is when these commands report CHECK CONDITION status and a NOT READY sense key).

Not Write-Protected

Write-Protected

The write-protected state determines whether an Initiator may or may not write information on a volume. This attribute is controlled by the user of the volume through the SAFE-switch on the cartridges.

The recording medium has two physical attributes called Beginning-Of-Tap

(BOT) and End-Of-Tape (EOT). BOT is at the end of the medium that is attached to the take-up reel. EOT is at the end of the medium that is attached to the supply reel.

2.2. Data Storage Characteristics

Serpentine Recording Method

The position on the medium where a pattern of recorded signal may be written by one write component is called a track or track set which may consist of either one, two or four tracks, depending on the tape format used. On a new volume, recording of a track set begins after mounting the volume and then by moving the tape from BOT toward EOT. When EOT is approached, the direction of recording is reversed and the Drive starts recording a track set from EOT towards BOT. This process repeats the number of times necessary to record all track sets. The total number of track sets depends on the tape format used. This method of recording is called serpentine.

Tandberg Data About Tape Streamers

2-2 Tandberg SLR Product Line SCSI Functional Specifications

BO T EO T

BO P

EO P

Figure 2-1: Serpentine Recording

A track set is read in the same sequence as when writing.

When reading data using a DC9250, SLR5 or SLRtape7 medium, the drive is aided by reference tracks to position the read/write head accurately.

For the DC9250 and SLR5 media the drive

, writing these tapes from BOT, has to provide the tape with the required reference bursts during write. The writing and reading of reference bursts are done automatically by the drive and are completely transparent to the user.

The reference bursts are pre-recorded on SLRtape7 media. The reading of reference bursts are done automatically by the drive and are completely transparent to the user.

On the other hand, SLRtape140, SLRtape100, SLRtape75, SLRtape60, SLRtape50, SLRtape40, SLR32 and SLRtape24 media have pre-recorded servo tracks, and the reading and writing of reference bursts are therefore not required. The servo tracks are used to position the read/write head both during the read and write operations. The reader should be aware of this distinction when meeting the term “ reference bursts” later in this manual. See [9] – [17] for further information on reference bursts and servo tracks.

From the Initiators point of view the recording medium may be looked upon as having a single large continuous logical track starting with an area called Beginning-Of- Media (BOM) and ending with an area called End-Of-Media (EOM). BOM is always on the same side of the volume as BOT. EOM may be located at the BOT or at the EOT depending on whether the total number of track sets is odd or even.

The logical track is split into several areas separated by markers. At least four parts may be identified:

Beginning-Of-Media End- Of-MediaEarly-Warning AreaRecording Area

BO T L P LEW EW EO T

Figure 2-2: Logical Track Areas

Drives covered by this manual can only read DC9250 and SLR5 media.

Tandberg Data About Tape Streamers

2-3 Tandberg SLR Product Line SCSI Functional Specifications

Beginning-Of-Media

This area holds no user data. It is used to record a special reference burst as defined in [9], [10]. This area starts with the BOT (Beginning Of Tape) tape marker and ends at the LP (Load Point) tape marker.

Recording Area

This area holds the user data. This area starts with the LP (Load Point) tape marker and ends at the LEW (Logical Early Warning) marker.

Early-Warning Area

When writing, the Initiator needs an indication that it is approaching the end of the Recording-Area. This position called Logical Early Warning (LEW) is reported to the Initiator at a position early enough for the Drive to write out any buffered data to the medium while still leaving enough room for additional recorded data or tapemarks. The actual amount of data is user configurable, see the LEW Position field in Section

14.3.9. This area ends at the EW (Early Warning) tape marker.

End-Of-Media

This is the short area between the EW (Early Warning) and the EOT (End-Of-Tape) tape markers. It is usually possible to complete the writing of a single last frame between EW and EOT. When this has been done or when EOT is found, all further write operations are discontinued even if the there are more data to be written in the data buffer.

2.3. Partitions Within a Volume

A volume may be split into several mini-volumes called partitions. Each partition has its own set of beginning and ending points. Each partition within a volume has defined its own Beginning-Of-Partition (BOP), RecordingArea, Early-Warning-Area and End-Of-Partition (EOP).

In the first partition (partition zero) on a volume with n partitions, BOP is identical to BOM. In the last partition on a volume (partition n-1), EOP is identical to EOM.

All volumes have a minimum of one partition called partition zero, or the default partition.

When a volume is mounted (that is inserted into the Drive and then loaded), the volume is logically positioned to the beginning of the default partition (partition zero). This is also true when a LOAD command (with the Load bit set to one) is executed. When a REWIND or ERASE command is received in any partition, the Drive positions to the beginning of the current partition.

Tandberg Data About Tape Streamers

2-4 Tandberg SLR Product Line SCSI Functional Specifications

2.3.1. Partitioning a Volume

Setting up a Medium Partition page and then executing a MODE SELECT command transferring this page to the Drive can partition a volume. The table below shows the parts of the Medium Partition Page(1) that relate to this procedure. Fields of no interest are marked with “ XX...XX” .

BYTE BIT 7 6 5 4 3 2 1 0

XX Page Code = 11h

01 Parameter Length = 4Eh

Maximum Additional Partitions

Additional Partitions Defined

FDP SDP IDP PSUM

XX XX XX

05 XXXX

06 XXXX

Partition Units

07 XXXX

Partition Size Descriptor 0

| . . .

Partition Size Descriptor 35

Table 2-1: The Medium Partition Page(1)

Note that the Tandberg SLR Product Line Tape Drives use the long form of the Medium Partition Page(1). This means that the Partition Size Descriptor fields are always present (both in MODE SELECT and MODE SENSE parameter data).

Note also that the Medium Partition Page(1) is not savable (the MODE SENSE command returns a Medium Partition Page(1) where the PS bit is set to zero).

Partitioning a volume is only allowed when the volume is positioned at BOM (the beginning of partition 0). Note also that a partition operation may lead to a loss of all existing data (the Drive runs an erase pass in each of the existing partitions if the new partitioning is different from the existing partitioning found on the medium).

The Tandberg SLR Product Line Tape Drives allow the Initiator to specify partitioning using two different methods. The first is the most general. The Initiator can specify both the number of partitions and their individual sizes. Using the second method the Initiator may only specify whether one or two partitions are requested. When a single partition is requested it will cover the whole medium. When two partitions are requested the Drive will allocate them fixed sizes (there will be one large and one small partition).

The Initiator Defined Partitions (IDP) bit is used to select between the two methods. When the IDP bit is one, the first method is used (the number and size of partitions may be specified).

Tandberg Data About Tape Streamers

2-5 Tandberg SLR Product Line SCSI Functional Specifications

When the IDP bit is zero, the second method is used. In this case the FDP bit is used to select one or two fixed size partitions. This method of using partitions is often called “ Quick File Access” (QFA).

Regardless of which method is used, executing a MODE SELECT command where the Medium Partition Page(1) is included in the parameter list, some form of partitioning will take place. This means that such a MODE SELECT command is only accepted at the beginning of partition 0 (BOP0).

The Partition Size Unit of Measure (PSUM) field defines the unit of the Partition Size Descriptors. A value of 2 specifies the unit to be MBytes and a value of 3 specifies the unit to be GBytes. MODE SELECT will ignore this field when the IDP bit is not set. MODE SENSE will return 2 (MBytes) in this field when the currently mounted volume has a capacity (uncompressed) of 65535 MBytes or less. MODE SENSE will return 3 (GBbytes) in this field when the capacity is above 65535 MBytes. This boundary is introduced because 65535 (FFFFh) is the maximum value of the Partition Size Descriptors.

The Partition Units field defines the unit of the Partition Size Descriptors when the PSUM field is set to

3. The Partition Units must be set to 9 when the PSUM field is set to 3 to indicate that GBytes (10

) is the

only supported unit. The Maximum Additional Partitions field is ignored by

the MODE SELECT command. When the Medium Partition Page(1) is returned by the MODE SENSE command this field is used to indicate the maximum number of partitions (in addition to the default partition) that the current medium is able to hold.

The Additional Partitions Defined field is used to specify the requested number of partitions (MODE SELECT) or it shows the actual number of partitions (MODE SENSE). The Partitions Size Descriptors is used to specify the requested size of each partition (MODE SELECT) or to show the current (estimated) size of each partition (MODE SENSE).

Note that setting the IDP or FDP bits to one is only allowed when the medium to be partitioned is a SLRtape140 to SLRtape7 or SLR32, medium. In this case a maximum of 24 partitions is allowed when using a SLRtape140, SLRtape100, SLRtape75, SLRtape60 or SLRtape40 medium. A maximum of 36 partitions is allowed when using a SLRtape50, SLR32 or SLRtape24 medium. A maximum of 3 partitions is allowed when using a SLRtape7 medium. (See also section 2.3.3 Using Initiator Defined Partitions).

When using other medium types, partitioning can not be used and only a single partition is available.

When partition sizes are quoted later on in this section, all sizes are uncompressed sizes.

Tandberg Data About Tape Streamers

2-6 Tandberg SLR Product Line SCSI Functional Specifications

2.3.2. Selecting a Partition

Two commands makes it possible to select a partition for read or write operations; the LOCATE command and the MODE SELECT command.

2.3.2.1. Selecting a Partition Using LOCATE

The LOCATE command can be used to specify the active partition. The Change Partition (CP) must be set to one and the Partition field must be set to the desired partition. The addressed partition must exist on the volume mounted. The Drive will then move to the specified partition before the actual block locate operation starts.

The READ POSITION and MODE SENSE commands may be used to read the current partition number in use.

When the partition has been changed, the Drive will stay in the selected partition for all media access commands until one of the following actions are taken (by the Initiator):

 A new LOCATE with CP set to one is executed.

 A MODE SELECT command is executed that changes the

state of the Active Partition field in the Device Configuration Parameters Page (page code 10h).

 A LOAD command is executed. This command always

brings the tape to the beginning of the default partition (partition 0).

 The cartridge is removed from the Drive. A newly

inserted cartridge must be loaded before it can be accessed. The load operation positions the tape at the beginning of the default partition (partition 0). This will be true whether the load operation happens as a result of an Auto Load or an explicit LOAD command (with the Load bit set to one).

2.3.2.2. Selecting a Partition Using MODE SELECT

The MODE SELECT command can be used to specify the active partition. The Device Configuration Parameters Page (page code 10h) is used for this purpose. The Change Active Partition (CAP) bit must be set to one and the Active Partition field must be set to the desired partition. The addressed partition must exist on the volume mounted. The Drive will then position the tape at the beginning of the specified partition.

The READ POSITION and MODE SENSE commands may be used to read the current partition number in use.

When the partition has been changed, the Drive will stay in the selected partition for all media access commands until one of the following actions are taken (by the Initiator):

 A new MODE SELECT command is executed that changes

the state of the Active Partition field.

Tandberg Data About Tape Streamers

2-7 Tandberg SLR Product Line SCSI Functional Specifications

 A LOCATE with CP set to one is executed.

Tandberg Data About Tape Streamers

2-8 Tandberg SLR Product Line SCSI Functional Specifications

 A LOAD command is executed. This command always

brings the tape to the beginning of the default partition (partition 0).

 The cartridge is removed from the Drive. A newly

2.3.3. Using Initiator Defined Partitions

When using a SLRtape140 to SLRtape7 or SLR32, medium the Initiator may specify both the number and the size of partitions (the IDP bit is set to one).

The following table shows the minimum number of tracks and track sets that can be allocated to a partition for each medium type. It also shows the maximum number of partitions and the minimum partition size.

Medium Type

Min. Track Sets

Min. Tracks

Max. Partitions

Min. Partition Size

SLRtape140 2 8 24 2916 MBytes

SLRtape100

2083 MBytes

SLRtape75 2 8 24 1563 MBytes

SLRtape60 2 8 24 1250 Mbytes

SLRtape50 2 4 36 694 Mbytes

SLRtape40 2 8 24 833 MBytes

SLRtape7 12 24 3 6666 Mbytes

SLR32 2 4 36 444 Mbytes

SLRtape24 2 4 36 333 Mbytes

Table 2-2: Minimum Partition Size

All partitions start at the BOT side of the medium. This

means that the Drive will be able to access any partition with a minimum of tape movements after an auto-load or a LOAD command.

Tandberg Data About Tape Streamers

2-9 Tandberg SLR Product Line SCSI Functional Specifications

Here is a simple example using an SLR32 medium. The volume has been partitioned into 3 partitions (one default partition and 2 additional partitions):

BOT EOT

BOP 0 EOP 0

EOP 2

BOP 2

BOP 1

EOP 1

Tracks in Partition 0 Tracks in Partition 1 Tracks in Partition 2

Figure 2-3: Multiple Partition Track Layout (SLR32 example)

The default partition (partition 0) has been allocated 2

track sets, partition 1 has been allocated 4 track sets while the last partition, which is partition 2 in this case, has been allocated the remainder of the medium (66 track sets).

When the IDP bit in the Medium Partition Page(1) is set to zero, the Drive allocates one or two fixed sized partitions on the volume mounted. See section 2.3.4. Quick File Access (All Formats) for information on how to use this method.

When the IDP bit is set to one, the Additional Partitions Defined field is used to specify the number of partitions requested. Note that this field specifies the number of partitions requested in addition to the default partition. If a total of 10 partitions are requested, the Additional Partitions Defined must be set to 9. The Partition Size Descriptors are used to specify the individual sizes of the requested partitions. Note that the size of the default partition (partition 0) must be included among the Partition Size Descriptors.

Since there is no provision in a Partition Size Descriptor to identify the number of a partition, the number of a partition is the relative position of the Partition Size Descriptor in the Medium Partition Page(1). The size of partition 0 is specified in the first Partition Size Descriptor, that is in bytes 8 and

9. Partition 1 is specified in bytes 10 and 11, etc.

Tandberg Data About Tape Streamers

2-10 Tandberg SLR Product Line SCSI Functional Specifications

When the Medium Partition Page(1) is used to specify the size of any partition, the Partition Size Descriptor fields are used to specify a requested size. As the actual partition size must match an integral number of track set pairs, the Drive will round the value requested up to the nearest number divisible by the minimum partition size given by the loaded medium (see table 2-2). As an example, if a partition is requested to have a value of 450 MBytes, the Drive will round the size up to 666 MBytes (333 * 2) when using an SLRtape24 medium, up to 694 MBytes when using an SLRtape50 medium and up to 2916 MBytes when using an SLRtape140 medium. If a partition is requested to have a size of 200 MBytes, the Drive will round the size up to 333, 444, 694 or 2916 Mbytes and so on. The Partition Size Descriptors returned by the MODE SENSE command may be used to read out the partition sizes after rounding has taken place.

The number of Partition Size Descriptors must always equal the Maximum Addition Partition plus one, specifying the size of the default partition plus all additional partitions. When the Additional Partitions Defined is less than the Maximum Additional Partitions, the last Partition Size Descriptors are set to zero. If the Maximum Additional Partitions field has a value of 35 and the Additional Partitions Defined has a value of n, then there must be n+1 non-zero Partition Size Descriptors. These descriptors must be the first n+1 descriptors (the medium has been partitioned into n+1 partitions numbered 0 through n). The last 35-n Partition Size Descriptors are all set to zero. The size of partition 0 (the default partition) must be greater than zero.

The sum of all sizes requested by the non-zero Partition Size Descriptors may be less than the total size on the medium. In this case some capacity remains inaccessible.

Sometimes it may be desirable to have the last partition cover the remainder of the medium. In this case the last non-zero Partition Size Descriptor can be set to a value of FFFFh.

If insufficient space exists on the medium for the requested partition sizes or if multiple Partition Size Descriptors are set to FFFFh, the Drive will return CHECK CONDITION status.

When medium information has been read by the Drive, the Medium Partition Page(1) returned by a MODE SENSE command indicates the actual partitioning on the currently mounted volume (the number and size of each partition). When medium information has not yet been read by the Drive, the Additional Partitions Defined and the Partition Size Descriptor fields are undefined and set to zeros. The SLR5 and DC9250 media have limited medium information and the Additional Partitions Defined and the Partition Size Descriptor fields are always undefined and set to zeros for these media.

Note however that the partitions sizes returned by a MODE SENSE command are estimated sizes. The sizes are based on the calculated capacity for a given number of pairs of track sets. The actual size achieved may differ from the estimated size due to such factors as tape length variations, the number of re-writes and amount of filler data written by the Drive.

Tandberg Data About Tape Streamers

2-11 Tandberg SLR Product Line SCSI Functional Specifications

2.3.4. Quick File Access

Partitions can be used to support the implementation of QFA (Quick File Access). QFA is a feature which provides support for two partitions on a volume, a directory partition and a data partition. When using QFA the initiator is not allowed to specify the size of those partitions. Only the number of partitions (one or two) can be specified.

The QFA mode can be enabled and disabled with the FDP bit of the Medium Partition Page(1). When not in QFA mode the Drive will implement a single partition covering the whole tape. This single partition is called the Default Data Partition.

Note that QFA is only supported when the medium to be partitioned is a SLRtape140 to SLRtape7 or SLR32 medium.

Partition

Use

0 Data

1 Directory Information

Table 2-3: Partitions Within A QFA Volume

In QIC compatible tape streamers like the Tandberg SLR Product Line Tape Drives, all partitions start on track set boundaries at the physical BOT end of the tape. Two track sets (a total of 4 tracks) are allocated to the directory partition when using the SLR6, QIC-5010-DC, MLR3 or ALRF-2 tape formats. For the ALRF-1 and ALRF-6 tape format two track sets (a total of 8 tracks) are allocated. All remaining track sets are allocated to the data partition.

The following table shows the directory track set allocation for all supported tape formats:

Medium Type

Approximate Size Directory Track (Set) Number

SLRtape140 2916 MBytes 46 and 47

SLRtape100 2083 MBytes 46 and 47

SLRtape75 1563 MBytes 46 and 47

SLRtape60 1250 MBytes 46 and 47

SLRtape50 694 MBytes 70 and 71

SLRtape40 833 MBytes 46 and 47

SLRtape7 1111 MBytes 34 and 35

SLR32 444 MBytes 70 and 71

SLRtape24 333 MBytes 70 and 71

Table 2-4: Directory Track set Allocation

When the medium has been identified the Medium Partition

Page(1) returned by a MODE SENSE command may indicate the actual partitioning on the currently mounted volume (the number and size of each partition). When the medium has not been identified the Additional Partitions Defined and the Partition Size Descriptor fields are undefined.

Note that the partition size(s) returned by a MODE SENSE command are estimated sizes. The sizes are based on the calculated capacity for a given number of tracks. The actual size achieved may differ from the estimated size due to such factors as tape length variations, the

Tandberg Data About Tape Streamers

2-12 Tandberg SLR Product Line SCSI Functional Specifications

number of re-writes and amount of filler data written by the Drive.

Tandberg Data About Tape Streamers

2-13 Tandberg SLR Product Line SCSI Functional Specifications

2.3.4.1. Setting QFA-Mode

The MODE SELECT command is used to enter or leave the QFA mode.

The Drive will be in non-QFA mode when the FDP bit in the Medium Partition Parameter Page (Page Code 11h) is set to zero. The Drive will be in QFA mode when the FDP bit is set to one.

When going into QFA mode the FDP bit must be set to one (the IDP bit must be set to zero). The Page Length field must be set to 4Eh (to keep the page at a non-changing length). The Maximum Additional Partitions field is ignored. The Additional Partitions Defined field is ignored (the Drive always partition the volume into two partitions when FDP set to one). The SDP and the IDP bits must be set to zero. The PSUM and Partition Units fields are ignored. The CLEAR and ADDP bits must be set to zero. The Medium Format Recognition field must be set to 01h (the value returned by the MODE SENSE command). The Partition Units field must be set to zero. The Partition Size Descriptor fields are ignored by the drive (they can have any value).

When leaving QFA mode (when partitioning the volume into a single partition) all fields have the values given above. The only exception is that the FDP bit is set to zero.

Setting or clearing the FDP bit is only legal when the tape is positioned at BOT. Note also that a change in partitioning (going from 1 to 2 partitions or the other way around) leads to loss of all existing data (the Drive runs an erase pass in each of the existing partitions).

Since the Medium Partition Page(1) is not savable, the Initiator is responsible for making sure that the correct QFA mode is active whenever a new cartridge is loaded.

When a QFA mode corresponding to the currently loaded volume is in effect, the Medium Partition Page(1) returned by the MODE SENSE command will indicate both the number of partitions in effect and the size of those partitions. In non-QFA mode the Additional Partitions Defined field returns 00h (indicating a single partition). The very first Partition Size Descriptor field returns the approximate size of the default (and only) partition. In QFA mode the Additional Partitions Defined field returns 01h (indicating a total of two partitions). The first Partition Size Descriptor field returns the approximate size of the data partition while the second Partition Size Descriptor field indicates the approximate size of the directory partition.

2.3.4.2. Changing Partition Using The LOCATE Command

The LOCATE command can be used to specify the active partition. The Change Partition (CP) must be set to one and the Partition field must be set to 0 (Data Partition) or 1 (Directory Partition). The Drive will then move to the specified partition before the actual block locate operation starts.

Tandberg Data About Tape Streamers

2-14 Tandberg SLR Product Line SCSI Functional Specifications

2.3.4.3. Changing Partition Using the MODE SELECT

Command

The MODE SELECT command can be used to specify the active partition. The Device Configuration Parameters Page (page code 10h) is used for this purpose. The Change Active Partition (CAP) bit must be set to one and the Active Partition field must be set to 0 (Data Partition) or 1 (Directory Partition). The Drive will then position the tape at the beginning of the specified partition.

2.4. Logical Elements Within a Partition

The Recording-Area on a volume may contain two types of Initiator accessible elements; data blocks and tape marks. These elements are controlled and transferred between the Initiator and the medium using READ, WRITE and WRITE FILEMARKS commands.

A unit of data supplied or requested by the Initiator is called a logical block. Logical blocks are stored according to the specifications of the tape format used (see [9] - [14]) and may be recorded as one or more physical blocks on the medium. When the physical block and the logical block are not recorded in a one-to-one relationship, it is the responsibility of the device to perform all blocking, de-blocking or padding of the logical block(s) sent to/from the Initiator.

Filemarks are one kind of tape marks. Filemarks are special recorded elements containing no user data. Initiators traditionally use filemarks to separate user data from labels and logical groupings of data from each other.

Setmarks are another kind of tape marks. Setmarks are special recorded elements containing no user data, providing a segmentation scheme hierarchically superior to filemarks.

Interblock gaps, the gaps between blocks and tapemarks, are introduced on the medium at the time a block or tapemark is written without explicit action by the Initiator. Minimum and maximum lengths for interblock gaps are defined in [9] - [14].

In addition to blocks, tape marks and inter-block gaps, erase gaps can be recorded. An erase gap is automatically recorded when a write operation is properly terminated. The erase gap acts as an end-ofdata marker on some tape formats.

In addition a single erase gap may be recorded on the medium through the use of the ERASE command. This erase gap will cover the whole - or a limited part - of the medium and the pre-recorded information in the erase gap will be written over and lost. Refer to section 8 for further details about the ERASE command.

Tandberg Data About Tape Streamers

2-15 Tandberg SLR Product Line SCSI Functional Specifications

After writing data from BOP on a partition, the medium is considered to be a contiguous grouping of blocks, filemarks, setmarks and gaps. Certain ANSI standards define gap lengths which, if exceeded, are to be considered as having reached blank medium. Depending on the tape format, this blank medium may be treated as an endof-data indication or an unrecoverable medium error causing an interchange error. Unrecorded volumes (new or erased) may exhibit blank medium characteristics if an attempt is made to read or space the volume before data have been written.

2.5. Overwrite

Normally tape streamers do not allow update of data already written on a volume. The structure and layout of the physical elements on the Recording-Area are such that they do not facilitate writing over old data with new data in the middle of a partition. When writing new data on a volume there are only two possibilities:

 A write operation may be started from BOP. All

existing data on that partition will be destroyed. This is true even if the new write operation writes less data than the amount of data already on the partition. The data “ left” on the partition can not be recovered.

 A write operation may append data to existing data on

a partition. In this case the medium must be positioned at end-of-data (EOD) before the write operation is started. The medium may be positioned at EOD by reading or spacing until the Drive indicates EOD. It is also possible to position the medium to EOD by executing a SPACE command with a Space Code of 3 (Space to EOD).

If the current tape position is not at BOP or EOD, a write operation will be terminated immediately with CHECK CONDITION status and the Error Code is set to W

RITE

PPEND ERROR

(Write After Read).

However, in addition to the two possibilities mentioned above, the Tandberg SLR Product Line Tape Drives offer under certain conditions the following two overwrite capabilities:

 The data following the very first data block in a

partition may be written over (leaving the first data block intact). The following restrictions do however apply:

The medium must be positioned after the 1st logical block in partition 0.

The length of this block must be less than 65536 (10000h) bytes when using a SLRtape140 to SLRtape7 or SLR32, medium or less than 32768 (8000h) bytes when using an SLR5 or DC9250 medium.

There must be no tapemarks either immediately in front of or immediately after the 1st data block.

Tandberg Data About Tape Streamers

2-16 Tandberg SLR Product Line SCSI Functional Specifications

BO P EO P

1s t block

Other existing data

New data written over ex ist ing data

Bef ore

After

Figure 2-4: Overwrite After the 1st Logical Data Block

 The last filemark in a partition may be written over.

The following restrictions do however apply:

The filemark must be the last of two sequential filemarks.

EOD must follow immediately after the filemark to be written over (no data must follow the two filemarks).

The medium must be positioned in front of the filemark to be written over.

BO P EO P

2nd fmk

New data written over last fmk

Before

After

EO D

!st fmk

EO D

Figure 2-5: Overwrite of Last Filemark

The overwrite capability is a configurable feature. It is controlled by the EOWR bit in the Miscellaneous Parameters Page of the MODE SELECT command.

Tandberg Data About Tape Streamers

2-17 Tandberg SLR Product Line SCSI Functional Specifications

2.6. Using Fixed and Variable Length Blocks

2.6.1. Variable and Fixed Length Blocks

The Drive is capable of supporting both fixed and variable length logical blocks. The concept of fixed or variable mode for writing and reading blocks only indicates the method by which the Initiator specifies the size of the logical block for transfer and not the method of recording physical blocks on the medium.

When writing the Drive transforms logical blocks passed from the Initiator into physical blocks to be stored on the tape. The length of a logical block may vary while the length of a physical block is defined by the tape format in use (see [9] - [14]). When reading the Drive transforms physical blocks from the tape back into logical blocks to be transferred to the Initiator,

When executing read and write commands two parameters must be specified; the block length and the number of blocks to read or write. As the SCSI read and write commands only have room for a single parameter, only the block length or the number of blocks can be specified directly. For this reason there are two different versions of each command capable of transferring tape data to or from the Initiator. The fixed length type commands can specify the number of blocks to be processed. For these commands all blocks will be of equal size. The actual size is given by the current value in the Block Size field in the Block Descriptor List of the MODE SELECT command. The variable length type commands can specify the length of the individual blocks, but the number of blocks written is always only one (for each command).

When the Block Size field has a value different from zero, the Drive is said to be in fixed block mode. In this mode both the fixed length type commands and the variable length type commands are allowed. When the Block Size field is zero, the Drive is said to be in variable block mode. In this mode only the variable length type commands are allowed (as writing fixed length blocks of size zero bytes does not make any sense).

Note that the illegal length handling in the READ command differs slightly depending on the current fixed/variable mode set by the MODE SELECT command.

2.6.2. Writing

When writing, the Drive groups the data transferred from the Initiator into blocks. Data blocks may be written with two different versions of the WRITE command; one with the fixed (FIX) bit set and one with the fixed (FIX) bit cleared.

When the FIX bit is cleared, a WRITE command will write a single block. The block length may be specified on a block-by-block basis. This is useful when writing blocks of varying length.

Tandberg Data About Tape Streamers

2-18 Tandberg SLR Product Line SCSI Functional Specifications

When the FIX bit is set to one, a WRITE command may write multiple blocks. The WRITE command must specify the number of blocks to write. All written blocks will be of the same length. The block size is controlled by the Block Size field in the MODE SELECT Block Descriptor List. Depending on the tape format used, different fixed block sizes are supported.

Note that when the Block Size field has been set to zero, the Drive is said to be in Variable Block mode and commands with the FIX bit set to one are not allowed.

2.6.3. Reading

When reading data off a tape, the Drive is able to determine the length of each block read. When reading, the expected block length must be specified. This can be done in two different ways with the two different versions of the READ command; one with the fixed (FIX) bit set and one with the fixed (FIX) bit clear.

When the FIX bit is clear, a READ command will read a single block. The expected block length may be specified on a block-by-block basis.

When the FIX bit is set to one, a READ command may read multiple blocks. The READ command must specify the number of blocks to read. The expected block length of all blocks is the same. The expected length is specified with the MODE SELECT command (the Block Size field of the Block Descriptor List). Note that when the Block Size field has been set to zero, the Drive is said to be in Variable Block mode and commands with the FIX bit set to one are not allowed.

2.6.4. Illegal Length Conditions when Reading

When the specified block length does not match the actual block length, the READ command will complain (if the Suppress Illegal Length Indicator (SILI) bit was not set in the READ Command Descriptor Block).

If the actual block length is smaller then the expected block length, the READ command will transfer the actual number of bytes found in the block. The READ command is then terminated with a CHECK CONDITION Status. The Illegal Length Indicator will be set in the Sense Data List. If the fixed-bit is cleared, the Information Bytes will be set to the difference between the expected number of bytes and the actual number of bytes. This will be a positive number in this case. If the fixed-bit is set, the Information Bytes will be set to the difference between the specified number of blocks and the actual number of blocks transferred. The block with the unexpected length is counted among the transferred blocks even if its length was wrong. Note that this means that the Information Bytes may read zero even if the READ command was terminated with CHECK CONDITION and Illegal Length Indication. The logical tape position will be at the beginning of the next block on the tape.

Tandberg Data About Tape Streamers

2-19 Tandberg SLR Product Line SCSI Functional Specifications

If the actual block length is larger than the expected block length, the READ command will transfer the expected number of bytes only. The READ command is then terminated with a CHECK CONDITION Status. The Illegal Length Indicator will be set in the Sense Data List. If the fixed-bit is cleared, the Information bytes will be set up with the difference between the expected number of bytes and the actual number of bytes. This will be a negative number in this case. The information bytes are presented as a 32 bit 2's complement number. If the fixed-bit is set, the Information bytes will be set up with the difference between the specified number of blocks and the actual number of blocks transferred. The block with the unexpected length is counted among the transferred blocks even if its length was wrong. Note that this means that the Information Bytes may read zero even if the READ command was terminated with CHECK CONDITION and Illegal Length Indication. The logical tape position will be at the beginning of the block following the block with the unexpected length. This means that the additional bytes in the block with the unexpected length are lost.

2.7. Data Buffering

2.7.1. Introduction

The Drive has a temporary storage area capable of holding one or more blocks or tapemarks – a data buffer. The data buffer may hold any combination of data blocks and tapemarks in the process of being written to the recording medium, or it may contain read-ahead data blocks and tapemarks transferred from the recording medium.

The data buffer operates as a FIFO queue, compensating for the different transfer rates on the SCSI-bus and the tape system. The Drive is usually only connected to the SCSI-bus for short amounts of time when bursts of data are transferred at a much higher speed than the normal tape transfer rate. The data buffer can be in one out of two modes; read mode or write mode. The data buffer is in write mode when executing WRITE or WRITE FILEMARKS commands. The data buffer is in read mode when executing LOCATE, READ, SPACE or VERIFY

commands.

VERIFY is not supported by SLR7 and SLR140

Tandberg Data About Tape Streamers

2-20 Tandberg SLR Product Line SCSI Functional Specifications

2.7.1.1. SLR50 Drive

The data buffer in the SLR50 Drive has been split into two distinct partitions, a host buffer and a tape buffer.

Host Buffer

The host buffer is used to buffer logical blocks transferred at high-speed (up to 20 MBytes/s) to/from the SCSI interface. Notice that only data are buffered in the host buffers. Tapemarks do not occupy any space in the host buffer. Tapemarks do however consume other resources in the buffers system so there is an upper limit to the number of tapemarks that the buffer system is able to hold.

Tape Buffer

The tape buffer is used to buffer formatted physical blocks going to/from the recording medium. The data in the tape buffer is transferred to/from the media at speeds varying from up to 2 MBytes/s (SLRtape50 medium) down to 480 KBytes/s (DC9250 medium).

A data formatting unit sits in the data path between the host buffer and the tape buffer. An optional data compression unit may also be placed in the data path between the two buffers.

Lo g i cal Block

Physica l Fr am e

SC SI Bus

Medium

Host Buffer Tape Buffer

Data

Fo r matter

Data

Com pression

Figure 2-6: The SLR50 Buffer System (Write Mode)

In a SLR50 Drive equipped with a 1 MByte physical data buffer the host buffer will occupy about 450 KBytes while the tape buffer is allocated 512 KBytes. For drives equipped with a 2 MByte physical data buffer the host buffer will occupy about 1450 KBytes while the tape buffer is allocated 512 KBytes.

Note that the use of the host buffer is optimized so that as long as most logical blocks are of the same size the complete allocated size for the host buffer is utilized to buffer data. When mixing blocks of different sizes or when a very large number of tapemarks are flowing through the buffer system, the effective host buffer size may be somewhat less than the allocated size.

Tandberg Data About Tape Streamers

2-21 Tandberg SLR Product Line SCSI Functional Specifications

2.7.1.2. SLR7, SLR60, SLR75, SLR100 and SLR140 Drives

The data buffer in the SLR7, SLR60, SLR75, SLR100 and SLR140 Drives consists of a single logical unit with a size of nearly 8 MByte.

Data Buffer

The data buffer is used to buffer data from logical blocks transferred at high-speed (up to 40 MBytes/s) to/from the SCSI interface. When moved into the buffer the data is first compressed and then formatted into frames of physical blocks. The data in the data buffer is then transferred to/from the media at speeds varying from up to 6 MBytes/s (SLR140) down to 3 Mbytes/s (SLR7).

Lo g ica l Block

Physica l Fr am e

SC SI Bus Medium

Data Buffer

Data

For matter

Data

Com pression

Lo g i cal to Physica l M apping

Figure 2-7: The SLR140, SLR100, SLR75, SLR60 and SLR7 Buffer System (Write

Mode)

Note that the use of the buffer is optimized so that as long as most logical blocks are of the same size the complete allocated size for the host buffer is utilized to buffer data. When mixing blocks of different sizes or when a very large number of tapemarks are flowing through the buffer system, the effective buffer size may be somewhat less than the physical size.

2.7.2. Data Formatting

When data are moved from the host buffer to the tape buffer it is grouped into physical blocks. The physical blocks are again grouped into frames. In addition to host related data and tapemarks the frames hold additional information like error correction codes (ECC) [9]

- [17]. The size of the physical block as well as the number of

blocks per frame depend on the tape format used [9] – [17].

The tape buffer is used to buffer complete frames. The maximum number of frames depends on the tape format in use. With an SLR32 medium for example, the tape buffer is able to hold up to 16 frames each holding a maximum

Tandberg Data About Tape Streamers

2-22 Tandberg SLR Product Line SCSI Functional Specifications

of 52 physical blocks of user data (each block is 512 bytes long). In addition each frame has 12 blocks of ECC information [13]. For other tape formats the values differs.

Tandberg Data About Tape Streamers

2-23 Tandberg SLR Product Line SCSI Functional Specifications

The size of a logical data block may be different than the size of a physical data block. In the tape buffer the size of a physical block is fixed (as given by the current tape format). In the host buffer the size of a logical block may vary since the Drive supports a wide range of fixed- and variable logical block sizes. The required blocking and de-blocking into logical blocks is handled by the data-formatting unit.

When the data compression option is not used, one logical block usually occupies at least one complete physical block. There is one exception. When writing 512 byte blocks with the FIXED bit set to one, a special option is used on SLR5 and DC9250 media allowing two logical blocks to share one physical block [9] - [10].

Large logical blocks may span several physical blocks. When data compression is used, the number of physical

blocks needed for each logical block depends on the compression ratio achieved.

2.7.3. Buffered Mode

The Drive is capable of operating in both a buffered mode and an unbuffered mode. Buffered mode is not applicable during read commands. When operating in buffered mode the Drive returns GOOD status for write operations when all data have been successfully transferred from the Initiator into the data buffer. When operating in unbuffered mode, GOOD status is not returned until all requested data or tapemarks are successfully recorded on the medium.

When issuing a buffered WRITE FILEMARKS command with the immediate bit set to one, GOOD status is returned as soon as the tapemark(s) has been moved into the data buffer. A WRITE FILEMARKS command with the immediate bit set to zero causes any buffered data blocks or tapemarks to be written to the medium. Upon successful completion of this process, which is called a synchronize operation, no data blocks or tapemarks remain in the data buffer. (Host may perform this synchronize operation without writing actual tapemarks by specifying zero in the number of tapemark field in the WRITE FILEMARK command).

Should an unrecoverable write error occur while in buffered mode, the Drive generates an error condition. The error is reported on the next applicable operation as a deferred error.

Note that when operating in buffered mode the Drive will pack physical blocks from one WRITE or WRITE FILEMARKS command together with physical blocks from the previous WRITE or WRITE FILEMARKS command into the same frame (when a frame based tape format is used). This is not true when operating in non-buffered mode. If a WRITE or WRITE FILEMARKS command does not transfer enough data (or tapemarks) to fill a complete frame, the rest of the frame is filled up with filler blocks. This may of course waste a lot of space on a tape. When reading a tape containing frames with filler blocks the Drive is able to remove the filler blocks without generating any discontinuities in the data stream on the SCSI-bus. Note, however, that even if filler blocks are inserted

Tandberg Data About Tape Streamers

2-24 Tandberg SLR Product Line SCSI Functional Specifications

and removed automatically by the Drive they will occupy buffer space both in read mode and write mode.

Tandberg Data About Tape Streamers

2-25 Tandberg SLR Product Line SCSI Functional Specifications

2.7.4. Read-Ahead

When operating in read mode the Drive always tries to fill up the buffer with read-ahead data. This means that when a read type command terminates the Drive continues to read data off the tape. The read-ahead only stops when the data buffer is full, if a non-read command is issued or if there is no more data to read. The readahead also stops when an uncorrectable block has been detected.

Read-ahead minimizes tape start and stops because when a read type command (including SPACE and VERIFY

) follows another read type command the wanted data may already be ready in the data buffer.

When an error has been detected by the tape system the Drive does not report the error until all data up to the point where the error occurred has been read out of the data buffer. This means that if an unrecovered read error has occurred this error is not reported until an Initiator requests the data block in error.

2.7.5. Underrun/Overrun

When the Drive is in write mode data is moved from the data buffer on to the tape. The tape write operation can only continue as long as there is any data left to write in the data buffer. If the buffer for any reason becomes empty the write operation must be stopped. This can happen if the Initiator is too slow transferring data or if the Drive is configured in non-buffered mode. When the tape is stopped in this way an underrun condition has occurred. Note, however, that the tape is not stopped immediately. The Drive is able to delay the stopping for some limited amount of time. This may be controlled by the Initiator by manipulating the Forced Streaming Count in the Miscellaneous Parameters Page (see the MODE SELECT command). When the data buffer later becomes non-empty the tape write operation can continue (see also the next section)

When the Drive is in read mode data is moved from the tape and into the data buffer. If the data buffer becomes full the tape read operation must be stopped immediately. This can happen if the Initiator is too slow transferring data or if the Initiator has stopped issuing READ, SPACE or VERIFY

commands (remember the tape read-ahead). When the tape is stopped in this way an overrun condition has occurred. When the data buffer later is emptied the tape read operation may continue (see also the next section).

The number of underrun/overrun conditions encountered may be reduced by utilizing the Automatic Velocity Control (AVC) and/or Forced Streaming. If the Automatic Velocity Control is enabled, the drive will select the speed which, based on the data transfer rate, will optimize streaming activity and minimize medium repositioning (only if more than one speed is available for the active tape format). If Forced Streaming is enabled, the drive will keep on rewriting old data a configurable number of times before going into an

VERIFY is not supported by SLR7 and SLR140

Tandberg Data About Tape Streamers

2-26 Tandberg SLR Product Line SCSI Functional Specifications

underrun condition. See 2.9. Optimizing Streaming Operation and the sections on MODE SELECT and MODE SENSE

commands for further details about Automatic Velocity Control, tape speeds and Forced Streaming.

2.7.6. Buffer Thresholds

Both the host buffer and the tape buffer act like large FIFOs. Data blocks and tapemarks (and for the tape buffer also control blocks, filler blocks and ECC blocks) are inserted in one end of the FIFO and the same data is later removed at the other end.

In write mode, when moving data into the host buffer, data is entered at a certain preferred amount at the time. If there is not enough free space in the host buffer for this amount, no logical data blocks are entered at all. The room required is called a buffer threshold. There must usually be at least room for an amount of data corresponding to the buffer threshold before new data are transferred in from the SCSI-bus. This will minimize the number of disconnects/reconnects. This buffer threshold is configurable.

SC SI -bus Side Data Formatter Side

Data in Buf ferData going in Data going out

Threshold Detector

Figure 2-8: Host Buffer Data flow During Write Operations

When transferring data blocks out of the tape buffer and onto the tape a similar mechanism exists. When the FIFO for some reason is empty (an underrun condition), the tape write operation will not be started again until the number of blocks (data, control, filler, ECC or tapemarks) ready in the FIFO is equal to or larger than another buffer threshold. This will have a tendency to minimize the number of tape start and stops (underrun conditions). This threshold is fixed and is not under user control.

In read mode there are also two buffer thresholds; one controlling the tape read operation and one controlling the SCSI-bus transfer. These thresholds work in much the same way as the threshold in write mode. The host buffer threshold is configurable while the tape buffer threshold is fixed

SC SI-bus Si de Data For matter Si de

Data in Buf ferData going out Data going in

Th reshold Detector

Figure 2-9: Host Buffer Data flow During Read Operations

Tandberg Data About Tape Streamers

2-27 Tandberg SLR Product Line SCSI Functional Specifications

Write Mode SCSIbus Threshold

This is the threshold used when moving data from the SCSI-bus into the data buffer in write mode. This threshold is controlled by the Write Buffer Empty Ratio. The Write Buffer Empty Ratio may affect the disconnect/reconnect frequency on the SCSI-bus during DATA OUT phases. See the Disconnect/Reconnect Page of the MODE SELECT command for details.

Read Mode SCSIbus Threshold

This is the threshold used when moving data from the data buffer to the SCSI-bus in read mode. This threshold is controlled by the Read Buffer Full Ratio. The Read Buffer Full Ratio may affect the disconnect/reconnect frequency on the SCSI-bus during DATA IN phases. See the Disconnect/Reconnect Page of the MODE SELECT command for details.

2.7.7. Disconnect/Re-connect

When the Drive is operating in write mode no data transfer will be initiated until the data buffer can accept the amount of data set up by the current write bus threshold. Instead the Drive disconnects from the SCSI-bus freeing the bus for use by other devices. As soon as the data buffer can accept the data, the Drive reconnects and transfers a burst of data. The size of the burst is upwards limited by either available buffer space or the Maximum Burst Size, whichever is smaller. The Maximum Burst Size is specified by the host by means of the Mode Select command. If still more data is requested for transfer, the Drive then disconnects again. A reconnect later will transfer another burst and the whole process repeats until all requested data have been transferred.

When the Drive is operating in read mode no data transfer will be initiated until the data buffer has ready the amount of data set up by the current read bus threshold. Instead the Drive disconnects waiting for enough data to become ready. When this happens the Drive reconnects and transfers a burst of data. The Drive will also reconnect if a tapemark is pending or if an exception has occurred. The size of the burst is limited by the Maximum Burst Size. If still more data is requested for transfer the Drive disconnects again. A reconnect later will transfer another burst and the whole process repeats until all requested data have been transferred.

Tandberg Data About Tape Streamers

2-28 Tandberg SLR Product Line SCSI Functional Specifications

2.7.8. Data Re-transfer

As a part of its bus parity error handling the Drive is able to re-transfer the last transferred burst any number of times.

When receiving data in write mode the Drive may optionally check for parity errors in the data received. When the complete data burst has been transferred the Drive changes from the data-out phase to the message-in phase. A RESTORE POINTERS message is then transferred to the Initiator. The Initiator should in response to this message reset its data pointer to where it was when the Drive instructed it to save its pointer the last time (with the SAVE DATA POINTER message sent each time the Drive disconnects) or if no such message has been received set the pointer back to where it was when the write type command was issued. The Drive then re-enters the data-out phase and transfers the last burst of data once more. The whole process can be repeated a number of times as long as the Drive detects parity errors in the received data.

When transferring data to the Initiator in read mode the Initiator may check for parity errors in the received data. When a parity error has been detected the Initiator may assert the ATN. After having transferred the complete data burst the Drive will honor the ATN condition by going from the data-in phase to the message-out phase. A message is then transferred from the Initiator. If this message is an INITIATOR DETECTED ERROR message the Drive will assume that the Initiator wants to have the last burst of data transferred once more. The Drive then goes to the message-in phase and transfers a RESTORE POINTERS message. The Initiator should then reset its data pointer to where it was when the Drive instructed it to save its pointer the last time (with the SAVE DATA POINTER message sent each time the Drive disconnects) or if no such message has been received set the pointer back to where it was when the read type command was issued. The Drive then re-enters the data-in phase and transfers the last burst of data once more. The whole process can be repeated a number of times as long as the Initiator asserts ATN during the data transfer.

See also the sub-section on Bus Parity Error Handling in section 7 for further information.

2.7.9. Buffer Parity Errors

The Drive is able to check for internal parity errors in its data buffer system. All data paths include a parity bit for every 8 bits of data. See the sub-section on Buffer Parity Error Handling in section 7 for details.

Tandberg Data About Tape Streamers

2-29 Tandberg SLR Product Line SCSI Functional Specifications

2.8. Data Compression

The Tandberg SLR Product Line Tape Drives support the Adaptive Lossless Data Compression (ALDC) algorithm. The ALDC is a variant of the LZ 1 (Lempel-Ziv 1) class of compression algorithms. For further details about the compression algorithm please refer to the description [8]. The actual compression and decompression are performed in hardware. For further details about the drive electronics, refer to the Tandberg SLR Product Line Reference Manual [1].

2.8.1. Background

Data compression can be used on all supported tape formats. When data compression is active, all data sent to the drive will be compressed (see section 2.8.2.2 for an exception). When compressed data is read the data is automatically decompressed. The Initiator may then issue a MODE SENSE command and check the Data Compression Parameter Page to investigate if the data on the tape is compressed or not.

When compression is enabled, the data will be compressed into groups of data called Compression Block Groups. Each of these groups will consist of an uncompressed Compression Header and compressed data. The data part is the result of compression of up to 64 KBytes (SLRtape24, SLR32 and SLRtape50 media) or 256 KBytes (SLRtape7, SLRtape40, SLRtape60, SLRtape75, SLRtape100 and SLRtape140 media) of original data [9] - [17].

8 Kbytes64 Kbytes 64 Kbytes64 Kbytes

Compressed data (uncompressed length given)

Compression Header

Compression Block Group

Original logical block with uncompressed length of 200 Kbytes, SLRtape24, SLRtape32 or SLRtape50

Figure 2-10: Packing Data in Compression Block Groups

The data compression ratio will depend on both the data pattern to compress and how the sequence of WRITE commands is issued to the Drive. Each start of a WRITE command with a block size different from the block size in the previous WRITE command will force a new Compression Block Group to be written, so a lot of WRITE commands transferring blocks of different sizes will result in a less than optimal compression ratio. Since the length of the Compression Header is 18 to 24 bytes large, the best performance is achieved by issuing WRITE commands writing large blocks or a large number of equal size blocks. Typical compression ratios are 2:1 to 3:1.

Tandberg Data About Tape Streamers

2-30 Tandberg SLR Product Line SCSI Functional Specifications

Note that in some cases data will actually expand when subjected to the compression process. This may for instance be the case when feeding already compressed data to the drive and the compression option is enabled.

When writing on a SLRtape50 media the drive will adaptively turn compression on and off. When a certain expansion has been detected the Drive decides that the data are not compressible and this piece of data will be written uncompressed. When data that is compressible again is presented to the drive (with a favorable compression ratio), compression is turned back on (if enabled at all).

When writing on SLRtape7, SLRtape40, SLRtape60, SLRtape75, SLRtape100 and SLRtape140 media the drive will keep compression on even if the data is not compressible. A worst case expansion of 12.5% may occur on certain data patterns.

When using SLRtape140 to SLRtape7 or SLR32, media any partition may hold a mix of compressed and uncompressed data.

When using an SLR5 or DC9250 medium the drive will not allow a mix of compressed and uncompressed data on the same medium. Before starting a write operation from BOM, compression may be turned on or off. At a later stage, it will not be possible to change the compression mode. When appending data on a tape, the drive will ensure that compression mode is preserved (on or off). This means that if the pre-written data on the medium is not compressed, the appended data will not be compressed (even if the DCE bit in the Data Compression Page has configured compression on). If, however, the pre-written data on the medium is compressed, the appended data will be compressed (even if compression has been configured off).

2.8.2. Controlling Data Compression

Data compression is controlled by fields in the Data Compression page of the MODE SENSE/SELECT page (page code 0Fh). This page is shown below. Fields of no interest are marked with “ XX” .

BYTE BIT 7 6 5 4 3 2 1 0

00 XX XX Page Code = 0Fh

01 XX

DCE DCC

DDE

XX XX



Compression Algorithm



Decompression Algorithm

12 XXXX

13 XXXX

14 XXXX

15 XXXX

Table 2-5: Data Compression Page Descriptor

Tandberg Data About Tape Streamers

2-31 Tandberg SLR Product Line SCSI Functional Specifications

The usage of the different fields varies slightly as a function of the tape format used, but the following general description is true for all tape formats.

The DCC (Data Compression Capable) bit is a status bit that indicates whether a device has compression/decompression capabilities. The MLR Series Drives are always compression/decompression capable. When transferring a Data Compression page with a MODE SELECT command this bit must be set to one (it is not changeable). When a MODE SENSE command is used to retrieve the Data Compression page the value returned is always one.

The DCE (Data Compression Enable) is a configuration bit used to turn data compression on or off. Some restrictions apply when using SLR5 and DC9250 media. See below for further details. The MODE SELECT command is used to set/reset this bit. When a MODE SENSE command is used to retrieve the Data Compression page the value returned reflects the value set by the last MODE SELECT command.

The DDE (Data Decompression Enable) bit is used in two different ways depending on the tape format in use. See the following sub-sections for further details.

The Compression Algorithm field is a configuration field used to select a specific data compression algorithm to be used when data are being compressed. The SLR Series Drives support two values; A value of 3h selects the ALDC algorithm. A value of 0h means “ no algorithm selected” . In this case the drive will not perform any data compression. This means that in order to have data compressed the DCE bit must be set to 1 and the Compression Algorithm field must be set to 3h. The MODE SELECT command is used to set a value in the Compression Algorithm field. When a MODE SENSE command is used to retrieve the Data Compression page the value returned in this field reflects the value set by the last MODE SELECT command.

The Decompression Algorithm field is a status field used by the drive to report which compression algorithm that was used when the data were written to a medium. The value in this field is valid only after this data have been read off the medium using a READ or VERIFY

command. A value of 3h indicates that the ALDC algorithm was used (the data on the medium is compressed). A value of 0h indicates that no algorithm was selected (the data on the medium is not compressed). A MODE SENSE command can be used to retrieve the Decompression Algorithm field. This field is ignored by the MODE SELECT command.

VERIFY is not supported by SLR7 and SLR140

Tandberg Data About Tape Streamers

2-32 Tandberg SLR Product Line SCSI Functional Specifications

2.8.2.1.

Using SLRtape140, SLRtape100, SLRtape75, SLRtape60, SLRtape50, SLRtape40, SLRtape7, SLR32 or SLRtape24 Media

When using SLRtape140 to SLRtape7 or SLR32 media it is possible to compress data on any partition. It is also possible to mix compressed and uncompressed data on the same partition. This means that the state of the DCE bit and the value of the Compression Algorithm field can be changed any time during write operations.

The DDE (Data Decompression Enable) bit is ignored when using these media. The Drive will, when transferring data to the Initiator, always decompress compressed data while uncompressed data are transferred without any decompression.

2.8.2.2. Using SLR5 or DC9250 Media

It is not possible to mix compressed and uncompressed data on a SLR5 or a DC9250 media.

None of the drives covered by this manual are able to write on these tapes. However, if the tapes are read, the drives will automatically enable data decompression if the tapes hold compressed data.

Tandberg Data About Tape Streamers

2-33 Tandberg SLR Product Line SCSI Functional Specifications

2.9. Optimizing Streaming Operation

For a tape streamer device underruns and overruns are typically time consuming. The stop, reposition and restart sequence may take a second or two. When this happens frequently, the average data transfer rate may suffer greatly.

In many cases using a combination of Automatic Velocity Control (AVC) and Forced Streaming can optimize streaming operation. AVC will automatically reduce tape speed if the transfer rate from host is weak or the compression ratio is high. Forced streaming will keep the drive streaming by writing dummy data while waiting for more data from host.

2.9.1. Forced Streaming

The Forced Streaming Count in the Mode Sense/Select Miscellaneous Parameters Page is used to enable or disable the Forced Streaming mechanism in the drive. The parameter is designed to give the maximum number of times to rewrite the same block while waiting for more data from host. When the maximum number is reached the drive is entering underrun mode and tape will stop and do a backspace while waiting for more data from host. The Forced Streaming Count is by nature a write parameter only. But, note that if forced streaming blocks are written to the tape, the read speed will reduce accordingly since the tape has to pass dummy data while reading.

The Forced Streaming feature trades performance for capacity. The number of underruns can be reduced, but some capacity will be lost to the rewrite process.

Note however that if the host system is able to write data at twice the speed of the native rate of the current media and the data can be compressed 2:1 or more, you will always have at least 2:1 compression no matter how many forced streaming blocks written.

On SLR140, SLR100, SLR75, SLR60 and SLR7 drives the Forced Streaming setting is ignored. Due to a large data buffer and short underrun time, there is no reason to use Forced Streaming.

2.9.2. AVC (Automatic Velocity Control)

The AVC bit in the Mode Sense/Select - Device Configuration page is used to enable or disable the AVC mechanism in the drive. When an underrun (during write) or an overrun (during read) occurs, the tape will stop and do a backspace while waiting for more data to be written or read from host. In this underrun/overrun situation the AVC algorithm will try to predict if a speed reduction is optimal or not. If a speed reduction is chosen, the AVC algorithm will, while streaming, try to predict if it is optimal to increase the speed again. A speed increase will perform a forced underrun/overrun.

Tandberg Data About Tape Streamers

2-34 Tandberg SLR Product Line SCSI Functional Specifications

2.9.3. Using both Forced Streaming and AVC

Using a combination of Forced Streaming and AVC has the following advantages while writing:

 The AVC algorithm does have one more parameter to use

when predicting optimal speed if the Forced Streaming is enabled (>0).

 The maximum “ average” Forced Streaming used is

reduced if the AVC algorithm is enabled. This results in a bit fewer forced streaming blocks written before underrun while at high speed, which results in less loss of capacity on the tape.

Note however that Forced Streaming and AVC are not available on all drives or medium types. For some drive and medium types one or both settings are ignored by the drive. See table 2-6 for details.

Drive Medium Forced Streaming AVC

SLR140

SLRtape140

Ignored

Available (2 speeds)

SLRtape100

Ignored

SLRtape75

Ignored

SLRtape60

Ignored

SLRtape50

Ignored

SLRtape40

Ignored

SLRtape7

Ignored

SLR100

SLRtape100

Ignored

Available (3 speeds)

SLRtape75

Ignored

Available (3 speeds)

SLRtape60

Ignored

Available (3 speeds)

SLRtape50

Ignored

SLRtape40

Ignored

Available (3 speeds)

SLRtape7

Ignored

SLR32

Ignored

SLRtape24

Ignored

SLR75

SLRtape75

Ignored

Available (2 speeds)

SLRtape60

Ignored

Available (2 speeds)

SLRtape50

Ignored

SLRtape40

Ignored

Available (2 speeds)

SLRtape7

Ignored

SLR32

Ignored

SLRtape24

Ignored

SLR5

Ignored

DC9250

Ignored

SLR60

SLRtape75

Ignored

Available (2 speeds)

SLRtape60

Ignored

Available (2 speeds)

SLRtape50

Ignored

SLRtape40

Ignored

Available (2 speeds)

SLRtape7

Ignored

SLR32

Ignored

SLRtape24

Ignored

SLR5

Ignored

DC9250

Ignored

Table 2-6: Forced Streaming and AVC On Different Medium Types (to

be continued…)

Tandberg Data About Tape Streamers

2-35 Tandberg SLR Product Line SCSI Functional Specifications

Drive

Medium

Forced Streaming

AVC

SLR50

SLRtape50

Available

Available (2 speeds)

SLR32

Available

Available (2 speeds)

SLRtape24

Available

Available (2 speeds)

SLR5

Ignored

DC9250

Ignored

SLR7

SLRtape7

Ignored

SLR5

Ignored

Table 2-6: Forced Streaming and AVC On Different Medium Types

2.9.4. Recommended Settings

SLR140, SLR100, SLR75, SLR60 and SLR7 Drives

 Forced Streaming = 0

 AVC = on

SLR50 Drives

 Forced Streaming = 1024

 AVC = on

The optimal Forced Streaming value is dependent on both host behavior, host transfer rate and compression ratio on data actually written. Fine-tuning should be done on “ real life systems” with “ real life data” .

2.10. Recorded Objects

The recording formats [9] - [14] specify that recorded elements (blocks and tapemarks) have identifiers included in the recorded information to help determining the write sequence and also to help detecting positioning errors. These identifiers are unique within the whole volume.

The identifiers are associated with physical blocks only. This means that a possible logical block number maintained by the host system will not be the same as the physical identifier of the same logical block because logical blocks may span over multiple physical blocks.

The physical identifiers are normally not visible to the host system. There are, however, some exceptions. In the Tandberg SLR Product Line Tape Drives, the READ POSITION and LOCATE commands may transfer physical block identifiers to/from the host system. The host system is, however, not expected to process this data in any way. The identifiers transferred from the Drive by the READ POSITION command should just be stored and then transferred unmodified back to the Drive at some future time by the LOCATE command.

Tandberg Data About Tape Streamers

2-36 Tandberg SLR Product Line SCSI Functional Specifications

2.11. TapeAlert

The TapeAlert information is accessed via LOG SENSE page 2Eh (see section 13.3.5 TapeAlert Page), and configuration is done via MODE SELECT page 1Ch (see section 14.3.9 Informational Exceptions Control Page). The Mode Sense/Select configuration of the TapeAlert interface is compatible with the SMART diagnostic standard for disc drives. The host software should first check the tape drive to determine whether it supports the TapeAlert Log Sense page 2Eh. By default the host software access to the TapeAlert Log Sense page is via polling.

The TapeAlert Log Sense page shall be read from a tape drive device at the following times as a minimum:



At the beginning of a write/read job, even if media is not loaded.



Immediately after a fatal error during the write/read job.



At the end of each tape when the write/read job spans multiple tapes. If the tape is to be ejected then the Log Sense page must be read BEFORE this.



At the end of a write/read job.

Though not mandatory, the host software may also poll the Log Sense page at regular intervals (e.g. every 60 seconds) while the tape drive is idle.

The TapeAlert Log page contains 64 one-byte alert flags. The specific conditions for any one flag to be set and cleared are defined in section 13.3.5.

There are three flag severity levels, listed below in order of severity:



Critical (C)



Warning (W)



Information (I)

Each time the host software reads the TapeAlert Log page, it should check all 64 flags to discover which are set (there may be more than one). The definitions of the 64 flags are device type specific, so that there is one definition for tape drive devices, and a different definition for stand-alone changer devices (in libraries). For each flag that is set, the host software shall communicate the defined error message and severity for that flag to the user, and also log it. If multiple flags are set simultaneously, they will be displayed together in ascending order of severity. At the beginning of each set of TapeAlert error messages, the tape device that initiated them must be identified. For the tape drive media-related flags (flags 4, 7 and 14) the software label of the media, if such exist, should be included in the TapeAlert error messages so that the user is aware what piece of media the error refers to. Such information could also be displayed with the messages for other flags as well if required. The information read in the TapeAlert flags should not in itself cause the software to stop a current backup/restore job.

3-1 Tandberg SLR Product Line SCSI Functional Specifications

Logical Characteristics

3.1. SCSI-bus Phases

The Drive implements the following SCSI-bus phases:

Bus management phases Information transfer phases

BUS FREE ARBITRATION SELECTION RESELECTION

COMMAND DATA-IN DATA-OUT STATUS MESSAGE-IN MESSAGE-OUT

Table 3-1: SCSI-bus Phases

The SCSI-bus will never be allowed to be in more than one phase at any given time. See also [2], [5] and [6]. See also [1] for timing diagrams showing SCSI-bus signal activity.

3.1.1. Bus Management Phases

3.1.1.1. Bus Free Phase

The BUS FREE phase is used to indicate that no SCSI device is actively using the SCSI-bus and that it is available for subsequent users.

The Drive enters the BUS FREE phase by releasing BSY after one of the following conditions:

 after a SCSI-bus RESET condition  after an ABORT message has been received  after a BUS DEVICE RESET message has been received  after a DISCONNECT message has been transmitted  after a COMMAND COMPLETE message has been transmitted  after certain SCSI protocol errors (see section 6.6

Unexpected Bus Free )

The Drive enters the BUS FREE phase by releasing SEL after the following condition:

 after an unsuccessful reselection of an Initiator

3.1.1.2. Arbitration Phase

The Arbitration phase allows the drive to gain control of the SCSI-bus so that it can resume or initiate the connection with an other device on the bus.

Tandberg Data Logical Characteristics

3-2 Tandberg SLR Product Line SCSI Functional Specifications

3.1.1.3. Selection Phase

The Selection phase allows the Initiator to select the drive for the purpose of having the Drive execute a command or physical path management function.

3.1.1.4. Reselection Phase

The Reselection phase allows the Drive to reconnect to an initiator for the purpose of continuing some operation that was previously started by that initiator, but was suspended by the Drive (e.g. it disconnected).

When attempting to reselect its Initiator, the Drive implements the recommended Reselection Time-out Delay according to ANSI [2] and [6].

If Reselection does not succeed, the Drive considers the current command for terminated (as if an ABORT message had been received). Buffered data will continue to be written to the tape if the time-out occurs during a write operation. Immediate commands will continue their execution. No sense data error information will be generated. A new command from the same or any other Initiator will execute normally (as if no time-out had occurred).

3.1.2. Information Transfer Phases

3.1.2.1. Command Phase

The COMMAND phase is used by the Drive to request command information from the Initiator.

The Drive transfers either six (Group 0 commands) or ten (Group 1 commands) command bytes in one single Command Phase.

3.1.2.2. Data In/Out Phases

The DATA-IN phase is used by the Drive to request that data is sent from the Drive to the Initiator.

The DATA-OUT phase is used by the Drive to request that data is sent from the Initiator to the Drive.

Synchronous and asynchronous Data Transfer are supported. 8 bit and 16 bit wide SCSI-busses are supported.

3.1.2.3. Status Phase

The STATUS phase is used by the Drive to request that status information is sent from the Drive to the Initiator.

Tandberg Data Logical Characteristics

3-3 Tandberg SLR Product Line SCSI Functional Specifications

3.1.2.4. Message-In/Out Phases

The MESSAGE-IN phase is used by the Drive to request message(s) be sent from the Drive to the Initiator.

The MESSAGE-OUT phase is used by the Drive to request message(s) be sent from the Initiator to the Drive.

3.2. SCSI-bus Phase Sequences

The order in which phases are used on the SCSI-bus follows a prescribed sequence. The RESET condition however can abort any phase and is always followed by the BUS FREE phase.

The sequences allowed by the SCSI standard are shown in the figure below. The normal progression is:

 From the BUS FREE phase to ARBITRATION

 From ARBITRATION to SELECTION or RESELECTION

and

 From SELECTION or RESELECTION to one or more of the

information transfer phases (COMMAND, DATA-IN/OUT, STATUS, or MESSAGE-IN/OUT)

BUS FREE ARBITRATION

RESELECTION

SELECTION

DATA IN

DATA OUT

STATUS

COMMAND

MESSAGE OUT

MESSAGE IN

Reset or protocol error

Figure 3-1: Phase Sequencing

Tandberg Data Logical Characteristics

3-4 Tandberg SLR Product Line SCSI Functional Specifications

Arbitration and Selection are always controlled by the Initiator. Once the Target is selected however, the Target controls the phase sequencing between the information transfer phases. If the Initiator wants to interrupt by sending a message to the Target, it may assert ATN.

3.2.1. Legal Phase Sequences

The sequences depend on the command(s) and the use of the ATN condition by the Initiators. The different command sections (sections 8 through 31) detail the possible sequences of each command.

3.2.1.1. Notation

The notation used in the sequences is called Backus-Naur Form or BNF. The elements of the notation used in this

manual are as follows:

::=

Rewriting rule or production. This symbol is read as ‘is defined as’ or ‘can be replaced by’.

Or. This symbol separates alternative definitions.

< >

Angle brackets are used to delimit the name of a defined sequence, e.g. <sequence>. Where it is defined, the name of a sequence is set in boldface.

{ }

Braces are used to denote a possible repetition of the enclosed symbols – one ore more times.

CAPS

The names of the different SCSI-bus phases are set in capital letters.

3.2.1.2. Sequence Elements

The sub-sequences or sequence elements shown below will be used in the different command sections to describe the actual sequences that are possible for each command.

<initiator-part> ::= BUS-FREE ARBITRATION SELECTION

The Initiator initiates a new connection by first waiting for the BUS FREE phase, then by arbitrating for the bus and – when the bus has been won – finally by selecting the Drive.

<message-out> ::= MESSAGE-OUT |

MESSAGE-OUT <message-out> |

MESSAGE-OUT MESSAGE-IN <message-out> |

MESSAGE-OUT BUS-FREE

Tandberg Data Logical Characteristics

3-5 Tandberg SLR Product Line SCSI Functional Specifications

A message-out sequence may transfer a single message or multiple messages. The message-out sequence may also include some message-in phases (as may be the case when the Drive rejects a message out). Finally the messageout sequence may end up in a BUS FREE phase (as it will after messages like ABORT or RESET).

<message-in> ::= MESSAGE-IN |

MESSAGE-IN <message-out>

A message-in sequence may transfer a message to the Initiator. After the message-in transfer a message-out may also occur (as may be the case when the Initiator rejects or aborts a message-in).

Note that the SCSI-2 and SCSI-3 specifications both call for a message out phase transferring an IDENTIFY message [2], [5].

<command> ::= COMMAND |

COMMAND <message-out>

The command sequence transfers a complete CDB (6 or 10 bytes). A message out sequence will follow if the Initiator asserts ATN during the command transfer.

<data-in> ::= DATA-IN |

DATA-IN <message-out>

The data in sequence transfers one or more data bytes from the Drive to the Initiator. A message out sequence will follow if the Initiator asserts ATN during the data transfer.

<data-out> ::= DATA-OUT |

DATA-OUT <message-out>

The data out sequence transfers one or more data bytes from the Initiator to the Drive. A message out sequence will follow if the Initiator asserts ATN during the data transfer.

<status> ::= STATUS |

STATUS <message-out>

The status sequence transfers a single status byte to the Initiator. A message out sequence will follow if the Initiator asserts ATN during the status transfer.

<completed> ::= <status> <message-in> BUS-FREE

When an unlinked command has completed execution, the Drive transfers a status byte and a message byte and goes to BUS FREE.

<completed-link> ::= <status> <message-in>

Tandberg Data Logical Characteristics

3-6 Tandberg SLR Product Line SCSI Functional Specifications

When a linked command has completed execution, the Drive transfers a status byte and a message byte. The bus is not going to BUS FREE.

Tandberg Data Logical Characteristics

3-7 Tandberg SLR Product Line SCSI Functional Specifications

<disconnect> ::= < message-in> BUS-FREE |

<message-in> < message-in> BUS-FREE

When the Drive disconnects it may send one or two message in bytes before going to the BUS FREE phase. When transferring a single message, this will be a DISCONNECT message. When transferring two messages, the first message will be a SAVE DATA POINTER message. The second message will be a DISCONNECT message. The Drive sends a single message when it disconnects after having transferred a CDB. It sends two messages in all other cases.

<reconnect> ::= BUS-FREE ARBITRATION RESELECTION <message-in>

The Drive reconnects to the Initiator by waiting for BUS FREE, arbitrating for the bus and then by performing the reselection. A single message in byte follows. This will be an IDENTIFY message with the Disconnect Privilege bit set to zero and with a zero LUN field.

3.2.1.3. Disconnects/Reconnects

Disconnection is the process of going through one or two MESSAGE-OUT phases and then to the BUS FREE phase in the middle of a command execution. When the Drive has disconnected, it will always try to reconnect at a later time. Reconnection is the process of going from the BUS FREE phase to the ARBITRATION phase, RESELECTION and MESSAGE-IN phase.

When disconnecting, the first of possibly two messages transferred will be a SAVE DATA POINTER message. The second or only message will be a DISCONNECT message. The Drive will then enter the BUS FREE phase.

When reconnecting, the message transferred will be an IDENTIFY message with the Disconnect Granted bit (Bit 6) set to zero. The six LUN bits will also be set to zero.

All commands except INQUIRY, REQUEST SENSE and TEST UNIT READY will disconnect if the Initiator grants disconnection. The first disconnect will always occur immediately after the CDB transfer.

Please refer to the section for each command to see if and when during the execution the command disconnects/reconnects.

Tandberg Data Logical Characteristics

3-8 Tandberg SLR Product Line SCSI Functional Specifications

3.2.1.4. Command Linking

Normally the Drive goes to the BUS FREE phase after a successful command completion. The Drive transfers the GOOD status byte and the COMMAND COMPLETED message before entering the BUS FREE phase.

If the initiator wants the Drive to go directly to a new COMMAND phase after successful command completion, the initiator has to link the commands. The Drive than transfers an INTERMEDIATE status byte followed by a LINKED COMMAND COMPLETE (or LINKED COMMAND COMPLETE W/FLAG) message byte before entering the COMMAND phase. The Drive expects immediately the transfer of a new command.

To link a command the Link bit has to be set to one in the Control Byte of the Command Descriptor Block (refer to section 4.2).

The command link function will operate as long as commands complete successfully and the link bit is set. When some error has been detected (any other than GOOD status would have been transferred for a non-linked command), the link will be broken, and the failing command will be terminated with the proper status byte and then with a COMMAND COMPLETE message byte.

Command linking modifies the command phase sequencing: The first command in a series of linked commands will

follow the prescribed phase sequence except that the <completed-link> sub-sequence will be used instead of the usual <completed> sub-sequence.

The second and every other command except the last will skip the <initiator-part> sub-sequence and go directly to the <command> sub-sequence. A <completed-link> subsequence will be used instead of the usual <completed> sub-sequence.

The last command in a series of linked commands will skip the <initiator-part> sub-sequence and go directly to the <command> sub-sequence.

Here is a simple example with three commands. The first two have their Link bits set to one.

<sequence-link> ::= <initiator-part> <command-1> <completed-link> <command-2>

<completed-link> <command-3> <completed>

Tandberg Data Logical Characteristics

3-9 Tandberg SLR Product Line SCSI Functional Specifications

3.3. SCSI Pointers

The SCSI architecture provides for two sets of three pointers within each Initiator. The first set of pointers is known as the current (or active) pointers. The second set of pointers is known as the saved pointers.

There are three kind of pointers in a pointer set. There are pointers pointing to a command, data or status area in the memory of the Initiator.

The pointers in the current pointer set are used to represent the state of the interface and point to the command, data and status byte to be transferred between the Initiator memory area and the Target for the current connection. There is only one set of current pointers in each Initiator. The current pointers are valid for the Target currently connected to the Initiator.

The pointers in the saved pointer sets represent the state of each command that the Initiator has active (whether or not it is currently connected). If the Initiator for instance has four different commands active in one or more targets (including the currently connected command), there will be four complete sets of saved pointers stored in the Initiator.

The saved command pointer always points to the start of the Command Descriptor Block for the current command.

At the beginning of each command, the saved data pointer points to the start of the data area. It remains at this value until the Target sends a SAVE DATA POINTER message to the Initiator. In response to this message, the Initiator stores the value of the current data pointer into the saved data pointer.

The saved status pointer always points to the start of the status area for the current command.

Tandberg Data Logical Characteristics

3-10 Tandberg SLR Product Line SCSI Functional Specifications

Data Buffer

Command Buffer

Status Buffer

Data Buffer

Command Buffer

Status Buffer

Command Pointer

Data Poi nter

Status PointerCurrent Status Pointer

Current D ata Pointe r

Current Co mman d Pointer

SCSI

BUS

Saved Comm and Pointers for all active command s:

Initiator Target

Saved Data Pointers for all active commands:

Saved Status Pointers for all active comman ds:

Figure 3-2: SCSI Pointers

The Target may restore the current pointers to their saved values by sending a RESTORE POINTERS message to the Initiator. The Initiator moves the saved value of each pointer into the corresponding current pointer.

Whenever a SCSI device disconnects from the bus, only the saved pointer values are retained. The current pointer values are restored from the saved values automatically upon the next reconnection.

Tandberg Data Logical Characteristics

3-11 Tandberg SLR Product Line SCSI Functional Specifications

3.4. Unit Attention

The Unit Attention Condition is created for each initiator whenever

 A new cartridge has been inserted

 LOG SELECT parameters have been changed by another

Initiator

 MODE SELECT parameters have been changed by another

Initiator

 The Drive microcode (ucode) has been changed

 A hard reset has occurred (either through the SCSI-

bus RST line or through processing of a BUS DEVICE RESET message)

 A power-on reset has occurred

 A LOG threshold condition has been met.

The Unit Attention Condition will persist for each Initiator until that Initiator clears the condition as described in the following paragraphs.

If an Initiator issues a command other than INQUIRY or REQUEST SENSE while a Unit Attention condition exists for that Initiator (prior to reporting CHECK CONDITION for the Unit Attention condition), the Drive will not perform the command and will report CHECK CONDITION status unless a BUSY or RESERVATION CONFLICT status (higher priority status) is also pending.

After reporting CHECK CONDITION status, the next command received from that Initiator should be REQUEST SENSE. Otherwise the Unit Attention condition will be cleared for that Initiator, the Drive will perform the command and report GOOD status.

If a REQUEST SENSE command is received after the Drive has reported CHECK CONDITION status to this Initiator for Unit Attention condition, the Drive will report UNIT ATTENTION Sense Key, and then clear the Unit Attention condition for that Initiator.

If a REQUEST SENSE command is received from an Initiator with a pending Unit Attention condition (before the Drive reports CHECK CONDITION), the Drive will discard any pending sense data, report UNIT ATTENTION Sense Key, and then clear the Unit Attention condition for that Initiator.

If an INQUIRY command is received with a pending Unit Attention condition (before the Drive reports CHECK CONDITION status), the Drive will perform the INQUIRY command, report GOOD status, and will not clear the Unit Attention condition. If the INQUIRY command is received after the Drive has reported CHECK CONDITION for a pending Unit Attention condition, then the Unit Attention condition will be cleared and the Drive performs the INQUIRY command.

Tandberg Data Logical Characteristics

3-12 Tandberg SLR Product Line SCSI Functional Specifications

3.5. SCSI-bus Conditions

3.5.1. Attention (ATN)

The ATTENTION condition allows an Initiator to inform a Target that the Initiator has a message ready. The Drive may get this message by performing a MESSAGE-OUT phase.

The Initiator creates the ATTENTION condition by asserting ATN at any time except during the ARBITRATION or BUS FREE phases.

The Initiator must assert the ATN signal before releasing ACK for the last byte transferred in a bus phase for the ATTENTION condition to be honored before transition to a new bus phase. An ATN asserted later might not be honored until a later bus phase and then may not result in the expected action.

The Initiator must keep ATN asserted if more than one message byte is to be transferred.

The Initiator may negate the ATN signal at any time except while the ACK signal is asserted during a MESSAGE-OUT phase. Normally, the Initiator negates ATN while REQ is true and ACK is false during the last REQ/ACK handshake of the MESSAGE-OUT phase.

If ATN occurs during a COMMAND phase, MESSAGE-OUT will occur after transfer of all Command Descriptor Block bytes has been completed.

If ATN occurs during a DATA IN or OUT phase while transferring parameters of a:

- INQUIRY

- LOG SELECT

- LOG SENSE

- MODE SELECT

- MODE SENSE

- READ BLOCK LIMITS

- READ POSITION

- RECEIVE DIAGNOSTIC RESULTS

- REQUEST SENSE

- SEND DIAGNOSTICS

command,

MESSAGE-OUT will occur after transfer of the complete Parameter List.

Tandberg Data Logical Characteristics

3-13 Tandberg SLR Product Line SCSI Functional Specifications

If ATN occurs during a DATA phase of a:

-READ

- READ BUFFER

- VERIFY

-WRITE

- WRITE BUFFER

command,

MESSAGE-OUT will occur after transfer of no more data than given by the Maximum Burst Size parameter in the Disconnect/Reconnect Page of the MODE SELECT command.

If ATN occurs during a STATUS phase, MESSAGE-OUT will occur after the status byte has been acknowledged by the Initiator.

If ATN occurs during a MESSAGE IN phase, MESSAGE-OUT will occur after the message byte has been acknowledged by the Initiator.

If ATN occurs during a SELECTION phase and before the Initiator releases the BSY signal, MESSAGE-OUT will occur immediately after the SELECTION phase.

If ATN occurs during a RESELECTION phase, MESSAGE-OUT will occur after the Drive has successfully sent it’s IDENTIFY message for that RESELECTION phase.

3.5.2. Reset

The RESET condition is used to immediately clear all SCSI devices from the bus. This condition will take precedence over all other phases and conditions. Regardless of prior bus phase, the bus enters the bus free phase. Any SCSI device may create the reset condition by asserting the RST signal.

The Drive will never create a RESET condition. The Drive is reset under the following conditions:

 At Power-on

 When RST is asserted by one of the devices on the

SCSI-bus

 When a BUS DEVICE RESET message is sent to the drive

 When the Drive microcode (ucode) is updated

 Internal watchdog function detecting uncorrectable

hardware or software errors

The Drive only supports the SCSI Hard Reset option.

Tandberg Data Logical Characteristics

3-14 Tandberg SLR Product Line SCSI Functional Specifications

3.5.2.1. Hard Reset

When the Drive detects a hard reset condition it will do the following:

 Release all SCSI-bus signals and go to BUS FREE.

 Clear all non-completed commands

 Release all SCSI-device reservations

 Return all SCSI device operating modes to their saved

values (MODE SELECT) or default values (PREVENT/ALLOW MEDIUM REMOVAL). The non-savable Log Parameters will be cleared to zero. The savable Log Parameters will remain unchanged, unless the Hard Reset is a result of a Power-up condition. In this case the savable Log Parameters will be initialized to their saved values.

 Position the media to BOM - if the drive has a

cartridge inserted when the reset condition occurred.

 UNIT ATTENTION condition will be set for all

Initiators (see Section 3.4. Unit Attention)

Tandberg Data Logical Characteristics

3-15 Tandberg SLR Product Line SCSI Functional Specifications

This Page Intentionally Left Blank

4-1 Tandberg SLR Product Line SCSI Functional Specifications

Commands

4.1. The Command Descriptor Block

A request to the Drive is performed by sending a Command Descriptor Block. For some commands the request is accompanied by a list of parameters sent during the DATAOUT phase.

The Drive will support Group 0 and Group 1 commands [2], [3], [4], [7]. Group 0 commands have 6 bytes in the Command Descriptor Block. Group 1 commands have 10 bytes in the Command Descriptor Block.

Examples of Six-Byte and Ten-Byte Command Descriptor Blocks are shown in the two following tables:

BYTE BIT 7 6 5 4 3 2 1 0

00 Operation Code

01 Logical Unit Number (LUN) Parameters

02 Parameters

03 Parameters

04 Parameters

05 Control Byte

Table 4-1: Typical Six-byte Command Descriptor Block

BYTE BIT 7 6 5 4 3 2 1 0

00 Operation Code

01 Logical Unit Number (LUN) Parameters

02 Parameters

03 Parameters

04 Parameters

05 Parameters

06 Parameters

07 Parameters

08 Parameters

09 Control Byte

Table 4-2: Typical Ten-byte Command Descriptor Block

Operation Code

The Command Descriptor Block always has an operation code as the first byte of the command. See Section 4.4. for operation codes supported by the Drive.

Tandberg Data Commands

4-2 Tandberg SLR Product Line SCSI Functional Specifications

LUN

The Drive supports only one Logical Unit Number (LUN). This field should/must be set to zero in all Command Descriptor Blocks for the Drive.

The Drive is a SCSI-1/2 compliant device. The Drive will ignore the LUN specified in the CDB if an IDENTIFY message was received. It is recommended that the LUN in the CDB be set to zero. See section 4.6.3. Commands Received Serially, LUN Not 0 for a description of the response to non-zero LUN values.

Parameters

See the specific commands for detailed information on the various parameter bytes.

Control Byte

See Section 4.2. for details on the Command Control Byte.

4.2. Command Control Byte

The Control Byte is the last byte of every Command Descriptor Block. A typical Control Byte is shown below.

BYTE BIT 7 6 5 4 3 2 1 0

Last Byte

X X RESERVED NACA Flag Link

Table 4-3: The Command Control Byte

Bits 6 and 7 are ignored by the Drive. The only exception is the SPACE command which uses control byte bit 7 as its FAST option bit.

RESERVED

These bits MUST always be set to zero.

NACA

The Normal ACA (NACA) bit must be set to zero.

Link

If the Link bit is set, an automatic link to the next command is made at a successful completion of the current command. The Drive first returns INTERMEDIATE status for the current command (instead of GOOD status), it then returns one of two linked command complete messages (as described under Flag below) before going directly to another Command Phase (there is no Bus Free phase in between commands).

When the Link bit and the IMMediate bit of a command (where applicable) are both set to one, the Drive returns CHECK CONDITION status. The Error Code is set to I

NVALID FIELD IN

CDB.

Flag

The Flag bit is only meaningful when the Link bit is set to one. If both bits are set, a LINKED COMMAND COMPLETE WITH FLAG message byte is sent after the INTERMEDIATE status byte. If only the Link bit is set, a LINKED COMMAND COMPLETE message byte is sent after the INTERMEDIATE status byte.

If only the Flag bit is set (the Link bit is zero), the Drive returns CHECK CONDITION status. The Error Code is set to I

NVALID FIELD IN

CDB.

Tandberg Data Commands

4-3 Tandberg SLR Product Line SCSI Functional Specifications

4.3. Reserved Fields

Reserved bits, fields, bytes and code values are set aside for future standardization. These bits, fields or bytes have to be set to zero. They are marked with the word RESERVED or the letter R in the Command Descriptor Blocks and Parameter Lists. If the Drive receives a reserved bit, field, byte that is not zero or receives a reserved code value, it will terminate the command with a CHECK CONDITION status.

If the offending bit, field, byte or code is located in a Command Descriptor Block, then the whole Command Descriptor Block (6 or 10 bytes) will be transferred before the command is terminated with CHECK CONDITION and the Error Code will be set to I

NVALID FIELD IN

CDB.

If the offending bit, field, byte or code is located in a Parameter List, then the whole Parameter List will be transferred before the command is terminated with CHECK CONDITION and the Error Code will be set to I

NVALID FIELD IN

PARAMETER LIST

Tandberg Data Commands

4-4 Tandberg SLR Product Line SCSI Functional Specifications

4.4. Command Set Summary

The Tandberg SLR Product Line Tape Drives supports the SCSI commands listed in the following table.

Description Group Media Type Operation Code

ERASE INQUIRY LOAD/UNLOAD LOCATE LOG SELECT LOG SENSE MODE SELECT MODE SENSE PREVENT/ALLOW MEDIA REMOVAL READ READ BLOCK LIMITS READ BUFFER READ POSITION RECEIVE DIAGNOSTIC RESULTS RELEASE REQUEST SENSE RESERVE REWIND SEND DIAGNOSTICS SPACE TEST UNIT READY VERIFY

WRITE WRITE BUFFER WRITE FILEMARKS

M N

Yes

Yes Yes

Yes

Yes Yes

Yes

O O O O M M O M M O O O M

M M M M M O M O M

19h 12h 1Bh

2Bh 4Ch 4Dh

15h 1Ah 1Eh

08h

05h 3Ch

34h 1Ch

17h

03h

16h

01h 1Dh

11h

00h

13h 0Ah 3Bh

10h

Table 4-4: SCSI Command Set

The Command Group is used to specify legal (and illegal) command sequences (see also Section 4.5).

Not supported by the Tandberg SLR7 and SLR140

The MODE SELECT command is a media access command only when the Device Configuration Page with the CAP bit set to one or when the Medium Partition Page(1) are included in the parameter list.

When executing a Selftest 1, the SEND DIAGNOSTICS is not a media access

command. When executing Selftest 2 the SEND DIAGNOSTICS is a media access command.

Tandberg Data Commands

4-5 Tandberg SLR Product Line SCSI Functional Specifications

Group M

These are Move Type commands.

Group R

These are Read Type commands. Data is read off the tape during command execution.

Group N

These are Navigate Type commands. Data is read off the tape during command execution, but no data transfer takes place on the SCSI-bus.

Group W

These are Write Type commands. Data is written to the tape during command execution.

All other commands

These are Neutral commands.

Commands marked with “ Yes” in the Media column are called Media Access Commands. The Drive will terminate the command with CHECK CONDITION status if any media access command is issued with no cartridge loaded. A cartridge is loaded when it is inserted and the Auto Load option is enabled or a LOAD/UNLOAD command has been executed with the Load bit set to one.

The Command Type is defined by the SCSI-standards [2] [3] [7] like this:

Type M

These commands must be implemented by a sequential access device in order to meet the minimum requirements.

Type O

These commands are optional for sequential devices.

Tandberg Data Commands

4-6 Tandberg SLR Product Line SCSI Functional Specifications

4.5. Command Sequencing

Usually the Initiator must issue a sequence of SCSI commands to be able to have the Drive perform a certain operation. As a general rule any sequence of SCSI commands is legal. There are, however, a few exceptions.

The tape and buffer system in the Drive can be in one of 4 different normal modes. In addition there are some exception modes. The Drive changes modes before starting to execute certain commands. The Drive may enter one of the exception modes when a command has failed. The action taken by the Drive when a command is received depends on the current mode. For the discussion of modes the command set is grouped into 5 different command groups according to the command mode specified in Section 4.4.

4.5.1. Normal Modes

There are 4 different normal modes. The default mode after power-up or reset is always MOVE. The Drive also enters MOVE mode when a new cartridge is inserted.

MOVE

The Drive attempts to enter the MOVE mode when a command from the move-group has been received. In this mode the data buffer is not used.

READ

The Drive attempts to enter the READ mode when a command from the read-group has been received. The data buffer system is set up to transfer data and tapemarks from the tape to the SCSI-bus.

NAVIGATE

The Drive attempts to enter the NAVIGATE mode when a command from the navigate-group has been received. The data buffer system is set up to transfer data and tapemarks from the tape.

WRITE

The Drive attempts to enter the WRITE mode when a command from the write-group has been received. The data buffer system is set up to transfer data (WRITE commands) or tapemarks (WRITE FILEMARKS commands) from the SCSI-bus to the tape.

The action taken by the Drive when a command is received depends on the current mode. Commands from one group can always follow a command from the same group with no special action taken. Note also that neutral-group commands can be inserted into any sequence of commands as they do not change the Drive's mode. When a command from one group follows a command from another group the Drive usually takes special action. In a few cases going from one group to another is not allowed. The command from the new group is then not executed at all. Instead it is just terminated with CHECK CONDITION.

The following table shows the actions taken by the Drive when a command from a certain group is received in the different normal modes:

Tandberg Data Commands

4-7 Tandberg SLR Product Line SCSI Functional Specifications

Current Mode Next Command Actions

MOVE move-group No action. The move-group command is executed.

read-group The buffer system is re-initialized (all buffered data is lost). The

Drive then calibrates the head position. The Density Code is updated. The read-group command is executed.

navigate-group

The buffer system is re-initialized (all buffered data is lost). The Drive then calibrates the head position. The Density Code is updated. The navigate-group command is executed.

write-group The Drive waits until the cartridge type is known. If the cartridge type

is suited for the selected tape format (Density Code) the buffer system is re-initialized (all buffered data is lost). If the tape format requires so the Drive then seeks the tape edge and writes the tape reference burst. The Density Code is updated. The write-group command is executed.

READ move-group The Drive stops any read-ahead operation. The buffer system is re-

initialized (all read-ahead data is lost). The move-group command is then executed.

read-group No action. The read-group command is executed.

navigate-group

The Drive just enters NAVIGATE mode. The navigate-group command is executed.

write-group

Since overwrite is illegal, the Drive will check if the tape is logically positioned at end-of-data (EOD). If the tape is not at EOD then the write-group command is terminated immediately with CHECK CONDITION status and the Error Code is set to W

RITE APPEND

ERROR

(Write After Read). The Drive continues to be in READ mode. If the tape is at EOD, or if overwrite is legal, the write-group command is executed. New data will be appended, beginning from the current position.

NAVIGATE move-group

The Drive stops any read-ahead operation. The buffer system is reinitialized (all read-ahead data is lost). The move-group command is then executed.

read-group The Drive just enters READ mode. The read-group command is

executed.

navigate-group No action. The navigate-group command is executed.

write-group The Drive will check if the tape is logically positioned at end-of-data

(Logical End Of Partition). If the tape is at EOD then the write-group command is executed. This will append new data after the last written block on the tape. If the tape is not at EOD then the write-group command is terminated immediately with CHECK CONDITION status and the Error Code is set to W

RITE APPEND ERROR

(Write After

Read). The Drive continues to be in NAVIGATE mode.

Table 4-5: Normal Mode Actions (to be continued...)

Tandberg Data Commands

4-8 Tandberg SLR Product Line SCSI Functional Specifications

Current Mode Next Command Actions

WRITE move-group Data and tapemarks remaining in the data buffer are written to the

tape. If this operation is successful the Drive enters the MOVE mode. The move- group command is executed.

read-group The Drive terminates the read-group command immediately with

CHECK CONDITION status. The Error Code is set to either E

ND-OF

DATA DETECTED

or END-OF-

DATA DETECTED AFTER

LEW depending on whether the tape is positioned before or after LEW. The Drive continues to be in WRITE mode.

navigate-group

If the navigate-group command is a SPACE forward command, the Drive terminates the command immediately with CHECK CONDITION status. The Error Code is set to either E

ND-OF-DATA

DETECTED

or END-OF-

DATA DETECTED AFTER

LEW depending on whether the tape is positioned before or after LEW. The Drive continues to be in WRITE mode. In other cases the Drive enters NAVIGATE mode and the navigate-group command is executed.

write-group A test is made to see if the current Density Code (tape format) has

changed. The write-group command is executed if the tape format has not changed.

Table 4-5: Normal Mode Actions

Tandberg Data Commands

4-9 Tandberg SLR Product Line SCSI Functional Specifications

4.6. Multiple Connections

4.6.1. Background

While the SCSI-bus is in the BUS FREE state, any Initiator may attempt a connection to the Drive. In most cases this will happen when the Drive has completed execution of some previous command (a STATUS byte and a COMMAND COMPLETED message has been sent). As soon as the SCSI-bus becomes free, an Initiator may again select the Drive. The Drive will respond to the selection and a new command may be transferred and later executed by the Drive. This is the usual scenario where commands are executed in a serial manner. The connecting Initiators may be the same or different Initiators.

If allowed, the Drive may disconnect from the current Initiator during command execution. The Drive reconnects automatically at a later stage, but this will temporarily leave the SCSI-bus in the BUS FREE state, even if the current command has not completed its execution. During these periods of BUS FREE phases, any Initiator (including the original), may seize the opportunity to select the Drive. In a multiple-initiator system, one Initiator may not even know that another Initiator already has established a connection to the Drive. When the Drive is logically connected to two Initiators at the same time (or the same Initiator twice), a concurrent command situation exists.

Regardless of whether a concurrent command situation exists or not, the different (or same) Initiator may connect to the same or different LUNs within the Drive. Even if the Drive only has a single LUN (LUN 0 (zero)), attempts may be made to connect to other LUNs.

Tandberg Data Commands

4-10 Tandberg SLR Product Line SCSI Functional Specifications

We have the following combinations of Initiator Ids, LUNs and concurrent/serial command execution:

Serial/Concurrent Connecting Initiator Specified LUN See Section

Serial Commands Any Initiator 0 4.6.2

Other than 0 4.6.3

Concurrent Command Same as original Initiator 0 4.6.4

Other than 0 4.6.5

Different Initiator 0 4.6.6

Other than 0 4.6.5

Table 4-6: Connections, LUN and Initiator ID Combinations

Initiator, ID = 6

Initiator, ID = 7

Drive, ID = 0

Drive, ID = 1

SC SI Bus

LU N 0

(LU N 1)

LU N 0

Client A

Client B

Client C

Client X

Client Y

C o nc urrent Commands, Different Initiators to Sam e LUN

Incorrect Initiator C o nne c tion

Unsupported LU N

Figure 4-1: Connections, Some Examples

4.6.2. Commands Received Serially, LUN Is 0

When the drive is not engaged in any command execution any initiator may select the Drive. The Drive will respond to the selection, the Initiator may send a new command, which later will be executed by the Drive.

Tandberg Data Commands

4-11 Tandberg SLR Product Line SCSI Functional Specifications

4.6.3. Commands Received Serially, LUN Not 0

The Drive responds to the selection and a new command may be transferred. If the command is an INQUIRY or REQUEST SENSE command it will be executed. The parameter data returned will however reflect an U

NSUPPORTED

LUN condition (see the INQUIRY and REQUEST SENSE command sections for details). All other commands will be terminated immediately with CHECK CONDITION status. The error code generated will be U

NSUPPORTED

LUN.

4.6.4. Concurrent Command, Same Initiator, LUN Is 0

The Drive will respond to the selection. If the Initiator does not send an ABORT or a BUS DEVICE RESET message during the same MESSAGE OUT phase as the IDENTIFY message, and Incorrect Initiator Connection condition exists [2], [5].

When the Drive detects an Incorrect Initiator Connection the Drive aborts the command already executing (the original command that disconnected) and terminates the current command with CHECK CONDITION status. The error code generated will be O

VERLAPPED COMMANDS ATTEMPTED

4.6.5. Concurrent Command, Any Initiator, LUN Not 0

NSUPPORTED

LUN condition (see the INQUIRY and REQUEST SENSE command sections for details). All other commands will be terminated immediately with CHECK CONDITION status. The error code generated will be U

NSUPPORTED

LUN.

4.6.6. Concurrent Command, Different Initiator, LUN Is 0

The Drive will respond to the selection and the Initiator may send a new command. If the command is one of the following commands,

 INQUIRY,

 REQUEST SENSE,

 TEST UNIT READY,

it will be executed as normal (concurrently with the command already executing in the Drive). All other commands are terminated with a BUSY status.

Tandberg Data Commands

4-12 Tandberg SLR Product Line SCSI Functional Specifications

This Page Intentionally Left Blank

5-1 Tandberg SLR Product Line SCSI Functional Specifications

Status Bytes

A status byte will be sent from the 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 a SCSI-bus reset condition.

The Drive supports the Status Bytes shown in the table below:

Status Byte Name Hex Code

BUSY STATUS 08

CHECK CONDITION STATUS 02

GOOD STATUS 00

INTERMEDIATE STATUS 10

RESERVATION CONFLICT STATUS 18

Table 5-1: The Status Set

BUSY Status

The command can be accepted, but it can not be executed because the Drive is busy.

Three different BUSY situations may occur:

The drive is already executing a command from another initiator and the new command can't be executed in parallel.

The Drive is executing an Immediate-type command (from any initiator) and the BSYI-option is turned on (please refer to the BSYI-bit in the Miscellaneous Parameters Page of the MODE SELECT command). This condition will prevail until the Immediate-type command has completed its execution (or the BSYIoption is turned off).

The drive is occupied by a drive-initiated preparation activity (as e.g. Autoload) and the BSYAoption is turned on (please refer to the BSYA-bit in the Miscellaneous Parameters Page of the MODE SELECT command). This condition will prevail until the drive-initiated activity has completed (or the BSYAoption is turned off).

CHECK CONDITION Status

An abnormal condition has occurred. The Initiator should issue a REQUEST SENSE command to get further information.

GOOD Status

The requested operation (the last command) was completed successfully.

INTERMEDIATE Status

This status is sent after a successfully completed command in a series of linked commands.

RESERVATION CONFLICT Status

The status is sent to an Initiator that attempts to access the Drive when it is reserved for another Initiator.

Tandberg Data Status Bytes

5-2 Tandberg SLR Product Line SCSI Functional Specifications

This Page Intentionally Left Blank

6-1 Tandberg SLR Product Line SCSI Functional Specifications

Message System

The message system allows communication between an Initiator and the Drive for the purpose of physical path management. The physical path may be broken and reestablished several times during the execution of a SCSI command if the Initiator has granted the Drive the privilege of disconnection.

The first message sent by the Initiator after the SELECTION phase has to be either the IDENTIFY, ABORT TASK SET or TARGET RESET message. The IDENTIFY message may be immediately followed by other messages. The IDENTIFY message establishes the physical path for a particular logical unit specified by the Initiator. Since the Drive has only one logical unit, the specified logical unit number should always be zero. The IDENTIFY message may also grant the Drive disconnection privilege. If the Drive is allowed to disconnect, it may do so by transferring a DISCONNECT message and then release the SCSI-bus (by entering the BUS FREE phase).

The Drive will always at some later stage reestablish the physical path by reselecting the Initiator. After the RESELECTION phase, the Drive's first message will be IDENTIFY. This allows the physical path to be reestablished for the Drive's specified logical unit number.

Whenever a physical path to an Initiator that is utilizing disconnection and reconnection is established, the Initiator must ensure that the active pointers of the physical path are equal to the saved pointers for that particular logical unit number (an implied restore operation should occur as a result of a reselection).

When the Drive has completed the execution of a SCSI command (successfully or not) it will signal the Initiator that it is about to break the physical path for good (for this selection sequence) by transferring a TASK COMPLETE message. The Drive will then enter the BUS FREE phase.

Tandberg Data Message System

6-2 Tandberg SLR Product Line SCSI Functional Specifications

6.1. Message In

The Drive supports the Message-In Bytes shown in the table below:

Message Name Input/Output Hex Code

TASK COMPLETE

DISCONNECT

EXTENDED MESSAGE

IDENTIFY

IGNORE WIDE RESIDUE

LINKED COMMAND COMPLETE

LINKED COMMAND COMPLETE W/FLAG

MESSAGE REJECT

RESTORE POINTERS

SAVE DATA POINTER

I/O

80+

Table 6-1: The Message-In Set

TASK COMPLETE

This message will be sent from the Drive to the Initiator to indicate that the execution of a command has terminated. Valid Status will have been sent. This message will always be sent next to a BUSY, CHECK CONDITION, GOOD or RESERVATION CONFLICT status byte. After successfully sending this message, the Drive will enter the BUS FREE phase.

DISCONNECT

This message will be sent from the Drive to the Initiator to inform the Initiator that the Drive is about to disconnect. This message is usually sent second to the SAVE DATA POINTER message. After successfully sending this message, the Drive will enter the BUS FREE phase.

EXTENDED MESSAGE

This message is sent by the Drive to the Initiator as the first byte of a multiple byte message. See Section

6.3 for a description of the extended messages.

IDENTIFY

IDENTIFY will be sent from the Drive to the Initiator immediately after a reconnect (this will be an IDENTIFY with LUN equal to the LUN in the incoming IDENTIFY message (sent during a message-out phase) and with the disconnect privilege bit set to zero, typically code 80h.

LINKED COMMAND COMPLETE

This message will be sent from the Drive to the Initiator to indicate that the execution of a linked command (without the FLAG bit set) has completed. This message will always be sent next to INTERMEDIATE status.

LINKED COMMAND COMPLETE W/FLAG

This message will be sent from the Drive to the Initiator to indicate that the execution of a linked command (with the FLAG bit set) has completed. This message will always be sent next to INTERMEDIATE status.

Tandberg Data Message System

6-3 Tandberg SLR Product Line SCSI Functional Specifications

MESSAGE REJECT

The Drive sends a MESSAGE REJECT message to the Initiator if it receives a message other then the messages listed in the message out table shown below (table 6-2). The Drive also sends a MESSAGE REJECT message to the Initiator if it receives a MESSAGE REJECT message following phases other than a MESSAGE-IN phase.

RESTORE POINTERS

This message will be sent by the Drive to the Initiator before the Drive attempts a retransfer of Data or a Status Byte.

SAVE DATA POINTER

This message will be sent by the Drive to the Initiator before sending the DISCONNECT message to inform the Initiator that a pointer save operation is advised.

IGNORE WIDE RESIDUE

This two-byte message will be sent by the Drive to the Initiator to indicate that the number of valid data bytes sent during the last REQ/ACK handshake of a DATA IN phase is less than the negotiated transfer width. The message will be sent immediately following the DATA IN phase. The second byte of the message contains the number of invalid data bytes transferred, and will be set to 01h.

6.2. Message Out

The Drive supports the Message-Out Bytes shown in the table below:

Message Name Input/Output Hex Code

ABORT TASK SET

TARGET RESET

EXTENDED MESSAGE

IDENTIFY

INITIATOR DETECTED ERROR

MESSAGE PARITY ERROR

MESSAGE REJECT

NO OPERATION

I/O

80+

Table 6-2: The Message-Out Set

ABORT TASK SET

Abort the current operation. The Drive goes to the BUS FREE phase. No Status or ending Message will be sent (see section 6.5 for details on the ABORT TASK SET message handling).

TARGET RESET

Abort the current operation. The Drive goes to the BUS FREE phase. No Status or ending Message will be sent. Possible data in the data buffer will neither be transferred to the Initiator nor written to the tape. Possible pending error conditions will be cleared and reservations are released. This message results in a hard reset, ref. section 3.5.2.

EXTENDED MESSAGE

This message is sent by the Initiator to the Drive as the first byte of a multiple byte message. See Section

6.3 for a description of the extended messages.

Tandberg Data Message System

6-4 Tandberg SLR Product Line SCSI Functional Specifications

IDENTIFY

The IDENTIFY message (see the table below) is sent by the Initiator to the Drive to establish an I T L nexus. It may also grant the privilege of disconnecting.

BIT 7 6 5 4 3 2 1 0

Idfy DscP LUN

Table 6-3: The IDENTIFY Message

Idfy The Identify bit is always set in the IDENTIFY

message.

DscP A Disconnect Privilege bit of one specifies

that the Initiator has granted the Drive the privilege of disconnecting. A DscP-bit of zero

means that the Drive will not disconnect.

LUN The Logical Unit Number field should be set to

zero because the Drive supports only Logical Unit 0.

INITIATOR DETECTED ERROR

The Initiator may send this message to the Drive to inform that an error has been detected in the Initiator. The source of the error may either be related to previous activities on the bus or may be internal to the initiator and unrelated to any previous activities on the bus. The drive generally responds to an INITIATOR DETECTED ERROR message by retrying the last phase. (i.e. by first going back to the previous phase and then by transferring the same information once more). See section 7.5 for details about how the drive responds to an INITIATOR DETECTED ERROR message received after a Status- or Data In phase.

MESSAGE PARITY ERROR

The Initiator may send this message to the Drive to inform that the last transferred message had a parity error. If the previous phase was a MESSAGE IN, the Drive will go back to the MESSAGE IN phase and the previous Message byte will be transferred once more. If a MESSAGE PARITY ERROR message is received when no message in has been sent, the Drive goes to the BUS FREE phase.

MESSAGE REJECT

The Initiator may send this message to the Drive to indicate that the last message transferred from the Drive was inappropriate or not implemented (see also section 6.4).

NO OPERATION

This message will simply be ignored.

Tandberg Data Message System

6-5 Tandberg SLR Product Line SCSI Functional Specifications

6.3. Extended Message

A value of one (01h) of the first byte of a message indicates the beginning of a multiple-byte extended message. The minimum number of bytes sent for an extended message is three. The extended message format is shown in the table below.

Byte Value Description

0 01h Extended Message

1 n Extended Message Length

2 y Extended Message Code

3 - n+1 x Extended Message Arguments

Table 6-4: Extended Message Format

The extended message length specifies the length in bytes of the extended message code plus the extended message arguments to follow. Therefore, the total length of the message is equal to the extended message length plus two. A value of zero for the extended message length indicates 256 bytes to follow.

The extended messages supported by the are shown in the table below. The extended messages are described in detail in Sections 6.3.1 and 6.3.2.

Extended Message Code Description

01h Synchronous Data Transfer Request

03h Wide Data Transfer Request

Table 6-5: Supported Extended Message Codes

To comply with the ANSI standard [2] and [5], Wide Data Transfer must be negotiated prior to negotiating Synchronous Data Transfer. If a Synchronous Data

Transfer agreement is in effect, it will be reset to asynchronous mode if the Wide Data Transfer Request is accepted.

If a parity error is detected during a MESSAGE-OUT phase the Drive will consume all the remaining bytes in the message and ask for a re-transfer of the whole message.

If a parity error is detected during a MESSAGE-IN phase the Initiator may signal a parity error by transferring a MESSAGE PARITY message back to the Drive. The Drive responds by transferring the whole message once more.

Tandberg Data Message System

6-6 Tandberg SLR Product Line SCSI Functional Specifications

6.3.1. Wide Data Transfer Request

The Wide Data Transfer Request (WDTR) message exchange establishes an agreement between the Initiator and the Drive on the width of the data path to be used for DATA phase transfers between the two devices. The agreement applies to DATA IN and DATA OUT phases only. All other information phases will use an eight-bit data path.

A WDTR message exchange is initiated by the Initiator whenever a previously arranged transfer width agreement may have become invalid, or whenever it is appropriate to negotiate a new transfer width agreement. Renegotiation at every selection is not recommended, since a significant performance impact is likely.

The agreement becomes invalid after a reset condition. The Initiator sets its transfer width to the maximum

data path width it elects to accommodate. If the Drive can also accommodate this transfer width, it returns the same value in its WDTR message. If it requires a smaller transfer width, it substitutes the smaller value in its WDTR message. The successful completion of an exchange of WDTR messages implies an agreement as follows:

WDTR Response: Implied agreement:

1) Non-zero transfer width Each device transmits and receives data with the specified transfer width.

2) Transfer width equal to zero Eight-bit data transfer

3) MESSAGE REJECT message Eight-bit data transfer

If the initiator recognizes that negotiation is required, it asserts the ATN signal and sends a WDTR message to begin the negotiating process. After successfully completing the MESSAGE OUT phase, the Drive responds with the proper WDTR message. If an abnormal condition prevents the Drive from returning an appropriate response, both devices shall go to eight-bit data transfer mode for data transfers between the two devices.

Following Drive response (1) above, the implied agreement for wide data transfers is considered to be negated if the initiator asserts ATN and the first message out is either MESSAGE PARITY ERROR or MESSAGE REJECT. In this case, both devices shall go to eight-bit data transfer mode for data transfers between the two devices. For the MESSAGE PARITY ERROR case, the implied agreement shall be reinstated if a re-transmittal of the second of the pair of messages is successfully accomplished. After a specific number of retry attempts, if the Drive receives a MESSAGE PARITY ERROR message, it will terminate the retry activity, and both devices shall go to eight-bit data transfer mode for data transfers between the two devices. The number of retry attempts is specified in the Bus Parity Error Retries field in the Miscellaneous Page of the Mode Select command. Please refer to section 14.3.9 for further details.

The implied transfer width agreement shall remain in effect until a TARGET RESET message is received, until a hard reset condition occurs, or until the Initiator

Tandberg Data Message System

6-7 Tandberg SLR Product Line SCSI Functional Specifications

elects to modify the agreement. The default data transfer width is eight-bit data transfer mode.

Tandberg Data Message System

6-8 Tandberg SLR Product Line SCSI Functional Specifications

The Wide Data Transfer Request Message has the following format:

Byte Value Description

0 01h Extended Message

1 02h Extended Message Length

2 03h Wide Data Transfer Request Code

3 m Transfer Width Exponent

Table 6-6: Wide Data Transfer Request

Transfer Width Exponent

Valid transfer widths are 8 bits (m=00h) and 16 bits (m=01h).

6.3.2. Synchronous Data Transfer Request Message

Synchronous Data Transfer Request (SDTR) message exchange can be initiated by the Initiator to change the data transfer agreement. The default data transfer mode is asynchronous data transfer mode.

An Initiator may initiate an SDTR message whenever it is appropriate to negotiate a new data transfer agreement (either synchronous or asynchronous). The Drive will never respond to an SDTR message with a MESSAGE REJECT message. Re-negotiation at every selection is not recommended, since a significant performance impact is likely.

The Synchronous Data Transfer Request Message has the following format:

Byte Value Description

0 01h Extended Message

1 03h Extended Message Length

2 01h Synchronous Data Transfer Request Code

3 m Transfer Period (multiples of 4 ns)

4 x REQ/ACK Offset

Table 6-7: Synchronous Data Transfer Request

Tandberg Data Message System

6-9 Tandberg SLR Product Line SCSI Functional Specifications

Transfer Period

The Transfer Period is the minimum time allowed between leading edges of successive REQ pulses and of successive ACK pulses, set by the device specifications for successful reception of data.

The minimum transfer period is 50 nanoseconds, m = 0Ch (12). The maximum transfer period is 544 nano seconds, m = 88h (136).

When a synchronous data transfer period of less than 200 ns is negotiated, “ fast synchronous data transfer” is used. This will result in different values for some SCSI-bus Timing Values, see [2] and [5].

When a synchronous data transfer period of less than 100 ns is negotiated, Ultra SCSI (Fast-20) data transfer is used. This will result in different values for some SCSI-bus Timing Values, see [18].

REQ/ACK Offset

The REQ/ACK Offset is the maximum number of REQ pulses allowed to be outstanding before the corresponding ACK pulse is received at the Drive. The value is limited by the size of the device's reception buffer. A REQ/AC

off- set of zero indicates asynchronous data transfer mode. The maximum REQ/ACK offset of the Drive is 15.

The default transfer mode is entered at power on, after a TARGET RESET message or a hard reset condition.

The Initiator sets its values to permit it to receive data successfully. If the Drive can also receive data successfully with these values (or smaller transfer periods or larger REQ/ACK offsets or both), it returns the same values in its SDTR message. If it requires a larger transfer period, a smaller REQ/ACK offset, or both in order to receive data successfully, it substitutes values in its SDTR message as required, returning unchanged any value not required to be changed.

The successful completion of an exchange of SDTR messages implies an agreement as follows:

Responding Device SDTR response: Implied Agreement:

1) Non-zero REQ/ACK offset Each device transmits data with a transfer period equal to or greater than and a REQ/ACK offset equal to or less than the values received in the other device's SDTR message.

2) REQ/ACK offset equal to zero Asynchronous transfer

3) MESSAGE REJECT message Asynchronous transfer

If the initiator recognizes that negotiation is required, it asserts the ATN signal and sends a SDTR message to begin the negotiating process. After successfully completing the MESSAGE OUT phase, the Drive will respond with the proper SDTR message. If an abnormal condition prevents the Drive from returning an appropriate response, both devices shall go to asynchronous data transfer mode for data transfers between the two devices.

Tandberg Data Message System

6-10 Tandberg SLR Product Line SCSI Functional Specifications

Following Drive response (1) above, the implied agreement for synchronous operation is considered to be negated if the initiator asserts the ATN signal and the first message out is either MESSAGE PARITY ERROR or MESSAGE REJECT. In this case, both devices shall go to asynchronous data transfer mode for data transfers between the two devices. For the MESSAGE PARITY ERROR case, the implied agreement shall be reinstated if a retransmittal of the second of the pair of messages is successfully accomplished. After a specific number of retry attempts, if the Drive receives a MESSAGE PARITY ERROR message, it will terminate the retry activity, and both devices shall go to asynchronous data transfer mode for data transfers between the two devices. The number of retry attempts is specified in the Bus Parity Error Retries field in the Miscellaneous Page of the Mode Select command. Please refer to section 14.3.9 for further details.

The implied synchronous agreement shall remain in effect until a TARGET RESET message is received, until a hard reset condition occurs, or until the Initiator elects to modify the agreement.

Tandberg Data Message System

6-11 Tandberg SLR Product Line SCSI Functional Specifications

6.4. Message Reject Message Handling

This section specifies the use of the MESSAGE REJECT message in both message-in and message-out phases.

6.4.1. Message-In Phase

The drive sends a MESSAGE REJECT message in the following instances:

 When an unsupported message has been received from

the Initiator. Supported messages are shown in table 6-2.

 When a MESSAGE REJECT message has been received after

a phase that was not a message-in phase.

6.4.2. Message-Out Phase

If, after a MESSAGE-IN phase, the Initiator asserts ATN and transfers a REJECT message, the Drive will react as shown in the table below:

Last Message-In Action Taken

DISCONNECT The Drive will not disconnect any more until an IDENTIFY message with the

Disconnect Privilege bit set to one is received.

IDENTIFY The current command is terminated and the drive goes to BUS FREE.

IGNORE WIDE RESIDUE The reject is ignored and the current command continues execution.

LINKED COMMAND COMPLETE

The current command is terminated and the drive goes to BUS FREE.

LINKED COMMAND COMPLETE W/FLAG

The current command is terminated and the drive goes to BUS FREE.

MESSAGE REJECT The reject is ignored and the current command continues execution.

SAVE DATA POINTER Same as for a DISCONNECT message.

SYNCHRONOUS DATA TRANSFER REQUEST

The Drive goes back to asynchronous data transfer.

TASK COMPLETE Ignore the REJECT message (the drive goes to BUS FREE).

WIDE DATA TRANSFER REQUEST

The Drive goes back to 8-bit data transfer.

Table 6-8: Response to MESSAGE REJECT

If the MESSAGE REJECT message is sent following phases other than MESSAGE-IN, the Drive will respond by sending a MESSAGE REJECT message.

Tandberg Data Message System

6-12 Tandberg SLR Product Line SCSI Functional Specifications

6.5. Abort Message Handling

If the Initiator transfers an ABORT TASK SET message during the MESSAGE IN phase, the Drive will immediately go to the BUS FREE phase. No Status or ending Message will be sent. The Drive will remain in its current state (MODE settings and current tape position must be kept).

When receiving a new command, the Drive will be able to continue where that last aborted command left off.

See Section 4.6. Multiple Connections for a description of ABORT TASK SET in overlapped command situations.

See also section 3.5.1. Attention (ATN) for details on ATN signal handling. The following is true for all commands:

 If the ABORT TASK SET message is transferred before

the Drive enters the Command phase, then the Drive will just go to the BUS FREE phase, effectively ignoring the whole selection.

 If the ABORT TASK SET message is transferred

immediately after the Command phase (ATN must be asserted during command transfer), then the Drive will just go to the BUS FREE phase, effectively ignoring the whole command. This is true even if the Drive detects errors in the command block (like Bus Parity Error).

 If the ABORT TASK SET message is transferred

immediately after the Status phase the command will have been executed and the Drive will go to the BUS FREE phase

The table on the following page describes ABORT TASK SET- for all commands when the ABORT TASK SET-message is transferred at any later stage in the phase sequencing.

Tandberg Data Message System

6-13 Tandberg SLR Product Line SCSI Functional Specifications

Command Action

ERASE LOAD/UNLOAD REWIND

These commands continue to execute as if executed in Immediate (or buffered) mode.

WRITE FILEMARK When the WRITE FILEMARK command is aborted after

disconnection, it has already been executed.

INQUIRY LOG SENSE MODE SENSE READ BLOCK LIMITS READ POSITION REQUEST BLOCK ADDRESS REQUEST SENSE

c If the ABORT TASK SET message is received before the parameter data is transferred, transfer is suppressed.

d Note that for the REQUEST SENSE command the parameter data is usually lost. The other commands may be executed again to get the parameter data once more.

MODE SELECT The actual mode selection will not be performed if the command

is aborted immediately after the parameter data transfer (ATN must be asserted during the data transfer). If the command is aborted in any later phase, the mode selection will, however, already have been performed.

PREVENT/ALLOW MEDIA REMOVAL RELEASE RESERVE TEST UNIT READY

These commands will be executed (the ABORT TASK SET came too late).

READ If the command is aborted immediately after a data transfer (ATN

must be asserted during the data transfer), the next Read operation will start on the following block. The Initiator may request sense information (Block Counter) to check the actual number of blocks read.

READ BUFFER This command will be aborted after the last data transfer. The

command may be issued again to retransfer the last transferred data and also the data not transferred when the command was aborted.

LOCATE The actual seek operation will not be performed if the command is

aborted immediately after the parameter data transfer (ATN must be asserted during the data transfer). If the command is aborted in any later phase, the seek operation will execute as if the command was executed in immediate mode.

SEND DIAGNOSTICS The actual diagnostic operation will not be performed if the com-

mand is aborted immediately after the parameter data transfer (ATN must be asserted during the data transfer). If the command is aborted later, the diagnostic operation has, however, already been performed.

SPACE

The space operation will be terminated immediately. Note however, that if the Drive does not disconnect and ATN is asserted some time after the command phase, the complete space operation will usually have completed before the Drive tests the ATN line (the Drive tests for ATN only when changing bus phases). The Initiator may request sense information (Block Counter) to check the actual number of spaced blocks.

VERIFY (Not supported by SLR7 and SLR140)

If the command is aborted immediately after a data transfer (ATN must be asserted during the data transfer), the next Compare operation will start on the following block. The Initiator may request sense information (Block Counter) to check the actual number of compared blocks.

WRITE If the command is aborted immediately after a data transfer (ATN

must be asserted during the data transfer), data in the last transfer will not be written on tape. If the last block continues over several transfers, then this complete block will not be written on tape.

WRITE BUFFER

This command will be aborted after the last data transfer. The transferred data has already been written into the data buffer.

Table 6-9: ABORT TASK SET Message Handling

Tandberg Data Message System

6-14 Tandberg SLR Product Line SCSI Functional Specifications

6.6. Unexpected Bus Free

An unexpected bus free occurs when the Initiator detects a BUS FREE condition any time the Initiator does not expect a BUS FREE condition. The drive uses an unexpected bus free to inform the Initiator of a protocol error. The Drive terminates the current command before going to the BUS FREE phase.

The Drive uses an unexpected bus free in the following cases:

 When the first message sent by the Initiator after a

successful selection is not an IDENTIFY, ABORT TASK SET or TARGET RESET message.

 When a MESSAGE PARITY ERROR message has been received

and the previous phase was not a message-in phase.

 When a second IDENTIFY message with a LUN field value

different from the value in the initial IDENTIFY message has been received in the same connection.

 When the Initiator rejects the IDENTIFY message sent

by the Drive during a reconnection.

 When a LINKED COMMAND COMPLETE or a LINKED COMMAND

COMPLETE W/FLAG message has been rejected.

 When a RESTORE POINTER message has been rejected.  When a fatal internal error has been detected.

Tandberg Data Message System

6-15 Tandberg SLR Product Line SCSI Functional Specifications

This Page Intentionally Left Blank

Tandberg Data SLR7-140 PROGRAMMER, SLR7, SLR50, SLR60, SLR75 User Manual

Specifications and Main Features

Frequently Asked Questions

User Manual