JP2779540B2 - Magnetic tape unit - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[Industrial applications]

The present invention relates to a magnetic tape device used as an external storage device of an information processing device,

[Prior art]

As one of such magnetic tape devices, there is a magnetic tape device in which data is divided into blocks, one frame of data is composed of a predetermined number of blocks, and the data is sequentially stored on a magnetic tape. In this magnetic tape device, when reading data from a magnetic tape by reproducing data with a magnetic head,
When there is a block in which an error has been detected, the magnetic tape is returned, the data of the same block is reproduced again, and an error retry for re-reading is performed. By the way, it is effective to add an error correction code to the data of one frame and apply an error correction method without performing an error retry when an error is detected during reading. However, the error correction code has a limit of correction capability, and when an error that cannot be corrected any more occurs in one frame of data, the magnetic tape is returned and a plurality of blocks of one frame are returned. It is necessary to perform the reproduction process again, that is, to retry the error. By the way, when performing an error retry,
All data read before the error retry is discarded, error detection is performed again on new data at the time of the error retry, and a new data is read into the buffer memory.

[Problems to be solved by the invention]

However, dust may be attached to the magnetic tape device as a cause of the error, but the position of the dust on the magnetic tape is often not always the same. For this reason, when an error is retried, there is no guarantee that the data of the block that was previously read as having no error can be read again as having no error. Therefore, in the method of discarding the data of the frame read before the error retry and performing the error retry, it is necessary to perform a large number of error retries until the error disappears or the number of error blocks decreases to the extent that the error can be corrected. , Time inefficient. Also, since all data is discarded, it is necessary to write all data to the buffer memory again at the time of error retry, which takes time.

[Means for Solving the Problems]

According to the present invention, data composed of one frame composed of a predetermined number of blocks is reproduced from a magnetic tape and stored in a memory while being divided for each frame. In a magnetic tape device having an error retry function for replaying data of the same frame on the tape, an error detecting means for detecting whether or not each block has an error is provided. A means is provided for storing, in the memory, only a block detected as having an error before the error retry, and only a block detected as having no error at the time of the error retry.

[Action]

At the time of error retry, attention is paid only to a block which was previously erroneous and could not be read, and if there is no error in this block this time, the block is taken into the memory.

【Example】

An embodiment of a magnetic tape device according to the present invention will be described below with reference to the drawings. In the case of this example, as shown in FIG. 2, one block includes, after data for one block, a block identification information ID composed of, for example, a block type indicating a block type, a track number, a block number, and the like. A code, for example, a CRC code is added. Then, one frame is formed by the 16 blocks. In this case, blocks 0 to 13 are data blocks, but blocks 14 and 15 are ECC blocks of error correction codes. As this error correction code,
For example, a redundancy-2 Reed-Solomon code can be used. By using the error correction code of this example, it is possible to correct errors of up to two blocks in one frame. In this example, data is recorded as serial data on a magnetic tape having a width of 1/4 inch, for example, and there is no gap between blocks. FIG. 1 is a block diagram of an embodiment of a data reading section of a magnetic tape device according to the present invention. Numeral 1 denotes a magnetic head. Data reproduced from a tape by the magnetic head 1 is supplied to an error detection circuit 2 where a CRC is provided.
The code performs error detection for each block.
From the error detection circuit 2, a flag FL indicating the result of the error detection is supplied to a control circuit 10 composed of a microprocessor, and data is supplied to a block buffer memory 3. The block buffer memory 3 has a memory capacity of, for example, three blocks, and performs a buffering operation of sequentially reproduced block data. And the control circuit
In 10, the block number detecting means 12 is used based on the identification information ID of the block stored in the block buffer memory 3.
In, a block number is detected. Further, the control circuit 10 is provided with an error flag memory 11, and the error flag from the error detection circuit 2 is stored in the error flag memory 11 in association with the block number from the block number detection means 12. . Therefore, at the time of error retry, this error flag memory 11 stores the information of the error flag of one frame at the time of previous reading. In the control circuit 10, the error flag information from the error flag memory 11 and the current error flag information from the error detection circuit 2 are supplied to the frame memory address and control signal generation means 13, and the block number detection means 12 Is supplied to the frame memory address and control signal generating means 13. This frame memory address and control signal generating means 13
Supplies the address information and the write control signal to the frame buffer memory 4 and controls the writing of the data of the block from the block buffer memory 3. The frame buffer memory 4 has a capacity capable of storing data for two frames or more, in this example, three frames. An error correction circuit 5 is connected to the frame buffer memory 4. The frame memory address and control signal generating means 13 of the control circuit 10 also controls the operation of the error correction circuit 5. In this example, if there are three or more error blocks in one frame, the correction becomes impossible. In such a case, an error retry is performed. Therefore, the frame memory address and control signal generating means 13 sends the error retry control section 14
And the error retry control unit 14 controls the tape running system 6. FIG. 3 is a flowchart of the operation of the control circuit 10 when reading data. That is, first, data of one block is read (step 101). Next, referring to the error flag from the error detection circuit 2, it is determined whether or not there is an error in the block (step 102). If there is an error,
The block data is discarded, and the process returns to step 101 to read the next one block. On the other hand, if there is no error, the block number of this block is confirmed by the information from the block number detecting means 12 (step 103), and it is determined whether the block data of the block number is necessary (step 104).
In this case, in this step 104, data reading
If it is the first time for that frame, it is determined that all data without error is necessary. On the other hand, if the data reading is an error retry, the error flag of each block at the time of the previous reading for the frame from the error flag memory 11 is referred to, and for the block determined to have no error at the time of the previous reading, this time Is ignored as unnecessary, and data is determined to be necessary for the block in which the previous error occurred. When it is determined that it is necessary, the "block valid flag" is set, and the data of the block is written to the address corresponding to the block number in the frame buffer memory 4 (step 105). When data is no longer needed, it is discarded. That is, it is not written to the frame buffer memory 4. Then, it is determined whether all the blocks for one frame have been read (step 106). If one frame has not been completed, the process returns to step 101, and the above steps are repeated. If one frame has been completed, it is determined whether or not the block valid flag is 14 or more (step 107). If the result of this determination indicates that the block valid flag is less than 14, error correction is not possible, and an error retry is performed. If the block valid flag is 14 or more, the process proceeds to step 108, where it is determined whether or not error correction is necessary. If the block valid flag is set for data blocks 0 to 13, it is determined that error correction is unnecessary because all the data has been read correctly, and the process returns to step 101 to read the next frame. On the other hand, even if the block valid flag is 14 or more, there is a data block having an error and the frame buffer memory 4
When the data block is not written,
It is determined that error correction is necessary, and the process proceeds to step 109, where the control circuit 10 sends an error correction start command to the error correction circuit 5. Then, the process returns to step 101 to start reading the data of the next frame. The error correction circuit 5 which has received the start command starts the error correction processing. At this time, the frame memory address and the control signal generating means 13 of the control circuit 10 access the frame buffer memory 4 for the error correction processing. . That is, the block data is read from the frame buffer memory 4 in the order according to the error correction code sequence, and the error correction circuit 5 generates the ECC syndrome, thereby restoring the error block. The restored block is written into the frame buffer memory 4 at the address of the corresponding block number. This error correction operation and reading of the next frame are performed simultaneously. The above operation will be further described with reference to FIG. Now, for example, it is assumed that the error detection at the time of first reading of 16 block data for one frame is in a state as shown in FIG. 4A. Here, OK indicates a block without an error, and NG indicates a block with an error. In this case, the data of the blocks 1, 6, 7, 10, and 14 determined as having an error by the error detection are not written to the frame buffer memory 4, but are discarded, and the remaining blocks 0, 2, 3, The data of 4, 5, 8, 9, 11, 12, 13, and 15 are respectively set with a block valid flag, and written into the frame buffer memory 4 at an address corresponding to the block number. At this time, since the number of erroneous blocks is five and error correction cannot be performed, the control circuit 10 controls the tape traveling system 6 to start an error retry operation. At the time of this error retry, it is determined that there is no error in the previous time, and the block having the block valid flag set is ignored even if it is determined that there is no error, and the data is discarded. Then, only for the "NG" block that was determined to have an error the last time, write control to the frame buffer memory 4 is performed based on the presence or absence of an error. For example, if the error detection result at the time of error retry is as shown in FIG. 4B for the block in which the previous error occurred, blocks 1, 6, 10, and 14 in which there is no error this time are stored in the frame buffer memory. 4 is written to the corresponding address, and the block 7 having the error is discarded. As a result, the number of error blocks is one, so that the error correction circuit 5 operates, and the block 7 is reproduced by the error correction processing. When data blocks 0 to 13 having no error are written to the frame buffer memory 4, these data blocks are transferred to, for example, a host device. In the above example, the case where the block number is added to the block data has been described. However, if the data length (the number of data) of one block is unchanged, which block of one frame is identified. Therefore, the present invention can be applied to the reading of data to which a block number is not assigned. In the above example, an error correction code block is added to one frame, and error correction can be performed in units of one frame. However, the present invention can be applied to a case where no error correction code is added. is there. The present invention is also applicable to a magnetic tape device that records data on a magnetic tape by providing a gap for each block and determines the block based on the gap.

【The invention's effect】

As described above, according to the present invention, at the time of error retry, at the time of the immediately preceding reading, the reading process is performed only on the block having an error. However, even if an error occurs at the time of reading processing in a later error retry, the error is not replaced. Therefore, the number of error retries can be reduced. Also, in the error retry, during the immediately preceding reading process, the blocks that did not have an error are ignored, and only the blocks that have an error are transferred and written to the frame buffer memory. Less time.

[Brief description of the drawings]

FIG. 1 is a block diagram of an embodiment of a magnetic tape device according to the present invention, FIG. 2 is a diagram for explaining a data structure, FIG. 3 is a flowchart for explaining an operation of a main part of the present invention, FIG. FIG. 4 is a diagram for explaining the reading process at the time of error retry. 1; magnetic head 2; error detection circuit 4; frame buffer memory 10; control circuit 11; error flag memory 12; block number detection means 13; frame memory address and control signal generation means 14; error retry control unit

Claims (1)

(57) [Claims] 1. A method comprising: reproducing data constituted by a predetermined number of plural blocks into one frame from a magnetic tape and storing the data in a memory divided into frames; In a magnetic tape device having an error retry function to reproduce the data of the same frame on the magnetic tape again, an error detecting means for detecting whether or not each block has an error is provided. A magnetic tape device comprising means for storing, in the memory, only a block detected as having an error before the error retry, and only a block detected as having no error at the time of the error retry.