JP2007323377A - Recording apparatus, method for writing management data and method for repairing management data - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To prevent increase in overhead in doubly recording management data in a recording medium and to suitably repair the management data when power supply is shut down. <P>SOLUTION: A request contents discrimination part 202 discriminates whether data to be written in an HDD 100 are management data or file data on the basis of information on partition. When a request is concerned with the management data, a double writing control part 204 writes the management data in a master management data area of the HDD 100 and then writes the management data in a slave management data area of the HDD 100. In a driver level, double writing processing is achieved. When the recording apparatus is started after power shut down, an area in which data are finally written when power supply is shut down is discriminated on the basis of final written LBA data in the HDD 100 and the master and slave management data are repaired on the basis of the discriminated result. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a recording apparatus that records file data on a recording medium such as an HDD (Hard Disk Drive) controlled by a driver based on a file system, and a management data writing method and a restoration method applied to the recording apparatus.

  Specifically, the present invention recognizes that the driver or the recording medium is management data based on the usage information of each area of the recording medium when the management data is written, and the management data is transferred to both the primary and secondary management data areas. This configuration relates to a recording apparatus or the like that prevents an increase in overhead due to double writing of management data.

  In addition to recording file data and primary / secondary management data on the recording medium, the present invention records data at the final writing position, and uses each area of the recording medium when the power is shut off during data writing. Based on the usage information and the data at the last writing position, it is determined which area of the recording medium was powered off, and the management data recorded in the primary and secondary management data area is restored according to the determination result Therefore, the present invention relates to a recording apparatus and the like that can appropriately repair management data.

  In recent years, there is a high need for digital video recording / playback in a system equipped with an HDD. In order to perform such digital video recording / playback (stream recording / playback), high throughput is required. To handle one high-quality stream, the system requires a throughput of 3 MB / sec. Higher throughput is required for file systems and HDDs. Further, when recording and reproducing a plurality of streams, a stricter throughput is required. Generally, as the number of recording and playback streams increases, the throughput tends to decrease exponentially.

  Household appliances equipped with file systems and HDDs have exploded in popularity, and the need for early recovery of equipment after power interruption due to external (such as power failure) and internal (such as abnormal temperature) failures is higher than PCs. Further, it is not desired that the recorded content is discarded by the process at the time of restoration.

  Here, there are (1) high throughput performance capable of handling a plurality of streams, (2) recovery / recovery from a high-speed failure occurrence, and (3) protection of recorded content from a failure as requirements for the file system and HDD.

  A typical technology aimed at speeding up file system recovery is the journaling file system. The main function of the journaling file system is the generation and recording of logs in write transactions.

  The content of the log is a change history of management data updated by a write transaction. Depending on the file system, the contents of the management data change are left as a history. The journaling file system refers to this log when recovering and repairing the file system, determines the management data that may have failed, checks the contents of the management data, and there is an inconsistency In this case, another content of the related management data is referred to, the management data having inconsistency is corrected, and the consistency of the file system is maintained.

  The above log is recorded on a recording medium as a log file. Here, the recording medium may be a recording medium on which management data is recorded or another recording medium. Data that the journaling file system writes to the recording medium is management data, file data, and log data.

  It is necessary to accurately observe the reflection order (writing order) of these three data on the recording medium. As an order, it is necessary to reflect management data after reflecting file data and log data on a recording medium. Otherwise, “the file data will be treated as a correct file with old contents”, “the gap between the log and the management data, which is the liver of the restoration process, will increase, and the restoration will take a long time”, “as a file system In order to maintain consistency, even normal files are discarded. ”

  Furthermore, there is no guarantee that the log file is in a normal state when a failure occurs, and there may be a case where a failure has occurred in the log file itself. The repair work in this case needs to check the contents of the management data without using a log to maintain consistency, and the recovery tends to increase time depending on the contents of the management data recorded.

  Thus, in order to use the journaling file system to satisfy the requirements of (2) recovery / recovery from high-speed failure occurrence and (3) protection of recorded content from failure for the file system and HDD. However, there is a problem that the request cannot be satisfied unless requirements such as the process of generating and recording the log (reflection order) must be performed and the log file is normal.

  In a system using a file system, the HDD is most used as a recording medium. This HDD is connected to a host controller (HostController) controlled by the HDD driver.

  When the set of these functional blocks is collectively referred to as an HDD unit (HDD Unit), in recent years, the HDD unit has been improved in technology for the purpose of increasing the capacity and speed. Regarding the increase in capacity, a capacity of 1 TB is expected several years after the current mainstream 500 GB, and the number of blocks to be managed by the file system increases.

  For high-speed technology, read / write requests to the HDD unit are buffered / cached in each block, the head seek is minimized within the HDD, read / write requests are issued and responses are efficiently read / written, etc. Efforts aimed at speeding up have been made.

  These are generalized as techniques such as reordering, NCQ (Native Command Queuing), TCQ (Tagged Command Queuing), and clustering. The heart of these technologies is queuing requests, optimizing for reading and writing, and issuing requests to the next functional block. Accordingly, the buffer / cache size of the HDD unit is also increasing.

  By queuing the request in each functional block, even if a read / write request from the application (Application) to the file system or HDD unit is made, a time lag occurs until the read / write to the recording medium is completed. As the buffer / cache size increases, the queued requests increase and the time lag also increases.

  As described above, in the HDD unit, the order of read / write requests is changed and aggregated by queuing, and a situation in which the requirements desired by the journaling file system are not satisfied occurs. Further, with respect to file system recovery / restoration when log files cannot be used, there is a problem that the time required for restoration increases because the capacity of the HDD increases and the number of objects to be managed increases.

  A possible solution to this problem is to synchronize the data that is forcibly buffered / cached with respect to the HDD unit to the recording medium. However, this is equivalent to not using queuing, and there is a problem that the speeding up of the HDD is sacrificed.

  As described above, (2) recovery / recovery from high-speed failure occurrence, (3) protection of recorded content from failure, and (1) multiple streams aimed by the journaling file system. It is technically contradictory to the high throughput performance that can be handled, and there is a trade-off relationship between devices that incorporate both.

  In addition, as the speed of HDDs is increasing as described above, the processing time (CPU processing) time of the file system and HDD driver is greatly affecting the overall throughput. Particularly in home appliances, the mainstream of a CPU incorporated therein is significantly inferior in performance (frequency and cache) to that of a PC, and is greatly affected. The next CPU processing is accommodated during the waiting time (IO waiting time) by reading and writing data to the HDD (IO processing) as long as the PC or the old HDD is used, but the powerless CPU and the high-speed HDD In this combination, the IO waiting time is short and the CPU processing tends to be long. In particular, it becomes a serious problem in a device that handles a plurality of streams.

In Patent Document 1, information (management data) indicating the location where the actual data of a file is recorded is recorded twice on a recording medium as regular information and preliminary information, and the regular information becomes faulty and cannot be read. In this case, it is described that the regular information can be restored using the preliminary information.
JP 2003-6017 A

  In the case where the file system management data is recorded twice on the recording medium as in Patent Document 1 described above, the problem that occurs when the file system is restored using log data as in the journaling file system described above is Disappear.

  However, each time file data (actual data) is recorded on a recording medium, management data needs to be recorded on the recording medium correspondingly, and this double writing processing is performed at the application level, and overhead There is a problem that becomes larger.

  Further, in the technique described in Patent Document 1, when the regular information becomes unreadable, the regular information is restored based on the preliminary information. It is not always in the correct state.

  An object of the present invention is to prevent an increase in overhead when file system management data is recorded twice on a recording medium. Another object of the present invention is to record management data twice on a recording medium, and to appropriately repair the management data when the power is turned off.

The concept of this invention is
A recording device for recording file data on a recording medium controlled by a driver based on a file system,
The recording medium has at least a file data area for recording the file data, a primary management data area and a secondary management data area for recording management data of the file system,
The driver or the recording medium has an information storage unit in which usage information of each area of the recording medium is stored.
When the management data is written, the driver or the recording medium recognizes that the management data is written based on the usage information stored in the information storage unit, and the management data is stored on the recording medium. The recording apparatus is characterized by writing to both the management data area and the sub management data area.

  In the present invention, file data is recorded on a recording medium based on the file system. This recording medium is controlled by a driver. The recording medium has at least a file data area for recording file data, and a primary management data area and a secondary management data area for recording file system management data, respectively.

  When file data is recorded in the file data area of the recording medium by the file system, management data is recorded in the primary and secondary management data areas of the recording medium in association therewith.

  In this case, when the driver or the recording medium is instructed to write the predetermined data to the predetermined area of the recording medium, the driver or the recording medium manages whether the predetermined area is a file data area based on the usage information of each area of the recording medium. Whether the predetermined data is file data or management data is recognized depending on whether it is a data area.

  When the predetermined data to be recorded is management data by the processing of the driver or the recording medium, the predetermined data is recorded in the predetermined area (primary management data area) and also in the sub management data area. . For example, the write position of the sub management data area (for example, LBA when the recording medium is an HDD) can be obtained by adding a preset offset value to the write position of the primary management data area.

  As described above, when writing the management data, the driver or the recording medium recognizes that it is management data based on the usage information of each area of the recording medium, and writes this management data to both the primary and secondary management data areas Since the processing for the double writing is performed at the driver level or the recording medium level, the overhead can be reduced as compared with the processing performed in the upper layer processing.

  The concept of the present invention is

  A recording device for recording file data on a recording medium based on a file system,

  The recording medium includes at least a file data area for recording the file data, a primary management data area and a secondary management data area for recording management data of the file system, and a final position data record for recording data of a final writing position, respectively. And having an area

  Any area of the recording medium based on the usage information of each area of the recording medium and the data of the final writing position recorded in the final position data recording area when the power is interrupted during data writing A management data restoration unit that judges whether or not the power supply has been shut down during writing to the management data and restores the management data recorded in the primary management data area and the secondary management data area according to the judgment result

  The recording apparatus is characterized by the above.

  In the present invention, file data is recorded on a recording medium based on the file system. This recording medium is controlled by a driver. The recording medium has at least a file data area for recording file data, a primary management data area and a secondary management data area for recording file system management data, respectively, and a final position data recording area for recording data of the final writing position. is doing.

  When the power is shut down during data writing, it is determined which area of the recording medium was powered off based on the usage information of each area of the recording medium and the data of the last writing position. The Then, the management data recorded in the primary / secondary management data area is repaired according to the determination result.

  For example, when the power supply is interrupted during writing to the primary management data area, the contents of the secondary management data area are copied to the primary management data area to be repaired. Further, for example, when the power is cut off when writing to the sub management data area, the contents of the main management data area are copied to the sub management data area, and the restoration is performed.

  As described above, the management data recorded in the primary / secondary management data area is repaired according to the determination result as to which power supply was interrupted when writing to which area of the recording medium, so that the primary / secondary management data is appropriately stored. The file system can be restored properly.

  According to the present invention, when writing the management data, the driver or the recording medium recognizes that it is management data based on the usage information of each area of the recording medium, and this management data is transferred to both the primary and secondary management data areas. The configuration is such that the overhead is increased due to double writing of management data.

  Further, according to the present invention, in addition to recording file data and primary / secondary management data on the recording medium, the data at the final writing position is recorded, and when the power is cut off, the usage information on each area of the recording medium and Based on the data at the final writing position, it is determined in which area of the recording medium the power was cut off, and the management data recorded in the primary / secondary management data area is restored according to this determination result. Yes, management data can be repaired appropriately.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 shows a configuration of a recording apparatus 10 as an embodiment. The recording apparatus 10 includes a software (SW) portion and a hardware (HW) portion.

  The SW part includes an application 106, a system call I / F 107, a file system 108, a block number (BlockNo) / LBA (Logical Block Addressing) conversion processing unit 110, , And an HDD driver 111.

  The HW portion includes a host controller 114 and an HDD 100 that constitutes a recording medium. The HDD driver 111 described above controls the HDD 100 via the host controller 114. The HDD 100 has a disk medium 104 in which data is stored. The file data 101, the management table 102, and the partition table (Partition Table) 103 are recorded on the disk medium 104, and the last written LBA data is stored in a nonvolatile memory space. 105 is recorded.

The above-mentioned data recorded in the HDD 100 is kept permanently even when the power is turned off. In the management table 102, a plurality of management data 109 are stored for each data type. The data type of the management data 109 includes, for example, a free block management table, a file name, a storage destination management table, and a file system version number.
In this embodiment, the management table 102 is written twice, but for the sake of explanation, each is identified as primary and secondary.

  FIG. 2 shows the partition structure of the HDD 100. There are a minimum of three partitions. One is a positive storage destination of the management table 102, one is a secondary storage destination of the management table 102, and one is a storage destination of the file data 101 that is the contents of the file, the primary management data area 301, and the secondary management A data area 302 and a file data area 303 are assumed. The primary management data area 301 and the secondary management data area 302 have the same partition size.

  In FIG. 2, there are four partitions, a primary management data area 301 and a secondary management data area 302 are set in the middle of the LBA, and a file data area 303 exists so as to sandwich these management data areas 301 and 302. The file data area 303 is divided into two partitions, but is handled as a continuous LBA.

  In FIG. 1, an application 106 is a program that handles files, and owns file data 112. An operation on the file is executed by calling a function of the system call I / F 107 from the application 106.

  The system call I / F gives the file system 108 a function for the file operation from the application 106 to the file system 108, the file identification, the memory address where the file data 112 is stored, and the file data 112 at the time of reading. The memory address and data size for storage and options / flags for processing are called as arguments.

  The file system 108 reads necessary management data 109 from the management table 102 recorded in the HDD 100 at the time of initialization. In the case of writing, the file system 108 refers to and searches the management data 109 according to the function called from the system call I / F, and manages the file system 108 in which the management data 109 to be operated is stored. Specify the unit block number.

  The file system 108 performs necessary rewriting for the block number. If there is no management data 109 to be operated, generation is performed. When writing the file data 112 or the management data 109 to the HDD 100, the flock number, which is the management unit of the file system 108, is transferred to the block number / LBA conversion processing unit 110 and converted to LBA. The block number / LBA conversion table is created as one type of management data when the file system 108 is constructed.

  The block number / LBA conversion processing unit 110 collects information necessary for reading and writing such as the converted LBA and the storage memory address of the file data 112 or the management data 109 as a request and passes it to the HDD driver 111. The HDD driver 111 determines the request content based on the information of the request, performs double writing processing as necessary, and controls the host controller 114 according to the content. The host controller 114 transmits and receives data to and from the HDD 100.

  FIG. 3 shows an example in which the function of double writing management data is implemented in the HDD driver 111. As shown in FIG. 3, the HDD driver 111 includes a write request queue (queue) 201, a request content determination unit 202, a partition usage information storage unit 203, a dual write control unit 204, and a command generation unit 205. And a host controller control unit 206.

  The write transaction will be described with reference to FIGS. When the application 106 writes the file data 112 to the HDD 100, the file system 108 rewrites the necessary management data 109, and uses the rewritten management data 109 as a request to the HDD 100 different from the file data 112. To the write request queue 201.

  The request includes address information as to which LBA of the HDD 100 the management data 109 is to be written. The HDD driver 111 refers to the partition table 103 and the partition usage information storage unit 203. Here, the partition table 103 simply has information on the partition of each partition, but the partition usage information storage unit 203 further records a usage table for each partition, that is, an area for recording file data. It has information such as whether it is an area to do.

  The request content determination unit 202 determines whether the data written to the HDD 100 is the management data 109 or the file data 112 based on the information related to the partition. If the request is for the management data 109, the management data 109 is written into the primary management data area 301.

  After the management data 109 is written to the primary management data area 301, the primary management data is transferred to the LBA of the HDD 100 in the content of the request used when writing the management data 109 to the primary management data area 301. The LBA offset of the area 301 and the sub management data area 302 is added, and the data is written to the sub management data area 302.

  In this case, these processes intended to reflect the contents recorded in the primary management data area 301 to the secondary management data area 302 are realized by the dual write control unit 204, the command generation unit 205, and the host controller 206. The

  If the management data 109 has not been written to the primary management data area 301, the management data 109 is not written to the secondary management data area 302. If the management data 109 has not been written to the secondary management data area 302, the management data 109 of the next request is not written to the primary management data area 301. In this way, synchronization is established between writing the management data 109 to the primary management data area 301 and the secondary management data area 302. That is, the primary management data area 301 and the secondary management data area 302 are in the same writing state.

  On the other hand, if the request is for the file data 112, the file data is written in the file data area 303.

  Also, the last written LBA data recording unit 207 of the HDD 100 calls the first LBA to be written and the write size from the content of the request as the last written LBA data 105 and records it in a non-volatile area. In this case, the last written LBA data 105 may be recorded in a predetermined area of the disk medium 104 or may be recorded in a separately provided non-volatile memory.

  FIG. 4 shows an outline of the above-described processing when the management data 109 and the file data 112 are written to the HDD 100. In this writing process, when the management data 109 is written to the HDD 100, the contents written in the primary management data area 301 are reflected in the secondary management data area 302, and mirroring at the partition level is performed.

  Next, in the recording apparatus 10 shown in FIG. 1, the restoration of management data when a failure occurs during a write transaction and the power is cut off will be described. FIG. 5 shows a management data recovery pattern by power-off. There are pattern 1 and pattern 2 as the repair patterns. Pattern 1 is a case where the power-off timing is being written to the primary management data area, and the destruction status is inconsistent with the contents of the primary management data area. Adopts a method of copying the contents of the secondary management data area to the primary management data area. In this case, the state after the restoration does not reflect the data of the last issued write command.

  Pattern 2 is when the power-off timing is during writing to the secondary management data area, and the destruction status is inconsistent with the contents of the secondary management data area. Adopts a method of copying the contents of the primary management data area to the secondary management data area. In this case, the data after the last issued write command is reflected as the state after the restoration.

  With reference to FIGS. 5A and 5B, the file system recovery process, that is, the management data repair process in the present embodiment will be further described.

  When the recording apparatus 10 is started after the power is shut off, the last written LBA data 105 is acquired from the HDD 100, the partition information and the usage information of the partition are acquired, and the last written LBA (last written LBA) is correct. The management data area 301, the secondary management data area 302, or the file data area 303 is determined.

  If the last written LBA is the file data area 303, the management system is not restored, and the file system is mounted based on the management data stored in the primary management data area 301.

  When the last written LBA is the primary management data area 301, all the data in the secondary management data area 302 is copied to the primary management data area 301, and then the management stored in the primary management data area 301 is stored. Mount the file system based on the data. The management data repair pattern in this case corresponds to the above-described pattern 1 of FIG.

  FIG. 6A shows the state of the file system when the last written LBA is the primary management data area 301, and FIG. 6B shows the state of the file system after repair. In FIG. 6, the corresponding file data is written to the HDD 100 after the management data is written. The same applies to FIG. 7 described later.

  When the last written LBA is the secondary management data area 302, all the data in the secondary management data area 302 is copied to the primary management data area 301 and then the management stored in the primary management data area 301 is stored. Mount the file system based on the data. The management data repair pattern in this case corresponds to the above-described pattern 2.

  FIG. 7A shows the state of the file system when the last written LBA is the secondary management data area 302, and FIG. 7B shows the state of the file system after repair.

  Note that the above-described repair processing can be implemented by software or as a firmware in the HDD 100.

  According to the recording apparatus 10 shown in FIG. 1, the HDD driver 100 is provided with a management data double-writing function. When the management data 109 is written, the management data 109 is managed in the primary and secondary management of the HDD 100 by the function. Double writing is performed in the data areas 301 and 302, and the overhead can be reduced as compared with the processing performed in the upper layer processing.

  At this time, the LBA offset of the primary management data area 301 and the secondary management data area 302 is added to the LBA of the HDD 100 in the content of the request used when writing to the primary management data area 301, and the secondary management data area The management data 109 is written to 302, and double writing can be realized with less CPU processing time.

  Further, according to the recording apparatus 10 shown in FIG. 1, when a power interruption occurs during a write transaction, any one of the HDDs 100 is determined based on the information of the last written LBA data recording unit 207 in the HDD 100 at the subsequent startup. Since it is determined whether or not the power has been shut down when writing to the area, and the management data is repaired according to the determination result (see FIGS. 5 to 7), the management data can be repaired appropriately. Recovery can be performed appropriately.

  In the above-described embodiment, the dual write function is implemented in the HDD driver 111 (see FIG. 3). However, it is conceivable that the dual write function is implemented in the HDD 100.

  FIG. 8 shows an example in which the function of double writing management data is implemented in the HDD 100. As shown in FIG. 8, the HDD 100 has a command queue 401, a command content determination unit 402, a partition usage information storage unit 403, a dual write control unit 404, and a media write control unit 405. is doing.

  A write command received from the HDD driver 114 via the host controller 114 in the write transaction enters the command queue 401. This command includes address information indicating which LBA of the HDD 100 is to be written. The command content determination unit 402 of the HDD 100 refers to the partition table 103 and the partition usage information storage unit 403 to determine whether the data to be written is management data 109 or file data 112.

  If the write command is for the management data 109, the management data 109 is written into the primary management data area 301. After the management data 109 is written to the primary management data area 301, the primary management data is transferred to the LBA of the HDD 100 in the content of the request used when writing the management data 109 to the primary management data area 301. The LBA offset of the area 301 and the sub management data area 302 is added, and the data is written to the sub management data area 302.

  The process intended to reflect the content recorded in the primary management data area 301 to the secondary management data area 302 is realized by the dual write control unit 404 and the media write control unit 405.

  If the write command is for the file data 112, the file data 112 is written to the file data area 303.

  As described above, even if the dual write function is implemented in the HDD 100 as described above, the dual write function is implemented by higher-level processing as in the case where the dual write function is implemented in the HDD driver 111. Compared to the overhead can be reduced.

  In the above-described embodiment, the present invention is applied to the recording apparatus 10 in which the recording medium is the HDD 100. However, the present invention can be similarly applied to apparatuses using other recording media. is there.

  The present invention can prevent an increase in overhead due to double writing of management data, and can appropriately repair the management data when the power is cut off. The present invention can be applied to a recording apparatus that records data on a recording medium such as an HDD.

1 is a block diagram illustrating a configuration of a recording apparatus as an embodiment. It is a figure for demonstrating the partition structure of HDD. It is a figure which shows the functional block of a HDD driver at the time of mounting the dual write function in a HDD driver. It is a figure which shows the outline of the write-in process of management data and file data. It is a figure which shows the restoration pattern of the management data by power supply interruption. It shows the state before and after the restoration of the file system when the last written LBA is the primary management data area. It shows the state before and after the restoration of the file system when the last written LBA is the secondary management data area. It is a figure which shows the functional block of HDD when the function of dual writing is mounted in HDD.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... Recording device, 101 ... File data, 102 ... Management table, 103 ... Partition table, 104 ... Disk medium, 105 ... Last write LBA data, 106 ... Application, 107: System call I / F, 108: File system, 109: Management data, 110: Block number / LBA conversion processing unit, 111: HDD driver, 112: File data, 113 ... Partition table, 114 ... Host controller, 201 ... Write request queue, 202 ... Request content determination unit, 203 ... Partition usage information storage unit, 204 ... Double write control , 205... Command generation unit, 206. ..., Last write LBA data recording unit, 301... Primary management data area, 302... Secondary management data area, 303... File data area, 401. ..Command content determination unit, 403... Partition use application information storage unit, 404... Dual write control unit, 405.

Claims (6)

A recording device for recording file data on a recording medium controlled by a driver based on a file system,
The recording medium has at least a file data area for recording the file data, a primary management data area and a secondary management data area for recording management data of the file system,
The driver or the recording medium has an information storage unit in which usage information of each area of the recording medium is stored.
When the management data is written, the driver or the recording medium recognizes that the management data is written based on the usage information stored in the information storage unit, and the management data is stored on the recording medium. A recording apparatus that writes in both the management data area and the sub management data area. File data is recorded on a recording medium controlled by a driver based on the file system, and the recording medium records file data for recording the file data and a primary management data area and a secondary management data area for recording management data of the file system, respectively. A method of writing management data in a recording apparatus having a management data area,
Whether the driver or the recording medium is instructed to write the management data to the predetermined area of the recording medium by the file system is based on the usage information of each area of the recording medium. A determination step for determining;
A write step in which the driver or the recording medium sequentially writes the predetermined data in the primary management data area and the secondary management data area when it is determined in the determination step that the management data is written. To write management data. A recording device for recording file data on a recording medium based on a file system,
The recording medium includes at least a file data area for recording the file data, a primary management data area and a secondary management data area for recording management data of the file system, and a final position data record for recording data of a final writing position, respectively. And having an area
Any area of the recording medium based on the usage information of each area of the recording medium and the data of the final writing position recorded in the final position data recording area when the power is interrupted during data writing A management data restoration unit that judges whether or not the power supply has been cut off during writing to the management data and restores the management data recorded in the primary management data area and the secondary management data area according to the judgment result A recording apparatus. The management data restoration unit copies the contents of the secondary management data area to the primary management data area when the power is shut down during writing to the primary management data area of the recording medium. The recording apparatus according to claim 3. The management data restoration unit is configured to copy the contents of the primary management data area to the secondary management data area when the power is shut down during writing to the secondary management data area of the recording medium. The recording apparatus according to claim 3. File data is recorded on a recording medium based on the file system, and the recording medium records file data for recording the file data, a primary management data area and a secondary management data area for recording the management data of the file system, respectively. A management data recovery method in a recording apparatus having a final position data recording area for recording data of a final writing position,
When power is interrupted during data writing, writing to any area of the recording medium based on usage information of each area of the recording medium and data of the final writing position recorded on the recording medium A judgment step for judging whether or not the power supply has been shut off during,
A management data repairing method comprising: a repairing step of repairing management data recorded in the primary management data area and the secondary management data area in accordance with a judgment result of the judgment step.

Images