JP2010176201A - Data loss prevention device for flash memory, and data loss prevention method for flash memory - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a data loss prevention device for a flash memory and a data loss prevention method for the flash memory, preventing the occurrence of data loss in the flash memory. <P>SOLUTION: One of a pair of blocks of the flash memory 6 is used as a rotation block for use in normal data writing, and the other is used as a mirror block for writing the same data as the rotation block. When an write error occurs in the rotation block in use, subsequent data read/write is carried out in the mirror block to continue data read/write after the occurrence of the write error. During initialization after a power supply is turned on again, a data group written into the mirror block is copied to the rotation block, and a mirror is reconstructed. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a flash memory which is a kind of non-volatile rewritable memory, and more particularly to a data loss prevention device and a data loss prevention method.

  2. Description of the Related Art Conventionally, various memories are mounted as storage destinations of various data necessary for operation in a computer that comprehensively manages the operations of apparatuses and devices. As one type of this memory, for example, a nonvolatile memory is used as a memory capable of electrically erasing and writing various data. For example, a flash memory (Flash Read Only Memory: as shown in FIG. 7) is used as this nonvolatile memory. 81) is widely used. The flash memory 81 is constructed by an aggregate of a plurality of blocks 82 to 85 (four in FIG. 7 from A to D) as a data writing destination, and data can be erased for each block unit. These blocks 82 to 85 are also referred to as erase blocks because they are data erasure units. Data reading / writing with respect to the flash memory 81 is managed by a driver 87 that operates according to a command from the application 86.

  By the way, EEPROM (Electrically Erasable PROM) and SRAM (Static Random Access Memory), which are other memory types than the flash memory 81, have an overwrite function capable of sequentially writing data on the same address, so that data has already been written. Even if a data write command is issued to this address, data can be sequentially written to this address. However, the above-described flash memory 81 has a characteristic that, when new data is written to an address where data has been written, this operation cannot be executed unless the entire block is erased once. There is.

  Therefore, in this type of flash memory 81, when data is written, the blocks 82 to 85 of A to D are rotated and used as shown in FIGS. At this time, when new data is first written to the flash memory 81, the driver 87 sequentially writes data in the empty area of the used block (A block 82 in FIG. 8) as shown in FIG. Then, when there is no free space in the A block 82, as shown in FIG. 9, the latest value in each data (ID1 to ID3 in FIG. 9) is copied to the next used block B block 83, and this B block 83 is used as a new data writing destination, and data writing is performed by sequentially repeating this operation each time there is no empty area in the used block. In addition, data is erased from a used block whose free area has run out before the next rotation.

JP 2008-243347 A

  By the way, when a write error to the used block occurs in the driver 87 as shown in FIG. 10 due to various factors, the driver 87 cannot execute data writing to the used block thereafter. As a result, even if the driver 87 receives new data from the application 86, the driver 87 cannot perform an operation of writing the new data to the flash memory 81, resulting in a problem of data loss. In addition, the problem of data loss includes, in a broad sense, that the data written in the used block is erased by the driver 87 that has operated in error. In short, it is a new data that can normally hold data in the flash memory 81. A countermeasure was requested.

  An object of the present invention is to provide a data loss prevention device for a flash memory and a data loss prevention method for a flash memory that can make it difficult to cause data loss in a flash memory.

  In order to solve the above problems, in the present invention, when a flash memory is used as a rewritable nonvolatile memory and a driver writes data to the flash memory by a write command from an application, the memory of the flash memory Among the two blocks set as a set in the data loss prevention device of the flash memory that reads and writes data while sequentially rotating and using a plurality of blocks that are data erasing units while constructing an area One of these is used as a rotation block used in the rotation, and the other is used as a mirror block of the rotation block. And a protection execution means for continuing reading and writing of the data using the mirror block when the abnormality is confirmed by the monitoring means. .

  According to this configuration, a plurality of blocks that the flash memory has as a memory area is divided into two, and one block in each set is used as a rotation block used for data reading and writing, and the other block is a mirror of the rotation block. Use as a block. When a driver access to the flash memory, that is, when an access abnormality occurs in the rotation block in use, data reading and writing is continued by using the mirror block of the rotation block in which the abnormality has occurred. . For this reason, even if an abnormality occurs in the used rotation block, it becomes possible to continue data reading and writing using the mirror block, so that it is difficult to cause a situation where data cannot be continuously read and data is lost. It becomes possible to do.

  In the present invention, when the abnormality is confirmed by the monitoring unit, the protection execution unit includes a data processing continuation unit that continues to read and write the data in the mirror block, and a mirror that reconstructs the mirroring destroyed by the abnormality. The gist of the present invention is that it comprises a reconstruction means.

  According to this configuration, when data read / write is continued in the mirror block of the same block in the case of abnormal data read / write in the rotation block, data processing can be continued, but mirroring between the rotation block and the mirror block can be performed. The situation is destroyed. However, in this configuration, when mirroring is destroyed, mirroring between the two blocks is reconstructed by the mirror reconstructing means, so after that, even when returning to the operation of reading and writing data in the rotation block, Normal data reading and writing can be executed without problems.

  In the present invention, when the mirror rebuilding unit receives a new data write command from the application during the rebuilding of the mirroring, the mirror rebuilding unit suspends the rebuilding and stores the data in the free memory area. The main point is that data can be written during mirroring reconstruction.

  According to this configuration, even if new data writing is accepted when rebuilding the mirroring that has been destroyed, the rebuilding is temporarily suspended and data writing is executed with priority. It becomes possible to do.

  In the present invention, a flash memory is used as a data rewritable nonvolatile memory, and when a driver writes data to the flash memory in accordance with a write command from an application, a data erasing unit is constructed while building a memory area of the flash memory. In a method for preventing data loss in a flash memory that reads and writes data while using a plurality of blocks that are rotated in order, rotation using one of the two blocks set as a set in the rotation Use as a block, use the other as a mirror block of the rotation block, and monitor whether or not there is an abnormality with respect to the driver access to the flash memory, and the monitoring means When confirming the normal, and summarized in that to continue to read and write the data using the mirror block.

  According to the present invention, it is possible to make it difficult for data loss to occur in a flash memory.

The perspective view which shows the external appearance of the driver's seat in a vehicle in one Embodiment. The block diagram which shows schematic structure including the software configuration of a computer. Explanatory drawing which shows operation | movement transition when data are written in a block. Explanatory drawing which shows the specific example when reading data from a block. Explanatory drawing which shows a state when the write error generate | occur | produced in the block. Explanatory drawing which shows operation | movement transition when rebuilding mirroring. The block diagram which shows schematic structure of the computer in the past. Explanatory drawing which shows operation | movement when writing data in the block of flash memory. Explanatory drawing which shows operation | movement when erasing data from the block of flash memory. Explanatory drawing which shows a state when a write error has occurred.

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a flash memory data loss prevention apparatus and flash memory data loss prevention method embodying the present invention will be described below with reference to FIGS.
As shown in FIG. 1, the vehicle 1 is provided with a steering angle detection device 3 that detects the steering angle of the steering wheel 2. The steering angle detection device 3 includes a steering angle sensor 4 as a sensor component (for example, a magnetic sensor) of the device 3 and a computer 5 as a control unit of the steering angle detection device 3. The computer 5 calculates the steering angle of the steering wheel 2 based on the sensor output value acquired from the steering angle sensor 4, and uses the calculated steering angle (calculated steering angle value) to other in-vehicle devices that need it. Output via a LAN (Local Area Network).

  As shown in FIG. 2, the computer 5 is provided with a flash memory (Flash ROM) 6 as a memory for writing various data. The flash memory 6 is a kind of nonvolatile memory capable of rewriting data in its own memory area. As described in the background art, the flash memory 6 is divided into a plurality of data erase units (in this example, A to F 6). Block) 7-12. That is, this type of flash memory 6 is a memory having a characteristic that data erasure can be executed only in block units, although data written in its own memory area can be rewritten. These blocks 7 to 12 have the same storage capacity by equally dividing the memory area of the flash memory 6. The A block 7, the C block 9 and the E block 11 constitute a block (rotation block).

  In the flash memory 6 of this example, when data is written, three blocks A, C, and E, one of the blocks A to F that are skipped, are rotated and used as data write destinations. To do. Here, of these three rotation blocks 7, 9, and 11, if the first block to be used is, for example, A block 7, first, various data related to the steering angle detection device 3 are set using the A block 7 as a data write destination. Write. The data written in the A block 7 at this time includes, for example, the time (ID1 data) when the engine is started by operating the ignition switch (engine switch) of the vehicle 1 to the ignition on position (IG ON position). The time during which the ignition switch is on (ID data 2), the sensor output value of the steering angle sensor 4 (ID3 data), and the like.

  By the way, when data is written to the flash memory 6, the data cannot be overwritten on the data-written address as a characteristic of the flash memory 6. Therefore, when new write data is received, the flash memory 6 performs an operation of sequentially increasing the received data in the empty area in the A block 7 in use. When the writing area becomes full, the used block is switched to the next used block (C block 9 in this example), and the latest value of each data of ID1 to ID3 is written to the C block 9 while writing the data to the C block. The data write is executed by continuing this operation and repeating this operation every time there is no free area in the block. It should be noted that data erasure is executed for a block in which data writing is full, when there is no free time during which data writing or data reading is not performed, until the next rotation starts.

  As shown in FIG. 2, the computer 5 includes, as software for operating the computer 5, an application 13, which is dedicated software for operating the steering angle detection device 3, and an application 13 and an OS (Operating System). There are provided middleware 14 that functions in an intermediate position, and a driver 15 that is software for operating the flash memory 6. The application 13 operates the computer 5 by outputting an operation command to the driver 15 via the middleware 14 based on various information and data input to the application 13. In addition, when there are a plurality of drivers 15, the application 13 can collectively manage these drivers 15.

  When the driver 15 receives an operation command from the application 13 via the middleware 14, the driver 15 operates according to the operation command and executes various processes such as data writing, data reading, and data erasing to the flash memory 6 (blocks 7 to 12). To do. At this time, the driver 15 receives an operation command from the application 13 as a trigger, and executes an operation based on this operation command regardless of the operation state of the application 13 even if the application 13 is executing another operation. Further, the data erasure of the blocks 7 to 12 by the driver 15 is executed when the driver 15 is free when the driver 15 is not performing data writing or data reading as described above.

  In the case of this example, the B, D, and F blocks 8, 10, and 12, which are skipped by one rotation, are used as mirror blocks that mirror the data group written in each block. The blocks 8, 10, and 12 in this example, which are mirror blocks, are blocks in which the same data as the data written in the rotation blocks forming a set is written in the same address order. In this example, the B block 8 is a mirror block of the A block 7, the D block 10 is a mirror block of the C block 9, and the F block 12 is a mirror block of the E block 11. The B block 8, D block 10, and F block 12 form a block (mirror block).

  The driver 15 is provided with a mirroring processing unit 16 that performs mirroring on the flash memory 6. The mirroring processing unit 16 functions so that two rotation blocks having the same configuration are generated in the flash memory 6 by logically dividing the memory area of the flash memory 6 into two. That is, the mirroring processing unit 16 sets a plurality of blocks 7 to 12 as a set of two, and in each of these sets, a rotation block that uses one block for actual data reading / writing (in this example, blocks 7, 9, 11), and the other block is used as a mirror block (blocks 8, 10, and 12 in this example) used for mirroring. The mirroring processing unit 16 corresponds to a mirroring processing unit.

  At the end of each of the blocks 7 to 12, a management area E that describes what the mirroring state of the block is is prepared. In this management area E, management information 17 for notifying what state the mirroring state of the own block is in is written. The management information 17 includes, for example, a mirror state normal notification that notifies the driver 15 that the mirroring state is normal, and a mirror state abnormality notification that notifies the driver 15 that the mirroring state is abnormal.

  Further, the driver 15 is provided with a data loss presence / absence confirmation unit 18 for confirming whether or not there is data loss with respect to the used rotation block. The data loss presence / absence confirmation unit 18 of the present example confirms whether or not the data writing can be normally executed when the driver 15 is executing data writing to the rotation block being used, thereby checking the rotation block being used. For example, when a write destination address is lost, it is recognized that a write error has occurred. When confirming a write error, the data loss presence / absence confirmation unit 18 updates the management information 17 from the previous mirror state normal notification to the mirror state abnormal notification. In addition, the data loss presence / absence confirmation unit 18 corresponds to a monitoring unit.

  The driver 15 is provided with a data processing continuation unit 19 that executes data writing and reading after the error detection in the mirror block forming a pair when the data loss presence / absence confirmation unit 18 detects a writing error in the rotation block. Yes. For example, when the data processing continuation unit 19 performs data writing in the A block 7 and confirms that a writing error has occurred in the A block 7, the data processing continuation unit 19 uses the B block 8 that is a mirror block of the A block 7. Continue to write or read data. The data processing continuation unit 19 constitutes a protection execution unit and a data processing continuation unit.

  Further, the driver 15 is provided with a mirror reconstruction unit 20 for reconstructing mirroring using a data group held in a mirror block corresponding to a write error in a rotation block in use. Yes. The mirror rebuilding unit 20 of this example checks the management information 17 at the read timing of the management information 17 such as when the computer 5 is turned on, and writes a mirror state abnormality into the mirror block when receiving a mirror status abnormality from the management information 17. The mirroring is reconstructed by copying the data group stored in step 4 to the rotation block that is paired with it. The mirror reconstruction unit 20 constitutes a protection execution unit and a mirror reconstruction unit.

Next, the mirroring operation executed by the computer 5 (driver 15) of this example will be described with reference to FIGS.
For example, when the driver 15 receives a data write command from the application 13 and newly writes data to the flash memory 6, as shown in FIG. Data is written in order. In the following description, the data that the driver 15 sequentially writes in the A block 7 (that is, various data of ID1 to ID3) is described as the first data D1, the second data D2,. .

  At this time, the mirroring processing unit 16 applies the same data content as the data written in the A block 7 (that is, the first data D1) to the B block 8 which is the mirror block of the A block 7 from the same address from the top. The data is written in order, and this A block 7 is mirrored by the B block 8. At this time, when the mirroring processing unit 16 confirms that the same first data D1 as that in the A block 7 can be normally written in the B block 8, the mirroring unit 16 recognizes that the mirror state is normal, and is provided at the end of the A block 7. The mirror state normal notification is recorded as the management information 17 and is also mirrored, and the same mirror state normal notification is recorded at the end of the B block 8. Then, when the driver 15 takes the operation of writing the second data D2, the third data D3, the fourth data D4,... In the A block 7 in order, the mirroring processing unit 16 sends the same data to the B block 8 in the same address order. Write sequentially.

  When the driver 15 receives a data read command from the application 13 during the data writing, the driver 15 reads data from the flash memory 6 according to the command. At this time, data reading of the flash memory 6 is executed only from the A block 7 which is a rotation block, as shown in FIG. As a result, the data read operation does not take the operation of reading data from both the A block 7 and the B block 8, so that the data read processing time can be shortened.

  By the way, as shown in FIG. 5, if an abnormality occurs in the data writing environment, for example, when a problem occurs in the format of the A block 7, the driver 15 may accept write data from the application 13. This cannot be written into the A block 7, and a write error occurs. After such a writing error occurs, data writing cannot be continued in the A block 7 that is a rotation block, which may lead to data loss as a result.

  However, in the case of this example, the presence or absence of a write error is monitored by the data loss presence check unit 18. Then, when the data loss presence / absence confirmation unit 18 confirms that the writing cannot be completed even after the predetermined time has elapsed even though the driver 15 enters the writing operation to the A block 7, for example, a writing error has occurred in the rotation block. Recognize. At this time, when confirming a write error, the data loss presence / absence confirmation unit 18 updates the management information 17 from the previous mirror state normal notification to the mirror state abnormal notification. Since the entire blocks 7 to 12 can be erased only in a lump, the mirror state abnormality notification is added to the empty area of the management area E and added to the mirror block.

  In addition, when the data processing continuation unit 19 detects a write error in the data loss presence / absence confirmation unit 18, data reading / writing is continued in the B block 8 that is a mirror block of the A block 7 instead of the A block 7 thereafter. . Thereafter, the A block 7 is left to be removed from the data read / write rotation, and the B block 8 is used as the rotation block. That is, when the writing in the B block 8 is full, the blocks are used in the order of the B block 8 → D block 10 → F block 12, and when the F block 12 runs out of free space, the write destination becomes B Returning to block 8, the above operation is repeated.

  Here, when the power is turned on again after the computer 5 is turned off, the mirror rebuilding unit 20 stores the management information 17 written at the end of the mirror block in order to rebuild the broken mirror. I'll go check. The reason why the mirror reconstruction is executed at the initialization after the power is turned back on is that the driver 15 takes an operation to check the management information 17 at the end of the block at the initialization. When the mirror reconstruction unit 20 confirms the management information 17 and reads the mirror state normality notification, the mirror reconstruction unit 20 recognizes that the mirroring is not broken and does not execute the mirror reconstruction operation.

  On the other hand, when the mirror reconstruction unit 20 confirms the management information 17 and reads the mirror state abnormality notification, the mirror reconstruction unit 20 recognizes that the mirroring has been broken, and reconstructs the broken mirror according to the procedure shown in FIG. Start building. At this time, since the flash memory 6 can erase data only in units of blocks, the mirror reconstruction unit 20 first erases the A block 7. After erasing the A block 7, the mirror reconstruction unit 20 writes all the data contents held in the B block 8 into the A block 7 as they are, that is, copies them. In other words, the mirror reconstruction unit 20 performs an operation of copying the data group written in the mirror block to the rotation block in order from the first address.

  Note that this copying is an operation that requires time to sequentially write the data of the B block 8 to the A block 7, so that a new data write command may be received from the application 13 during the copying. At this time, the mirror reconstruction unit 20 temporarily interrupts the mirror reconstruction being performed at that time, and new data (the fifth in FIG. 6) is added to the vacant area in both the A block 7 and the B block 8. After the data D5) is written and written, the mirror reconstruction is continued. The mirror rebuilding unit 20 may receive a data read command from the application 13 at the time of mirror rebuilding. At this time, the mirror rebuilding is temporarily interrupted to cope with this.

  Then, when all the data has been copied, the mirror reconstruction unit 20 writes a mirror state normality notification to the management area E of both the A block 7 and the B block 8, and stores the management information 17 Update and record that the mirror status is normal. Thereby, after the power is turned on again, the A block 7, the C block 9, and the E block 11 are rotated as usual, and the reading and writing of data is executed. In addition, although the mirroring of A block 7 and B block 8 was demonstrated here, since operation of both mirroring of C block 9 and D block 10 and mirroring of E block 11 and F block 12 is the same, explanation is Omitted.

  In this example, one of the two blocks forming the set is used as a rotation block, and the other is used as a mirror block, so that data is mirrored for each rotation block. For this reason, for example, even if a write error occurs in the rotation block, the subsequent data writing is continued in the mirror block, so that no data leakage occurs, and as a result, data loss is prevented. In addition, since the mirror is reconstructed at the initialization after the power is turned on again, it is possible to restore the rotation block to the state where the regular data has been written. It becomes possible.

According to the configuration of the present embodiment, the following effects can be obtained.
(1) The same block as the rotation block used for normal data read / write is provided as a mirror block, and when a write error occurs in the rotation block being used, data read / write is continued using the mirror block thereafter. . For this reason, even if a write error occurs in the rotation block, it is possible to make it difficult to cause a situation where data cannot be continuously written and data is lost.

  (2) When mirroring is destroyed, mirroring reconstruction is executed at initialization after power-on, so that these two blocks are restored to the original state in which the rotation block and mirror block have the same data contents. Can do.

  (3) When a new data write command is received from the application 13 while the mirroring reconstruction is being executed, the mirroring is temporarily interrupted and a new data write is executed. Therefore, even if a new data write request is generated during mirroring reconstruction, it can be handled without any problem.

  (4) When reading data from the flash memory 6, data is read only from the rotation block of the two pairs of blocks. For this reason, since the operation of reading data from both the rotation block and the mirror block is not performed, it does not take time to read the data.

  (5) The management information 17 for notifying the mirror state is written at the end of the blocks 7 to 12, and the management information 17 is checked to determine whether the mirroring is normal. Therefore, confirmation of whether the mirror state is normal can be executed by a simple method of viewing the management information 17.

Note that the embodiment is not limited to the configuration described so far, and may be modified as follows.
-Handling data loss is not necessarily limited to a method in which data writing is continued in the mirror block when a write error occurs and the mirror is reconstructed at initialization. For example, when the driver 15 detects data loss in the used block, the erased data may be recovered by copying the data previously held in the mirror block to the used block.

  -The target of detecting an access error is not necessarily limited to a write error. For example, the driver 15 sees whether or not the access destination for the flash memory 6 has been lost, the flash memory 6 (blocks 7 to 9) is empty, and the data written in the flash memory 6 itself. It may be a thing to confirm that has become garbled.

The position of the management area E is not necessarily limited to the last area of the block, but may be another place.
-Rebuilding a mirror is not limited to initialization. In short, as long as the management information 17 is read, the execution time is not particularly limited.

  The data written to the flash memory 6 is not limited to the ignition on start time, the ignition on execution time, and the sensor value of the steering angle sensor indicated by ID1 to ID3, and may be data exchanged during operation of the steering angle detection device 3. Anything can be used.

The middleware 14 is not always necessary and may be omitted.
The number of blocks of the flash memory 6 is not limited to a total of 6 from A to F, and may take other numbers.

The first use block for data writing is not necessarily limited to the A block 7, and may be the C block 9 or the E block 11.
The computer 5 of this example having an access protection function is not limited to being mounted on the steering angle detection device 3 and is particularly limited as long as it includes the computer 5 having the flash memory 6 as a memory. It is not a thing.

Next, technical ideas that can be grasped from the above-described embodiment and other examples will be described below together with their effects.
(1) In any one of claims 1 to 3, management information for notifying a current state of the block is written in the block, and the protection execution unit checks the management information to perform the rotation. The presence or absence of the abnormality in the block to be used is confirmed. According to this configuration, for example, confirmation of the presence / absence of abnormality in the rotation block in use can be performed by a simple method of viewing the management information.

  (2) In any one of claims 1 to 3 and the technical idea (1), when a data read command is received from the application, data is read by reading data only from the block on the rotation block side. Data reading means for executing reading is provided. According to this configuration, when data is read, data is read only from the rotation block side of the two blocks holding the same data group, so that it is possible to shorten the data reading time.

  6 ... Flash memory, 7 ... A block constituting block (rotation block), 8 ... B block constituting block (mirror block), 9 ... C block constituting block (rotation block), 10 ... block (mirror block) ) Constituting D), 11... E block constituting block (rotation block), 12... F block constituting block (mirror block), 13... Application, 15. 18, a data loss presence / absence confirmation unit as a monitoring unit, 19 a protection execution unit, a data processing continuation unit constituting a data processing continuation unit, 20 a protection execution unit, and a mirror reconstruction unit constituting a mirror reconstruction unit.

Claims (4)

When a flash memory is used as a rewritable nonvolatile memory and a driver writes data to the flash memory in response to a write command from an application, a plurality of data erasing units are constructed while building the memory area of the flash memory. In the flash memory data loss prevention device that reads and writes data while rotating and using the blocks in order,
Mirroring processing means for using one of the two blocks set as a set as a rotation block used in the rotation and using the other as a mirror block of the rotation block;
Monitoring means for monitoring whether there is an abnormality with respect to access of the driver to the flash memory;
A flash memory data loss prevention apparatus, comprising: a protection execution unit that continues reading and writing of the data using the mirror block when the monitoring unit confirms the abnormality. The protection execution means
Data processing continuation means for continuing reading and writing of the data in the mirror block when the abnormality is confirmed by the monitoring means;
2. The flash memory data loss prevention apparatus according to claim 1, further comprising mirror rebuilding means for rebuilding mirroring destroyed by the abnormality.   When the mirror rebuilding unit receives a new data write command from the application during the rebuilding of the mirroring, the mirror rebuilding unit temporarily interrupts the rebuilding and writes the data to the free memory area. 3. The flash memory data loss prevention apparatus according to claim 2, wherein data writing during mirroring reconstruction is possible. When a flash memory is used as a rewritable nonvolatile memory and a driver writes data to the flash memory in response to a write command from an application, a plurality of data erasing units are constructed while building the memory area of the flash memory. In the flash memory data loss prevention method that reads and writes data while rotating and using the blocks in order,
One of the two blocks set as a set is used as a rotation block used in the rotation, and the other is used as a mirror block of the rotation block, so that there is an abnormality in the driver access to the flash memory. A method for preventing data loss in a flash memory, wherein monitoring is performed by monitoring means, and when the abnormality is confirmed by the monitoring means, reading and writing of the data is continued using the mirror block.

Images