JP4257559B2 - Data processing apparatus and data reproduction method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention provides a data processing device. Placement The data reproduction method can be applied to a tape streamer that backs up data such as a server. The present invention provides a continuity of first identifiers assigned to recording tracks before and after a recording track in which an error is detected, Continue By ignoring the error by determining the attribute of the recording track, user data can be reliably sent to the host device even if an error occurs in the reproduced data.
[0002]
[Prior art]
Conventionally, a tape streamer is used for backup of a server or the like, for example, by adding an error correction code and recording desired data at a high density.
[0003]
That is, in the tape streamer, data to be used for backup is divided in predetermined block units, and an error correction code in the product code format is generated. Further, the tape streamer outputs these data and error correction codes to the recording head in a predetermined order, and records these data on the magnetic tape while forming recording tracks on the magnetic tape in an oblique manner. At this time, the tape streamer scans the scanning trajectory of the recording head with the reproducing head, performs read after write, and executes a so-called retry process for a portion where data cannot be reproduced correctly.
[0004]
On the other hand, at the time of reproduction, the tape streamer scans the magnetic tape with the reproducing head and processes the reproduction signal, thereby sequentially reproducing the data recorded on the magnetic tape, and error by the error correction code added at the time of recording. Correct it. Further, the retry processing is repeated for the portion where error correction processing is difficult as described above. As a result, the tape streamer can back up desired data with high reliability.
[0005]
[Problems to be solved by the invention]
By the way, in this type of tape streamer, an error is detected in units of a plurality of blocks and a retry process is executed, and the plurality of blocks include data recorded other than user data.
[0006]
That is, in the tape streamer, dummy data may be recorded instead of the user data, and such dummy data does not affect the reproduction of the user data even if it cannot be reproduced correctly. Even in such a case, if the retry process is repeated in units of multiple blocks, the retry process will be repeated uselessly, and if the retry fails, the user data can be reproduced correctly. Nevertheless, transmission of the reproduced user data to the host computer is stopped.
[0007]
Such a phenomenon is considered to increase as the recording density on the magnetic tape increases. For example, in the case of connecting writing, it is considered that the phenomenon becomes prominent at the connecting portion.
[0008]
The present invention has been made in consideration of the above points. A data processing device capable of reliably sending user data to a host device even when an error occurs in reproduced data. Placement And a method for reproducing data.
[0009]
[Means for Solving the Problems]
Claim 1 or claim for solving the problem 2 In the invention, the continuity of the first identifier assigned to the recording tracks before and after the recording track in which the error is detected, and the recording track in which the error is detected are applied to the data processing apparatus or the data reproducing method. In Continue Based on the determination based on the second identifier set in the recording track, the reproduction data is processed ignoring the detection of the error.
[0012]
Claim 1 or Claim 1 In the configuration, the continuity of the first identifier assigned to the recording tracks before and after the recording track where the error is detected, and the recording track where the error is detected Continue According to the determination based on the second identifier set in the recording track, the attribute of the data recorded in the recording track in which the error has occurred can be determined based on the first and second identifiers set at the time of recording. it can. Therefore, if the reproduction data is processed by ignoring the detection of the error by this determination, an error about unnecessary data can be ignored, and even if an error occurs in the reproduced data, the user data can be surely obtained. Can be sent to the host device.
[0015]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings as appropriate.
[0016]
(1) Configuration of the first embodiment
(1-1) Overall configuration
FIG. 1 is a perspective view showing a backup system according to an embodiment of the present invention. The backup system 1 is configured by arranging a tape streamer 3 and the like on a large console 2. That is, in this backup system 1, a tape streamer 3, a power supply unit 4, and a CPU unit 5 are sequentially arranged from the lower front side, and a tape cassette transport mechanism 6 is arranged on the rear side.
[0017]
Here, the power supply unit 4 supplies power to each part of the backup system 1, and the CPU unit 5 controls the operation of the entire backup system 1. The tape streamer 3 records data from the host computer on the tape cassette loaded by the transport mechanism 6, and reproduces the data recorded on the tape cassette and outputs it to the host computer.
[0018]
The transport mechanism 6 is configured so that a plurality of tape cassettes can be stored in the shelf 6A. Under control of the CPU unit 5, the tape cassette held on the shelf 6A is transported to and loaded into the tape streamer 3. In contrast, the tape cassette discharged from the tape streamer 3 is transported to the original shelf 6A. For this reason, the transport mechanism 6 is mounted on the elevator 6B, which is movable in the vertical direction of the console 2, and a tape cassette is inserted into and removed from the shelf 6A, and between the tape streamer 3 and the tape streamer 3. The hand block 6C for taking in and out is arranged. In the tape cassette applied to the tape streamer 3 according to this embodiment, a non-contact type IC tag is disposed on the back surface, and a reader / writer for accessing the IC tag is disposed on the shelf 6A and the elevator 6B. It is made so that.
[0019]
FIG. 2 is a block diagram showing the CPU unit 5 together with the peripheral configuration. In the CPU unit 5, a host computer interface (host computer IF) 5A is connected to the host computer 8 by, for example, SCSI (Small Computer System Interface), takes in commands from the host computer 8 and data used for recording, and inputs them to the internal bus BUS. On the contrary, the status data to the host computer 8 output to the internal bus BUS, the reproduced data, etc. are output to the host computer 8.
[0020]
The streamer interface (streamer IF) 5B is connected to the tape streamer 3 by, for example, SCSI, and outputs a predetermined control command output to the bus BUS and data for recording to the tape streamer 3, and conversely, the tape streamer 3B. The status data output from 3 and the reproduced data are received from the tape streamer 3 and output to the internal bus BUS.
[0021]
The control panel 5D is arranged in front of the console 2, accepts various operations by the operator, and notifies the CPU 5C of the operations by the operator via the internal bus BUS. The driver 5E drives the elevator 6B and the hand block 6C under the control of the CPU 5C.
[0022]
The CPU 5C secures a work area in the memory 5F and executes a predetermined processing procedure, thereby controlling the operation of the entire backup system 1 in response to a command from the host computer 8. That is, when an access command is input from the host computer 8, the tape streamer 3 is instructed to record / reproduce the tape cassette. At this time, if the access command from the host computer 8 is for a tape cassette different from the tape cassette loaded in the tape streamer 3, the transport mechanism 6 is controlled to load the corresponding tape cassette into the tape streamer 3. When the operator instructs the so-called virgin tape format, similarly, the tape cassette 3 is loaded into the tape streamer 3 under the control of the transport mechanism 6 and the tape streamer 3 is instructed to perform initialization processing.
[0023]
(1-2) Configuration of tape streamer
FIG. 3 is a plan view showing the front and back of the tape streamer 3. The tape streamer 3 has an insertion port 3A formed in the front (FIG. 3 (A)) so that a tape cassette can be inserted and discharged from the insertion port 3A. The tape streamer 3 is provided with various connection connectors on the back surface thereof, and can be directly connected to the host computer 8 by omitting the CPU unit 5, for example.
[0024]
(1-2-1) Tape streamer format
FIG. 4 is a schematic diagram showing the format of a recording track by the tape streamer 3. As shown in FIG. In the tape streamer 3, longitudinal tracks are formed above and below the magnetic tape 10, and oblique tracks are formed between the longitudinal tracks. Here, one of the upper and lower longitudinal tracks is assigned to the controller track 10A, and a control signal for tracking control is recorded. The other longitudinal track is assigned to the ID track 10B, and information for managing the oblique track is recorded by the management table.
[0025]
On the other hand, the oblique track is formed by sequentially arranging a set of recording tracks with positive and negative azimuth angles. Each oblique track is formed so as to be divided into two by a substantially central portion, and the recording track on the preceding side of one set of recording tracks is used for tracking control in the divided portion in the center. A region TP in which a pilot signal is recorded is formed.
[0026]
Further, the oblique track is formed by using two sets and four recording tracks as one track set, and this track set is used as a unit. Each track set is set with a track set ID used to identify each track set, and this track set ID is assigned to a part of information recorded on the ID track 10B. The track set ID is set such that a physical track set ID that simply increases in all track sets and a logical track set ID that simply increases except for dummy, EOD, and the like. The track set ID is converted into a time code by SMPTE and recorded on the ID track 10B.
[0027]
Each oblique track is assigned to an A track, a B track, a C track, and a D track corresponding to each channel of the recording / reproducing system from the leading side of the track set.
[0028]
The track set forms a user track set used for recording user data, a tape mark track set assigned to a tape mark indicating a file delimiter, an EOD (End Of Data) track set indicating the end of data, and a dummy track. A dummy track set is distinguished.
[0029]
FIG. 5 is a schematic diagram showing the relationship between the track set formed in this way and the ECC block which is an error correction processing unit. The tape streamer 3 divides user data and the like used for recording into predetermined block units and adds error correction codes to form ECC blocks. In the tape streamer 3, data of eight ECC blocks are allocated to one track set. At this time, each ECC block indicated by numerals 0 to 7 is divided into four equal parts in the direction of the outer parity and assigned to each track, thereby interleaved between the tracks and recorded. In addition, the data assigned to each track is recorded after being interleaved and subjected to word interleaving.
[0030]
FIG. 6 is a schematic diagram showing the ECC block. One ECC block is assigned user data of 77 × 190 bytes, and 27-byte outer parity is generated for the 77-byte data string, and this outer parity is added to the data string. After a block ID of 2 × 104 bytes is further added, a 12-byte inner parity is generated for a 192-byte data string including the block ID in a direction orthogonal to the outer parity. Appended to the column.
[0031]
As a result, the ECC block is formed by adding the error correction code in the product code format and adding the synchronization pattern SYNC of 4 × 104 bytes. Here, the block ID is an ECC block identification ID, and the synchronization pattern SYNC is a specific pattern indicating the start of the ECC block.
[0032]
As a result, in the tape streamer 3, 77 × 190 bytes of user data and the like are allocated to form one ECC block, and further, eight ECC blocks are allocated to one track set. It is possible to allocate user data of 11,040 bytes to one track set.
[0033]
FIG. 7 is a schematic diagram showing a data structure of 11,040 bytes allocated to one user track set. One track set has a format ID of 4 bytes at the head. Here, the format ID is an ID for format identification. In this embodiment, a value FFFF0000h indicating that user data is recorded in a track set of four tracks is described.
[0034]
Subcodes are assigned to the subsequent 136 bytes, and tables such as VSIT, VIT, and BST, which will be described later, user data track sets, identification data of tape mark track sets, logical track set IDs, etc. are assigned to manage track sets. Data, logical track set ID, etc. are recorded.
[0035]
Further, the number of bytes obtained by removing the data length of the block management table from the following 116884 bytes is allocated to the recording of user data and the like. When user data is allocated to this area, dummy data is allocated if the user data cannot be satisfied.
[0036]
The block management table is set to a maximum of 4096 bytes. Here, the tape streamer 3 is configured to be able to record / reproduce blocks in units of a set of track sets, and a table used for managing these blocks is recorded in the block management table. Further, in the track set, the subsequent 12 bytes are allocated to the reservation, and the remaining 4 bytes are allocated to EOD (End Of Data) indicating the end of one track set.
[0037]
In this way, the tape streamer 3 records user data using the track set as a basic unit, and also records various management data.
[0038]
FIG. 8 is a schematic diagram showing the overall layout of the magnetic tape 10. The magnetic tape 10 has a recording start position LBOT (Logical Beginning) on the side where the magnetic tape 10 has traveled a predetermined distance from the starting end PBOT (Physical Beginning Of Tape) of the physical magnetic tape 10 excluding the leader tape connected to the reel. Of Tape) is set, and a recording end position LEOT (Logical End Of Tape) is set on the scanning start end side of the magnetic tape 10 from a similar physical end PEOT (Physical End Of Tape). Various data is recorded on the magnetic tape 10 between the recording start position LBOT and the recording end position LEOT, and as a result, the magnetic tape 10 is used while avoiding an area near the leader tape where the error rate is relatively easy to deteriorate. . The magnetic tape 10 forms one physical volume between the recording start position LBOT and the recording end position LEOT.
[0039]
The magnetic tape 10 has a run-up area formed by a predetermined distance from the recording start position LBOT, and the magnetic tape traveling system can be servo-locked in this run-up area. For this reason, oblique tracks are sequentially formed on the magnetic tape 10 from this run-up area.
[0040]
On the magnetic tape 10, a VSIT (Volume Set Table) is recorded following this run-up area, and the top physical track set ID for recording this VSIT is set to 0ID. Here, VSIT records various data for managing the logical volume between the recording start position LBOT and the recording end position LEOT. That is, the VSIT records the volume name of the magnetic tape, the file name recorded on the magnetic tape, the physical track set ID of the VIT assigned to each file, and the tape streamer 3 accesses this VSIT. The magnetic tape 10 can be discriminated, and each file recorded on the magnetic tape 10 and the recording position of each file can be confirmed. Here, one VSIT is formed by one track set (1ID), and the same contents are repeatedly recorded ten times to ensure high reliability.
[0041]
The magnetic tape 10 has a VSIT retry area for 90 IDs following the VSIT recording area, so that the VSIT can be re-recorded in the retry area and the recovery process can be executed as necessary. Has been made.
[0042]
Further, in the magnetic tape 10, a margin band is formed for a predetermined ID, so that even if VSIT is updated, recorded data in the subsequent area is not affected at all. The magnetic tape 10 is allocated to the area of the logical volume between this position margin band and the recording end position LEOT.
[0043]
FIG. 9 is a schematic diagram showing logical volumes allocated to this logical volume area. In the magnetic tape 10, data is recorded in units of files with a tape mark TM that is a file delimiter code in between (FIG. 9A), and each file is configured by blocks. (FIGS. 9B and 9C).
[0044]
In the tape streamer 3, a unit including DIT (Directory Information Table) and EOD (End Of Data) in one file constitutes one logical volume. Here, the DIT is a table for managing one logical volume to which the DIT is allocated, and one length is formed by 40 IDs. As shown in FIG. 9 (D), the same contents of this 40ID are repeatedly recorded seven times in the DIT, thereby ensuring high reliability.
[0045]
Each DIT is assigned with the run-up area described above in FIG. 8 at the head so that the servo system can be locked. Subsequently, a position margin band is arranged on the rear side, so that the recorded data in the subsequent area is not affected even when updated, as in the case of VIST.
[0046]
As shown in FIG. 9E, the DIT is a VIT (Volume Information Table) using 1 ID, a BST (Bad Spot Table) using 1 ID, a LIDT (Logical ID Table) using 1 ID, a FIT (File Information Table) using 20 ID, and a 1 ID. A UI (User Information Table) is set. The DIT is set so that the remaining 16 IDs are reserved.
[0047]
Here, the physical track set ID of the VIT is recorded in the VSIT, so that the VIT can be accessed based on the recording of the VSIT, and further, the entire DIT can be accessed. In the VIT, the logical track set ID is set so as to match the physical track set ID. The VIT is assigned the volume label of the logical volume to which this DIT is assigned, the first physical track set ID of the user track set, and the last physical track set ID.
[0048]
BST is assigned position information of an area (hereinafter referred to as a butt spot) that is set to be difficult to use by the tape streamer 3 due to scratches on the magnetic tape 10 or the like. That is, in the tape streamer 3, if a bit error that is difficult to correct by error occurs due to read after write or retry processing at the time of recording, the data after this area is recorded again on the magnetic tape. In the tape streamer 3, with respect to the data re-recorded on the magnetic tape in this way, an area where this data should be recorded is defined as a bad spot. In the BST, the physical track set ID at the start end and the physical track set ID at the end end of such a bad spot are recorded.
[0049]
The LIDT (Logical ID Table) is used for high-speed search for each block, and data of a file number, a physical track set ID, and a block number is recorded for every 200 logical track set IDs. As a result, the tape streamer 3 can roughly detect the position of the block to be accessed and search at a high speed.
[0050]
The FIT is formed by recording the physical track set ID of the tape mark and the block number of each block. The UT is information indicating whether or not the volume has been updated, and status data indicating whether or not the volume has been updated is set to FFFFFFFFh before the update and 00000000h after the update.
[0051]
(1-2-1) Schematic configuration of tape streamer
FIG. 10 is a block diagram showing a tape streamer 3 according to such a format. In the tape streamer 3, a SCSI interface (SCSI IF) 21 is connected to the CPU unit 5 by a SCSI interface, and is directly connected to a host computer when the CPU unit is not used. The SCSI interface 21 inputs user data to be recorded and outputs the user data to the memory controller 22 and outputs the reproduced data output from the memory controller 22 to the host computer. Further, the SCSI interface 21 notifies a command from the CPU unit 5 and the host computer 8 to the main CPU unit 23 via the memory controller 22. On the contrary, the status output from the main CPU is received via the memory controller 22, and the status is output to the host computer 8 and the CPU unit 5.
[0052]
The memory controller 22 temporarily holds the user data input via the SCSI interface 21 in the buffer memory 24, and further outputs the data held in the buffer memory 24 at a timing according to the processing of the subsequent ECC encoder (ECCENC) 25. To do. The memory controller 22 receives the data of the block management table, the above-mentioned DIT, etc., the corresponding data from the main CPU unit 23, temporarily holds it in the buffer memory 24, and outputs it to the ECC encoder 25. On the contrary, at the time of reproduction, the memory controller 22 temporarily stores the user data output from the ECC decoder (ECDECEC) 26 in the buffer memory 24 at a timing synchronized with the processing of the ECC decoder 26. The held user data is output to the SCSI interface 21. In the above-described DIT or the like, the data output from the ECC decoder 26 is output to the main CPU unit 23. Further, the error detection result input together with the user data is notified to the main CPU unit 23 by accessing the main CPU unit 23.
[0053]
The buffer memory 24 processes user data that is sequentially input in a bank structure in units of capacity corresponding to the predetermined number of track sets described above, thereby constituting a bank memory. As a result, the tape streamer 3 executes the recording / reproducing process and the retry process using one bank configured in the buffer memory 24 as a unit of processing. Thus, in the track set, when user data is recorded with a data amount less than one bank, a dermy track set in which dermy data is recorded instead of the user data is formed.
[0054]
The ECC encoder 25 forms the ECC block described above with reference to FIG. 6 by generating and adding an error correction code to the data output from the memory controller 22 and adding a synchronization pattern or the like. Further, the ECC encoder 25 outputs data by the ECC block by a plurality of systems corresponding to the arrangement of the magnetic head. Further, at this time, the ECC encoder 25 outputs these data in a predetermined order, thereby executing inter-track interleaving and word interleaving processing.
[0055]
The equalizer 28 converts data output from the ECC encoder 25 into a serial data string by a plurality of systems corresponding to the arrangement of the recording heads, and further modulates the data by a method suitable for recording on the magnetic tape 10. Thereby, the equalizer 28 generates a drive signal for driving the recording head, and drives the recording head mounted on the rotary drum 29 by this drive signal.
[0056]
The rotating drum 29 is provided with a recording head of a predetermined system and a reproducing head that scans the scanning locus of the recording head, and rotates at a predetermined rotation speed under the control of the servo circuit 31. As a result, the tape streamer 3 sequentially forms oblique tracks on the magnetic tape 10 running at a predetermined speed and records user data and the like, and the recording result can be monitored by a reproducing head.
[0057]
The equalizer 30 receives a reproduction signal from a reproduction head mounted on the rotary drum 29 during recording and reproduction, and reproduces the data recorded on the magnetic tape 10 by performing waveform equalization and demodulation processing on the reproduction signal. .
[0058]
The ECC decoder 26 takes in the output data of the equalizer 30 and performs an error correction process using an error correction code added at the time of recording. Further, the ECC decoder 26 outputs the user data obtained by the error correction process in this way to the memory controller 22 together with the error detection result which is the error correction process result. As a result, the tape streamer 3 outputs the data recorded on the magnetic tape 10 to the host computer 8 via the buffer memory 24, and the main CPU unit 23 can acquire data such as VIST if necessary. Yes.
[0059]
In addition, the ECC decoder 26 notifies the main CPU unit 23 of the error detection result obtained in this way via the buffer memory 24, and can thereby execute processing such as retry as needed under the control of the main CPU unit 23. It is made like that. When the user data is recorded, the ECC decoder 26 simply executes an error detection process, and records the error detection result obtained as a result together with the user data or the like in the buffer memory 24, thereby performing a read after write process. It is possible to determine whether or not the data recorded in step 1 has been correctly reproduced. The ECC decoder 26 stores such an error detection result in the buffer memory 24 by setting a corresponding error flag.
[0060]
As a result, the main CPU unit 23 secures a work area in the random access memory (RAM) 33 and executes a central processing unit that executes a predetermined processing procedure in response to a command input via the SCSI interface 21. The operation of the entire tape streamer 3 is controlled, and the status is notified to the host computer 8 as necessary. In this process, the buffer memory 24 is accessed as necessary to acquire information such as VSIT and DIT reproduced from the magnetic tape 10, and the operation of the drive system and the like of the magnetic tape 10 is controlled based on this information. To do. After the user data is recorded, the overall operation is controlled so that the corresponding DIT and VSIT are rewritten.
[0061]
The system controller (system controller) 34 controls the operation of the mechanical system of the tape streamer 3 by exchanging data with the main CPU unit 23 that is executed via a dual port RAM (DP-RAM) 35. That is, in the tape streamer 3, the sensor 37 is arranged at a cassette insertion slot or the like, and detects, for example, whether or not the magnetic tape 10 can be loaded and notifies the servo circuit 31. A capstan motor (M) 36 causes the magnetic tape 10 to run under the control of the servo circuit 31. The fixed head 38 forms the longitudinal tracks 10 </ b> A and 10 </ b> B of the magnetic tape 10 and outputs a reproduction signal of the longitudinal tracks 10 </ b> A and 10 </ b> B to the servo circuit 31.
[0062]
The servo circuit 31 drives a predetermined drive system based on the detection result of the tape cassette by the sensor 37 and executes processing such as loading and unloading of the tape cassette, and processing such as loading and unloading of the magnetic tape 10. Execute. The servo circuit 31 drives the rotary drum 29 to rotate, and controls the rotational speed of the capstan motor 36 by determining the reproduction result of the CTL track 10A based on the rotational phase of the rotary drum 29, thereby tracking control. Execute the process. During reproduction, the servo circuit 31 executes the tracking control process with reference to the pilot signal for tracking control assigned to the oblique track of the magnetic tape 10.
[0063]
On the other hand, when the magnetic tape 10 is initialized, the servo circuit 31 drives the fixed head 38 while the magnetic tape 10 is traveling at a predetermined traveling speed, whereby the longitudinal tracks 10A and 10B are applied to the magnetic tape 10. Form. The servo circuit 31 processes the reproduction signal obtained from the fixed head 38 to reproduce the management data, and outputs this management data to the system controller 34. The time code assigned to the management data is output to the time code recorder (TCR) 40.
[0064]
In this initialization process, the time code recorder 40 sequentially generates time codes and outputs them to the servo circuit 31. The servo circuit 31 uses the time codes and various data output from the system controller 34 for management. Data is generated and the fixed head 38 is driven. On the other hand, at the time of recording and reproduction, the time code obtained from the servo circuit 31 is converted into a track set ID and output.
[0065]
In addition to controlling the operation of the servo circuit 31, the system controller 34 outputs the track set ID output from the time code recorder 40 to the ECC encoder 25 and the like via the main CPU unit 23.
[0066]
The interface CPU (IFCPU) 41 exchanges data with the main CPU unit 23 via the DP-RAM 42, thereby enabling information on the tape cassette to be acquired in the shelf 6A and the elevator 6B, and further to other computers. Data communication with the system is possible.
[0067]
That is, in the tape streamer 3, the reader / writer arranged on the shelf 6A or the like transmits a radio wave from the antenna (ANT) 44 to the IC tag 46 arranged on the back surface of the tape cassette 45, and a response is received from the IC tag 46 by this radio wave. When obtained, various information regarding the tape cassette 45 is transmitted to and received from the IC tag.
[0068]
The display CPU 49 inputs / outputs data to / from the IC tag via the memory label interface 47 to / from the interface CPU 41. The display CPU 49 displays various data on a predetermined display screen according to data input via the interface CPU 41. This display includes information on the tape cassette loaded in the tape streamer 3.
[0069]
The SIO 50 is a serial interface such as RS-232C or RS-422, and transmits / receives maintenance information to / from an external device. The Ether IO is an Ethernet interface, and is connected to a predetermined network so that various kinds of information can be transmitted and received.
[0070]
(1-2-1) Detailed configuration of tape streamer
FIG. 11 is a schematic diagram showing a part of subcodes set to each track set by the main CPU unit 23. In FIG. 11, for simplicity of explanation, one bank is described as being assigned to an 8-track set. Also, the track set where the error occurred is shown in a solid color.
[0071]
The main CPU unit 23 sets a track identifier indicating the attribute of the track set for each track set. Here, the track identifier is a dummy ID (denoted by a symbol D) indicating a track set by dummy data, a user data ID (denoted by a symbol U) indicating a track set by user data, and a tape mark indicating a track set by a tape mark. It is divided into an ID (indicated by a symbol T) and a data end ID (indicated by a symbol E) indicating a track set by EOD.
[0072]
Of these, the data necessary for processing in the host computer 8 corresponds to track set data based on user data and track set data based on tape marks, as shown in FIG. 11B. It will be. Note that the track set by EOD is valid when four track sets are connected at the time of continuous writing, and a tape mark or the like is subsequently recorded so as to leave one of the four track sets. Therefore, the data end ID generated in the middle of the second file (FILE2) is generated when this track set is set to the temporary EOD and subsequently connected and recorded.
[0073]
The main CPU unit 23 records the tape mark on the magnetic tape 10 using three track sets, and controls the overall operation so that dummy data is recorded for the head and end track sets. Accordingly, the main CPU unit 23 controls the entire operation so that unnecessary data is recorded in at least one track set in a portion where the data is recorded following the end of the data.
[0074]
The main CPU 23 increments the logical track set ID (logical ID) by a value of 1 in the case of a meaningful track set in the host computer 8 as described above. The logical track set ID assigned to the track set is assigned as it is.
[0075]
In this way, the main CPU unit 23 sets the subcode, records the track set, and processes the data fetched into the buffer memory 24 by executing the processing procedure shown in FIG. To do.
[0076]
That is, the main CPU unit 23 proceeds from step SP1 to step SP2, and determines whether an error has occurred by accessing the buffer memory 24 and determining an error flag. If a negative result is obtained here, the main CPU unit 23 proceeds to step SP3 and ends this processing procedure. As a result, the main CPU unit 23 switches the processing to the reproduction data of the subsequent bank. Therefore, as indicated by reference numeral (1) in FIG. 11A, when an error does not occur in each bank, the main CPU section 23 selectively sends the user data portion to the host computer 8.
[0077]
On the other hand, if a positive result is obtained in step SP2, the main CPU unit 23 proceeds to step SP4. In this case, for the track set in which an error has occurred in this way, the main CPU unit 23 determines the attribute of the track set in which the error has occurred because the logical track set ID and the track identifier are also uncertain. The main CPU unit 23 determines the attribute and the continuity of the preceding and following logical track set IDs. Continue Execute from the attribute of the track set.
[0078]
That is, in step SP4, the main CPU 23 determines whether or not the increase in the previous and subsequent track set IDs is due to a monotonous increase. In this case, if an error is detected for only one track set, it can be determined that the increase is due to a monotonous increase when the increase in the preceding and following track set IDs is 2. If an error is detected in two consecutive track sets, it can be determined that the increase is due to a monotonous increase when the increase in the preceding and following track set IDs is 3. Here, when the increase in the track set ID before and after is monotonously increased, as indicated by reference numeral (3) in FIG. The detected track set is determined as a track set in which data necessary for the host computer 8 is recorded. Accordingly, when a positive result is obtained in step SP4, the main CPU unit 23 proceeds from step SP4 to step SP5, sets the track set in which an error has occurred to a necessary area, and proceeds to step SP5.
[0079]
On the other hand, if a negative result is obtained in step SP4, the main CPU unit 23 proceeds to step SP7. Here, in the case where the increase in the track set ID before and after is not due to the monotonous increase, for example, when an error is detected in only one track set, the increase in the track set ID before and after is the value. 1 In this case, in FIG. 11, as indicated by reference numeral (2), when unnecessary data is allocated to the track set in which an error has occurred, and as shown by reference numeral (4), to the track set in which an error has occurred. There are two cases where necessary data is allocated and unnecessary data is allocated to the following track set.
[0080]
Accordingly, in the subsequent step SP7, the main CPU unit 23 determines the attribute of the track set in which an error has occurred with reference to the track set identifier of the subsequent track set of the adjacent track sets. Specifically, it is determined whether or not the subsequent track set is an unnecessary track set, and thereby the attribute of the track set in which an error has occurred is determined. That is, when a positive result is obtained in step SP7, the main CPU unit 23 proceeds to step SP5, sets the track set in which an error has occurred to a necessary area, and proceeds to step SP6.
[0081]
On the other hand, if a negative result is obtained in step SP7, in this case, the track set in which this error has occurred can be determined as a track set to which unnecessary data is allocated, and the process moves to step SP8. Here, the main CPU unit 23 sets the track set in which the error has occurred to an unnecessary area, and proceeds to step SP6.
[0082]
In this manner, when determining the attribute of the track set in which the error has occurred, the main CPU unit 23 determines whether or not there is another error in step SP6. If an affirmative result is obtained here, the process proceeds to step SP3. Return. On the other hand, if no other error has occurred, the main CPU unit 23 proceeds from step SP6 to step SP9 and determines whether or not a retry is necessary.
[0083]
Here, if the track set in which an error has occurred is set in the necessary area, the main CPU unit 23 determines that retry is necessary, moves to step SP10, instructs the retry process, and then moves to step SP3. On the other hand, when all the track sets in which errors have occurred are set in unnecessary areas, the main CPU unit 23 ignores the occurrence of errors in this case. That is, the main CPU unit 23 proceeds directly from step SP9 to step SP3, and ends this processing procedure.
[0084]
Thus, when executing the retry process, the main CPU unit 23 rewinds the magnetic tape 10, reproduces the data recorded on the magnetic tape 10 for the corresponding bank, and executes this processing procedure again. . Further, even if the retry is repeated, only if the necessary data cannot be correctly reproduced for this bank, the error status is sent to the host computer 8 and the processing is stopped.
[0085]
(2) Operation of the first embodiment
With the above configuration, in this backup system 1 (FIGS. 1 and 2), the tape cassette stored in the shelf 6A by the management of the CPU unit 5 and the like is loaded into the tape streamer 3 and then loaded, and then the fixed head 38 is sequentially loaded. As a result, the ID track 10B and the CTL track 10A, which are longitudinal tracks of the magnetic tape 10, are recorded on the magnetic tape 10 (FIGS. 4 and 10). As a result, the track set ID based on the time code is recorded on the magnetic tape 10 as one of the management data in units of one track set. Thereafter, after the magnetic tape 10 is rewound, VSIT is recorded at a position traveled by a predetermined distance from the leader tape (FIG. 8), thereby recording the volume name and the like of this tape cassette and performing initialization processing. Complete. The tape cassette initialized in this way is transported to and stored in a predetermined position on the shelf 6A under the management of the CPU unit 5.
[0086]
On the other hand, when data backup is instructed by the host computer 8, if the tape cassette corresponding to the tape streamer 3 is not loaded by the management of the CPU unit 5, the corresponding tape cassette is taken out from the shelf 6A. Set on the tape streamer 3. Further, the tape cassette that has been loaded in the tape streamer 3 until then is discharged from the tape streamer 3 and stored in the shelf 6A.
[0087]
When a tape cassette is loaded into the tape streamer 3 in this way, the tape streamer 3 loads the magnetic tape, reproduces the VSIT, and can access the file designated by the host computer 8 from the VSIT. 10 is run. That is, when the first file is recorded on the magnetic tape 10, the DIT is recorded at a predetermined interval from the VSIT, and then the data output from the host computer 8 is sequentially recorded on the magnetic tape 10. When appending to a previously recorded file, the DIT of the corresponding file is reproduced based on the VSIT recording, the end of the last block is detected from this DIT, and input from the host computer 8 sequentially from this end. Record data to be recorded. When the host computer 8 instructs the reproduction of data, the corresponding file position is detected from the VSIT, the DIT is accessed, and the data recorded on the magnetic tape 10 is reproduced according to the recording of the DIT (FIG. 9). .
[0088]
That is, at the time of recording, an error correction code in the product code format is added to user data sequentially input by the ECC encoder 25 (FIG. 6), and the tracks are interleaved by four tracks constituting one track set ( 5), and sequentially recorded on the magnetic tape 10 by interleaving between words. Further, the data is recorded on the magnetic tape 10 in units of banks set in the buffer memory 24.
[0089]
At this time, in the tape streamer 3, a track identifier indicating the attribute of the track set is set for each track set by the management of the main CPU unit 23, and a track set based on dummy data, a track set based on user data, and a tape are determined based on the track identifier. Various data are recorded so that the track set by the mark and the track set by the EOD can be identified (FIG. 11).
[0090]
In addition, for a track set that is meaningful for processing in the host computer 8, a logical track set ID (logical ID) is sequentially incremented and recorded. For a track set that has no meaning, a logical track assigned to the previous track set is recorded. The set ID is assigned and recorded as it is.
[0091]
A tape mark is arranged between the files, and the tape mark is set to record dummy data for the head and end track sets. Further, a track set in which unnecessary data is recorded in at least one track set is arranged at a portion where data is recorded following the end of the data.
[0092]
On the other hand, during reproduction, a reproduction signal is obtained from the reproduction head mounted on the rotary drum 29, and this reproduction signal is processed by the equalizer circuit 30 to obtain reproduction data. The error correction process is performed. Further, the data subjected to error correction processing is recorded in the buffer memory 24 together with an error flag which is an error correction processing result.
[0093]
In the tape streamer 3, the error flag recorded in the buffer memory 24 is accessed by the main CPU unit 23, and only when an error occurs in necessary data, the retry process is executed for each bank, and the retry is repeated. However, output of user data to the host computer 8 is stopped only when an error occurs in necessary data.
[0094]
That is, the tape streamer 3 determines whether or not there is an error by accessing the buffer memory 24, and if an error is detected, the continuity of the logical track set ID before and after the track set in which the error has occurred and the track in which the error has occurred. From the data recorded in the rear track set of the track sets adjacent to the set, it is determined whether or not the track set in which the error has occurred is a track set in which data necessary for processing of the host computer 8 is recorded.
[0095]
That is, when the increase in the preceding and following logical track set IDs is due to a monotonous increase, the track set in which an error has occurred is determined as a track set in which data necessary for processing by the host computer 8 is recorded. If the increase in the preceding and following logical track set IDs is not due to monotonous increase, the attribute of the track set in which the error has occurred is determined from the attribute of the subsequent track set, and the track set in which the error has occurred is It is determined whether or not the track set records data necessary for processing by the computer 8. That is, in this case, when the track set in which data unnecessary for the processing of the host computer 8 is recorded, such as a dummy track set, and an error occurs in only one track set, an error occurs. The track set is determined as a track set in which data necessary for processing by the host computer 8 is recorded.
[0096]
As a result, when an error occurs in the track set that does not record data necessary for processing by the host computer 8, the tape streamer 3 can ignore this error and not execute retry processing. It is possible to effectively avoid a decrease in access speed due to repeated useless retries. Further, even when an error occurs in such a place, necessary reproduction data can be sent to the host computer 8, whereby user data can be reliably sent to the host computer.
[0097]
(3) Effects of the embodiment
According to the above configuration, when an error occurs, the continuity of the logical track set ID before and after the track set in which the error is detected and the track set in which the error is detected are determined. Continue By ignoring the occurrence of an error based on the determination based on the track identifier set in the track set, even if an error occurs in the reproduced data, the user data can be reliably transmitted to the host computer which is the host device.
[0098]
(4) Second embodiment
By the way, when the track pitch is dense, the writing accuracy with respect to the track width is relatively lowered. For this reason, when linked writing is performed, the track width of a part of the previous track set becomes narrower due to overwriting at the linked portion, and there is a possibility that data cannot be correctly reproduced for this track set. In addition, if the track pitch changes at the connected part, tracking is temporarily disturbed, and there is a possibility that data cannot be correctly reproduced for the first track set that is connected.
[0099]
Thus, in this embodiment, even in such a case, the user data can be correctly reproduced, and the user data is surely combined with the determination of the attribute of the track set in which the error has occurred as described in the first embodiment. To be sent to the host computer.
[0100]
In this case, the data transfer speed of user data sent from the host computer 8 is lower than the data transfer speed in the recording system of the tape streamer 3, thereby stopping the running of the magnetic tape 10. It is a case that repeats.
[0101]
In this embodiment, the track set setting processing by the main CPU unit 23 is the same as the configuration according to the first embodiment except for the point different from the first embodiment described above. In this embodiment, only this different configuration will be described, and redundant description will be omitted.
[0102]
In this embodiment, when stopping the running of the magnetic tape 10, the main CPU unit sets the last four track sets as EOD track sets, so that at least the last track set is unnecessary. Assign data. That is, as shown in the first and third times in FIG. 13, when the file 1 or the file 3 is recorded and the running of the magnetic tape 10 is stopped, and as shown in the second time in FIG. 13, a tape mark is recorded. When the traveling of the magnetic tape 10 is stopped, the last four track sets are set as EOD track sets. In the EOD track set, the main CPU unit manages the bank of the buffer memory 24 in consideration of this fact by being effective by the 4-track set.
[0103]
Also, for tape marks, there is a possibility of adding data by adding blocks, so the first and last track sets of the three track sets are set to dummy track sets, so that these first and last tracks are set. Unnecessary data is assigned to the set.
[0104]
On the other hand, when connecting and writing, the main CPU unit leaves at least one track set to which unnecessary data is allocated in a part before the boundary to be connected and written.
[0105]
That is, when connecting and writing from the EOD track set portion, the main CPU section invalidates the EOD track set by setting the physical track set ID issued to the servo system, and at least of the EOD track set. For one, the subsequent track set is recorded so that it remains recorded. Specifically, the main CPU unit in FIG. 13 continues the track so as to leave only the first track set of the EOD of the four track sets as shown in the tape image from the first recording to the second recording. Record the set. In FIG. 13, the joint is indicated by an arrow.
[0106]
When data is additionally recorded with the tape mark portion invalidated, the main CPU makes the tape mark invalid and keeps at least one track set of the tape mark recorded. Record the following track set. Specifically, the main CPU section continues the track so as to leave only the first track set of the tape marks by the three track sets as shown in the tape image by the second to third recordings in FIG. Record the set.
[0107]
On the other hand, when updating a file, the immediately preceding tape mark is kept valid and the following track set is recorded. Specifically, the main CPU unit displays a track set following the last track set of the tape marks by the three track sets, as shown in FIG. 13 from the third recording to the fourth recording tape image. Record.
[0108]
Further, in the case of connecting and writing in this way, unnecessary data is allocated to the first track set by setting the first track set at the start of connecting writing to a dummy track set.
[0109]
Accordingly, as shown in FIG. 13, in the first to the second time, when connecting and writing from the EOD portion, the main CPU portion records the track set in which unnecessary data, which is one of the EOD track sets, and the tape mark. The dummy track set assigned before the track set is assigned before and after the joint. Similarly, when appending to file 2 as in the second to third times, a track set in which unnecessary data, which is the first track set of the tape mark track set, and the first track set of the file to be added are recorded. The dummy track set assigned to is assigned before and after the joint. Similarly, when the file 3 is rewritten as in the third to fourth times, the track set in which the last unnecessary data is recorded by the tape mark arranged before the file 3 and the first track set of the file 3 are recorded. Assign the track set with unnecessary data before and after the joint.
[0110]
As a result, the main CPU section selectively assigns track sets based on unnecessary data before and after joints, which are parts where errors occur frequently. Note that the main CPU sets the track identifier and the logical track set ID so as to correspond to the setting of the track set by such unnecessary data.
[0111]
As with the main CPU unit 23 described above for the first embodiment, the main CPU unit generates an error during playback by determining the continuity of the preceding and following logical track set IDs and the attributes of adjacent track sets. ignore. As a result, the main CPU section reliably sends the user data to the host computer even when an error occurs.
[0112]
That is, in the case where no error occurs (indicated by reference numeral (1)) as shown in FIG. 14 in comparison with FIG. 11, the image at the time of reproduction after the fourth recording in FIG. 13 is stored in the buffer memory 24. The stored user data can be selectively output to the host computer 8 to supply data necessary for processing of the host computer 8.
[0113]
In addition, when an error occurs in a track set that is not before or after the boundary at which continuous writing is recorded, the relationship between the track identifier and the logical track set ID is set to be the same as described above with reference to FIG. Also in this case, user data stored in the buffer memory 24 can be selectively output to the host computer 8 to supply data necessary for processing of the host computer 8.
[0114]
On the other hand, the first implementation is performed by determining the attribute of the track set in which the error has occurred from the continuity of the preceding and following logical track set IDs and the attribute of the adjacent track set even in the part that is connected and recorded. An error can be processed in the same manner as in the first embodiment.
[0115]
In this embodiment, the main CPU determines the attribute of the track set in which an error has occurred from the attribute of the subsequent track set with reference to the increase in the preceding and following logical track set IDs. That is, in the arrangement of the track set as in the first embodiment, except for the EOD track set, the track set in which necessary data is recorded is always arranged before and after the track set in which unnecessary data is recorded. Will be. On the other hand, in this embodiment, a track set in which exceptionally unnecessary data is recorded continues before and after the boundary in which linked recording is performed.
[0116]
Thus, for example, as indicated by reference numeral (2) in FIG. 14, when an error is detected in only one track set, the increment before and after is the value 1, and necessary data is allocated to the subsequent track set. In the case where the error has occurred, the attribute of the track set in which the error has occurred can be determined and the correct determination result can be obtained in the same manner as in the case where an error has occurred in a portion different from the vicinity in which the linked recording is performed.
[0117]
On the other hand, as shown by reference numerals (3) and (4) in FIG. 14, when an error occurs in one track set before and after the boundary, the increment of the logical track set ID becomes 0. Occurs. However, in this case, since the increment is 0, it can be determined that unnecessary data is continuously allocated to the three track sets including the preceding and following track sets. In such a case, the attribute of the track set in which the error is detected is determined as the track set in which unnecessary data is recorded.
[0118]
Even in the vicinity of the portion recorded in this way, as indicated by reference numeral (5) in FIG. 14, when an error is detected in only one track set, the increment before and after is the value 1, If unnecessary data is assigned to a later track set, the attribute of the track set in which the error occurred is set to the required data, as in the case where an error occurs in a different part from the neighborhood where the linked recording is performed. Can be determined as a recorded track set, and a correct determination result can be obtained.
[0119]
On the other hand, as indicated by reference numerals (6) to (8) in FIG. 14, in an area that is adjacent to the boundary, an error may be detected in two consecutive track sets. . In this case as well, as described above, the main CPU determines the attribute of the track set in which the error has occurred from the continuity of the preceding and following logical track set IDs and the attribute of the adjacent track set. Can do.
[0120]
That is, when an error occurs in two consecutive track sets, the increment of the preceding and following logical track set IDs is 1 and necessary data is allocated to the following track set, or the preceding and following logical track set IDs If the increment of the value is 0, in these two track sets, it is possible to determine the track set to which unnecessary data is allocated.
[0121]
Also, when an error occurs in two track sets, if the increment is 2 and the necessary data is assigned to the following track set, unnecessary data is allocated to these two track sets. Can be determined as a track set to which necessary data is allocated.
[0122]
Thus, in this embodiment, after determining the attribute of the track set in which an error has occurred, the same processing as in the first embodiment is executed based on the determination result.
[0123]
By assigning unnecessary data to the track sets before and after the joint as in the second embodiment, user data can be recorded while avoiding parts with high error occurrence frequency. Even if an error occurs in the reproduced data at a site where such an error occurs frequently, user data can be reliably sent to the host computer.
[0124]
At this time, user data can be transmitted to the host computer more reliably by ignoring the occurrence of an error by determining the continuity of the logical track set ID and the data recorded in the adjacent track set.
[0125]
(5) Third embodiment
By the way, when reproducing from the middle of the magnetic tape, the attribute of the track set in which the error has occurred may not be determined by the above-described method. Therefore, in this embodiment, a linkage flag indicating recording by linkage is set. In this embodiment, the series of processes related to the setting of the linkage flag is the same as the configuration according to the first embodiment except that the series of processes is different from that of the first embodiment described above. In this embodiment, only this different configuration will be described, and redundant description will be omitted.
[0126]
That is, as shown in FIG. 15, the main CPU unit sets a joint writing flag indicating joint recording at a predetermined position behind the boundary by the joint recording when performing joint recording. Here, the main CPU section is related to the processing of the same bank as the track set for starting the linked recording, and the track set that is closest to the boundary of the linked writing to which the user data necessary for the processing of the host computer is allocated. Set the linkage flag. Specifically, the linkage flag is set in the second track set after the linkage boundary.
[0127]
In this joint writing recording, as in the second embodiment described above, the track set is set so that unnecessary data is allocated to the track set before and after the joint.
[0128]
On the other hand, at the time of reproduction, the main CPU ignores the occurrence of an error only for the part that is connected by the process based on the connection flag. In other words, after various examinations, the fact that user data cannot be transferred to the host computer due to the occurrence of an error is due to an error in the linked part, and an error has occurred in other parts. It was found that sufficient characteristics could be secured in practice without determining the attributes of the track set.
[0129]
Thus, in this embodiment, the main CPU unit processes the error detection result by executing the processing procedure shown in FIG. That is, the main CPU proceeds from step SP11 to step SP12, where it determines whether an error has occurred by accessing the buffer memory 24 and determining an error flag. If a negative result is obtained here, the main CPU moves to step SP13 and ends this processing procedure.
[0130]
On the other hand, if a positive result is obtained in step SP12, the main CPU proceeds to step SP14. Here, the main CPU determines whether or not a linkage flag is set at a predetermined position for the track set in which the error has occurred. Here, when the linkage flag is set after one or two track sets from the track set in which the error has occurred, the CPU determines that the linkage flag is set at a predetermined position. When the linkage flag is set at a different position, and when no linkage flag is set, it is determined that the linkage flag is not set at a predetermined position.
[0131]
If a negative result is obtained here, the main CPU proceeds to step SP15, sets this error occurrence location in the necessary area, and then proceeds to step SP16. On the other hand, if a positive result is obtained in step SP14, the main CPU proceeds to step SP17. Here, this error occurrence is set as an unnecessary area, and the process proceeds to step SP16.
[0132]
In step SP16, the main CPU determines whether or not there is another error. If a positive result is obtained here, the main CPU returns to step SP13. On the other hand, if no other error has occurred, the main CPU proceeds from step SP16 to step SP18 and determines whether a retry is necessary.
[0133]
Here, when the track set in which an error has occurred is set in the necessary area, the main CPU determines that a retry is necessary, moves to step SP19, instructs the retry process, and moves to step SP13. On the other hand, when all the track sets in which an error has occurred are set as unnecessary areas, the main CPU ignores the occurrence of the error in this case. That is, the main CPU proceeds directly from step SP18 to step SP13, and ends this processing procedure.
[0134]
As a result, as shown in FIG. 15 (1), when an error occurs in one track set after the joint during reproduction from the joint, the main CPU ignores this error by setting the subsequent joint flag. Can be processed. On the other hand, as shown in FIG. 15 (2), when an error occurs in the two track sets after the joint in the reproduction from the same joint, it is necessary because the joint writing flag cannot be detected in this case. A track set in which various data is recorded can be processed without ignoring errors.
[0135]
On the other hand, as shown in FIG. 15 (3), when an error occurs in the track set as in FIG. 15 (1) in the reproduction after one track set from the joint, in this case, The reproduced data itself cannot detect an error and causes no problem. On the other hand, as shown in FIG. 15 (4), when an error occurs in the track set as in FIG. 15 (2) in the playback after one track set from the joint, Since it cannot be detected, an error of a track set in which necessary data is recorded can be processed without being ignored.
[0136]
As shown in FIG. 15 (5), when an error occurs in two track sets across the joint in the case of playback from one track set before the joint, the two track sets are indicated by the joint flag. Can be handled ignoring errors. On the other hand, as shown in FIG. 15 (6), in the case where playback is performed from one track set before the joint, if an error occurs in two track sets straddling the joint and one track set that follows the joint, Since the write flag cannot be detected, it is possible to process without ignoring the error of the track set in which necessary data is recorded.
[0137]
Furthermore, as shown in FIG. 15 (7), when an error occurs in two track sets across the joint in the case of playback from the two track sets before the joint, the two track sets are indicated by the joint flag. Can be handled ignoring errors. On the other hand, as shown in FIG. 15 (8), when an error occurs in two track sets straddling the joint and the previous one track set during playback from the same track set, Since the flag cannot be detected, it is possible to process without ignoring the error of the track set in which necessary data is recorded. Also, as shown in FIG. 15 (9), when an error occurs in two track sets straddling the joint and the preceding and following track sets during playback from the same track set, the linkage flag is also set in this case. Since it cannot be detected, an error of a track set in which necessary data is recorded can be processed without being ignored.
[0138]
As a result, when data is reproduced from the middle of the magnetic tape, user data can be reliably sent to the host computer even if an error occurs in the reproduced data.
[0139]
According to the third embodiment, a track set in which unnecessary data is recorded is arranged before and after the boundary that is connected, and further connected to a track set at a predetermined position after the boundary that is connected. By setting a linkage flag indicating a boundary, user data can be reliably transmitted to the host computer even if an error occurs in the reproduced data.
[0140]
(6) Other embodiments
In the first and second embodiments described above, the case has been described in which errors are ignored based on the attributes of the track set on the back side of the track sets adjacent to the track set in which the error has occurred. However, the present invention is not limited to this, and an error is ignored on the basis of data recorded in this track set, which is an attribute of the preceding track set, among the track sets adjacent to the track set in which the error has occurred. Alternatively, the track set based on such a criterion may be switched and processed as appropriate according to the error occurrence location in one bank.
[0141]
In the third embodiment described above, the case where the linkage flag is set after the two-track set of the joint has been described. However, the present invention is not limited to this, and the essential point is that a certain relationship is established with respect to the joint. What is necessary is just to set a connection flag so that it may have, and it can set it in various places as needed.
[0142]
In the third embodiment described above, the case where only one type of identifier based on the linkage flag is set has been described. However, the present invention is not limited to this, and a plurality of types of identifiers that specify the boundary of linkage linkage are set. It is possible to specify the joint from any identifier. In this case, for example, it is conceivable to use, as the identifier, a count value that is sequentially incremented in units of track sets from the joint.
[0143]
In the third embodiment described above, a case has been described in which errors are simply ignored based on the linkage flag. However, the present invention is not limited to this, and the determination method according to the first embodiment is also included. May be applied.
[0144]
Further, in the above-described embodiment, a case where one track set is configured by four tracks and an error is determined in units of the track sets has been described. The present invention can be widely applied to a case where a track set is configured and an error is determined in units of the track set, and further, an error is determined and processed in units of tracks.
[0145]
In the above-described embodiment, the case where the present invention is applied to the tape streamer associated with the combination with the transport mechanism has been described. However, the present invention is not limited to this. Can also be widely applied.
[0146]
In the above-described embodiments, the case of recording computer data has been described. However, the present invention is not limited to this, and can be widely applied to, for example, processing video signals and audio signals.
[0147]
【The invention's effect】
As described above, according to the present invention, the continuity of the first identifier assigned to the recording tracks before and after the recording track in which the error is detected, Continue By ignoring errors by determining the attributes of recording tracks Re Even if an error occurs in the generated data, the user data can be reliably sent to the host device.
[Brief description of the drawings]
FIG. 1 is a perspective view showing a backup system according to an embodiment of the present invention.
FIG. 2 is a block diagram showing the backup system of FIG. 1;
3 is a front view and a rear view showing a tape streamer of the backup system of FIG. 1. FIG.
4 is a schematic diagram showing a recording format by the tape streamer of FIG. 3. FIG.
FIG. 5 is a schematic diagram for explaining inter-track interleaving in the tape streamer of FIG. 3;
6 is a schematic diagram for explaining an ECC block in the tape streamer of FIG. 3; FIG.
7 is a schematic diagram for explaining a track set in the tape streamer of FIG. 3; FIG.
8 is a schematic diagram for explaining a physical volume of the entire magnetic tape in the tape streamer of FIG. 3;
FIG. 9 is a schematic diagram for explaining each volume in the tape streamer of FIG. 3;
10 is a block diagram showing the tape streamer of FIG. 3. FIG.
11 is a schematic diagram for explaining error processing in the tape streamer of FIG. 3; FIG.
12 is a flowchart showing the processing procedure of FIG. 11. FIG.
FIG. 13 is a schematic diagram for explaining setting of a track set in a tape streamer according to a second embodiment of the present invention.
14 is a schematic diagram for explaining error processing in the tape streamer of FIG. 13; FIG.
FIG. 15 is a schematic diagram for explaining error processing in a tape streamer according to a third embodiment of the present invention;
16 is a flowchart showing the procedure of the process of FIG.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Backup system, 3 ... Tape streamer, 5 ... CPU unit, 6 ... Conveyance mechanism, 8 ... Host computer, 10 ... Magnetic tape, 23 ... Main CPU part, 24 ... Backup memory, 25 …… ECC encoder, 26 …… ECC decoder

Claims (2)

In a data processing apparatus that reproduces data recorded on a magnetic tape in units of a plurality of recording tracks, detects an error in the reproduced data, and outputs the reproduced data to an external device based on the detection result of the error ,
The recording track is
A first identifier One not a predetermined value only for the recording track for recording necessary data value is set to change, a second identifier indicating the attribute of the recorded data is recorded,
Of the plurality of recording tracks of the tape mark, unnecessary data is recorded on the first and last recording tracks, and / or at least one recording track is unnecessary at a portion where data is recorded following the end of the data. Data is recorded,
The data processing device includes:
According to the continuity of the first identifier assigned to the recording tracks before and after the recording track in which the error is detected, and the second identifier set in the recording track following the recording track in which the error is detected the decision must be disposed of the reproduced data by ignoring the detection of the error
Data processing equipment. A data reproduction method for reproducing data recorded on a magnetic tape by using a plurality of recording tracks as a unit, detecting an error in the reproduction data, and outputting the reproduction data to an external device based on the error detection result In
The recording track is
A first identifier One not a predetermined value only for the recording track for recording necessary data value is set to change, a second identifier indicating the attribute of the recorded data is recorded,
Of the plurality of recording tracks of the tape mark, unnecessary data is recorded on the first and last recording tracks, and / or at least one recording track is unnecessary at a portion where data is recorded following the end of the data. Data is recorded,
The method for reproducing the data is as follows:
The continuity of the first identifier assigned to the recording tracks before and after the recording track in which the error is detected, and the second identifier set in the recording track following the recording track in which the error is detected the decision must be disposed of the reproduced data by ignoring the detection of the error
The method of reproduction data.