JP3518232B2 - Portable media library device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of operating a highly available storage device system for a portable storage medium, and more particularly to a process for discharging a defective portable storage medium.

[0002]

[Prior art]

(1) Regarding the disk array, as a known example closest to the storage device system to which the present invention is applied, a paper by Patterson (D. Patterson, et al, “A.
Case for Redundant Arrays of Inexpensive Disks (RAI
D) ”, ACM SIGMOD Conference Proceeding, Chicago, I
L, June 1-3, 1988, pp.109-116). This paper discloses a technique for arranging data on a disk array.

The disk array is a mechanism for realizing high performance and high reliability of the disk system. In a disk array, multiple disk devices are physically made to appear as one disk device to a processing device for high performance, while a disk device storing data fails for high reliability. If the error occurs, redundant data for data recovery is stored in another disk device.

Here, a data recovery method when parity is used as redundant data described in Patterson's paper will be described. When data is written, it is divided into n (n ≧ 1) into bits, bytes, or the size of a block which is a unit of writing from the processing device, and the plurality of divided data are written into different disk drives. Then, the parity is created by calculating the exclusive OR of the divided plurality of data, and the parity is written to a disk drive different from the data. For example, when the data is divided into five, if each data is D1 to D5 and the parity is P, "P = D1 xor D2 xor D3 x
or D4 xor D5 (xor is an exclusive OR operator) ”holds. This set of D1 to D5 and P is called a parity group, and when the values of D1 to D5 are updated, P is also updated so as to maintain the value according to the above formula. If any one disk fails,
The data can be restored by calculating the exclusive OR of the data on the remaining disks, and the restored data can be written to the spare disk and the system can be restored by replacing the failed disk with the spare disk. .

(2) Portable media library device By the way, in computer systems, there are magnetic tape and optical storage devices as storage devices often used in addition to disk devices.
k) is receiving attention. The characteristics of these storage devices are that a storage medium and a read / write device (hereinafter referred to as a data transfer device) are separated, and the storage medium is loaded into an arbitrary data transfer device and data on the storage medium is loaded. The point is to read and write. Generally, these storage media are also called portable media. In a large-scale computer system, a library is introduced to easily manage a large number of portable media. The library includes, in addition to the storage medium / data transfer device, a storage for storing a large number of storage media, a robot (hereinafter, referred to as a transport means) for transferring the storage medium between the storage and the data transfer device, and the like. . Since there is a very high need for improving the availability of such a library device, it is effective to realize high availability by applying the concept proposed in Patterson's paper.

As a technical example to which such a concept is applied,
DVD application (Alan E. Bell (IBM Research Divi
sion): “DVD Applications”, COMDEX 96, Nov. 20,
1996). In this paper, RAIL (Redundant Arrays of Inexpensive) that has redundancy by combining multiple ordinary libraries consisting of DVDs, data transfer devices, transporting means, etc.
Libraries) are proposed. Hereinafter, a library device configured by a portable medium will be simply referred to as a library device.

[0007]

Portable media generally have no apparent difference in whether the media is defective or not.
That is, the user cannot visually identify whether or not the medium is defective. When a medium is ejected from the library device, if it is not possible to determine whether or not the medium has a failure, there is a possibility that the failed medium will be mistakenly used again (problem 1). For example, if the faulty medium is used as a part of the medium constituting the library as usual, the entire library is eventually unusable. Further, when the failed medium is to be reused as a spare medium, the processing performed using the medium needs to be redone when it is determined to be the failed medium.

Further, it is generally judged that the storage medium has failed, not when all the blocks in the medium have errors in reading, but in the block in which the error has occurred in the medium. It is the time when the number of exceeds a certain value. Therefore, a block that can be read by the library device without error remains in the failed medium. Here, “readable without error” means that an error does not occur when the library device accesses the block, and that the error of data is caused by error correction information (hereinafter referred to as ECC information) inside the block at the time of reading. It means that it was confirmed that there is no. In this way, if a block that can be read without error remains in the faulty medium, even if the physical identification of Problem 1 is possible, if the medium is thrown into the library device, the If the transfer device receives a request from the host device for a block that can be read without error, it cannot be determined that the medium is a defective medium. In other words, the medium is regarded as a medium that can be used until there is a request for a block that cannot be read, and is operated (Problem 2).

Due to the above two problems, it has been found that the operation of the library device is seriously hindered when the faulty medium is put into the library device. Therefore, in the present invention, the library device is smoothly operated by solving both of the above two problems.

[0010]

In order to solve the above-mentioned problems, in the present invention, the defective medium is subjected to two processes in the library device and then discharged.

(1) Means for Problem 1 (Enables Physical Identification of User) For Problem 1, the user can visually identify the medium as a defective medium (hereinafter referred to as physical identification). ), The library device is provided with a failed medium physical identification processing device for processing the failed medium. The failure medium physical identification processing device is
Before the faulty medium is ejected from the library device, a fault identifier is added to the faulty medium so that the user can clearly identify the medium as the faulty medium. Then, the failure medium to which the failure identifier is added is ejected from the library device.

(2) Means for Problem 2 (Enables Logical Identification of Library Device) With respect to problem 2, even if the faulty medium is loaded in the data transfer device, the library device becomes the faulty medium. A failure medium logical identification processing means for processing the data of the failed medium by the library device is provided so that the defective medium can be immediately identified upon access (hereinafter referred to as logical identification). The failed medium logical identification processing means makes it possible to identify that the medium is a failed medium with respect to an access when recognizing the medium.

[0013]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention based on the solving means will be described below with reference to FIGS.

(1) Configuration of Library Device of Embodiment In FIG. 1, 100 is a host device, 200 is a library control unit, 300 is a library group, and 400 is a storage medium carrying unit. The library control unit 200 includes a main control unit 201 that controls the array control unit 200 and a storage medium 8.
Storage medium failure detection means 202 for detecting a failure of 00,
Fault medium logical identification processing means 203 for logically identifying the fault medium 900 and host input / output means 2
04, a storage medium transport control unit 210 for controlling the storage medium transport unit, and data transfer device input / output units 221 to 22.
5, the library control unit internal bus 20
Connect at 5. The host input / output unit 204 is connected to the host device 100 via the host interface 500.

It is assumed that the library group 300 is composed of five library groups 301 to 305 in this embodiment. Further, it is assumed that the storage medium 800 that constitutes the library is stored in the library groups 1 to 4 and the data is divided using the concept of the disk array in which the redundant data is the parity. The library group 5 holds a spare medium and is used for a recovery process of data of a failed medium. Library group 1
The internal configuration of each of to 5 will be described using the library group 1 as an example. The library group 1 includes a data transfer device 311 that reads / writes data in the storage medium 800 and transfers the data, and a storage medium storage unit 321 that stores the storage medium 800. Library group
Reference numeral 1 denotes a data transfer device input / output unit 221 of the library control unit 200 via the data transfer device interface 701.
Connect with.

The storage medium transport unit 400 includes a storage medium 800.
Storage medium loading / discharging means 420 for loading / unloading the storage medium from outside and storage medium transporting means 410 for transporting the storage medium between the storage medium loading / unloading means 420 and the library apparatus group 300. To be done. Further, the storage medium loading / ejecting means 420 adds a failure identifier 901 that can be identified by the user to the failure medium 900 received from the storage medium transporting means 410, and a storage medium from the outside. Storage medium loading / unloading device 4 for loading 800 / ejecting storage medium 800 to the outside / ejecting failure medium 900 with failure identifier 901 to the outside
22. Storage medium transport unit 40 based on the above configuration
0 is the storage medium transport control means 2 of the library control unit 200 via the storage medium transport control interface 600.
Connect with 10.

The above is the configuration of the library apparatus according to the embodiment of the present invention.

(2) Failure Medium Recovery Process Next, the operation of the failure medium recovery process including the solving means in the embodiment of the present invention will be described with reference to FIGS. 1 and 2. Here, for the purpose of explaining the recovery processing, it is assumed that the storage medium in which the failure has occurred is in the data transfer device 311 of the library group 1.

(2-1) Flow of Failure Medium Recovery Processing In FIG. 2, the storage medium failure detection means 202 operates the storage medium 80 of the data transfer device 311 in the same manner as in the conventional example.
When it is determined that 0 is the failure medium 900, the recovery process is started (11). First, the head block of the failed medium 900 is accessed (12). As in the conventional example, it is checked whether the block can be read without error (13). If the data can be read, the main control unit 201 reads the block data of the defective medium (16) and writes the data in the same block of the spare medium loaded in the library group 5 (17). If the data cannot be read, the main controller 20
1 reads the data of the same block of all other storage media loaded in the library groups 2 to 4 (1
4) Then, block data of the defective medium is created by performing an exclusive OR operation (15), and the data is written to the same block of the spare medium loaded in the library group 5 (17).

After the processing (17) is completed, the main controller 20
1 uses the failed medium logical identification processing means 203 to enable logical identification of the data of the block judged (13) in the failed medium 900 (18). This "fault medium logically identifiable process" refers to a process that puts the block into a state in which reading and writing cannot be performed normally. A specific example will be described later.

After the processing (18) is completed, the main controller 20
1 checks whether the block was the last block of the failed medium 900 (19). If the following block exists, the next block of the failed medium 900 is accessed (2
0), return to the determination (13) and repeat the processing. If it is the final block, the main control unit 201 blocks the failed medium from the library device (21).

After the processing (21) is completed, the main controller 20
1 carries the failed medium 900 in the data transfer device 311 of the library group 1 to the failed medium physical identification processing device 421 using the storage medium carrying means 410 so that the failed medium can be physically identified (22). This “faulty medium physical identifiable process” is performed by the faulty medium identifier 90 for the faulty medium 900.
It means a process that allows the user to visually recognize that the medium is a faulty medium by adding 1. A specific example will be described later.

After the processing (22) is completed, the main controller 20
The reference numeral 1 designates the fault medium 900 with the fault identifier 901 from the fault medium physical identification processing device 421 to the storage medium loading / unloading device 42.
The defective medium 900 with the failure identifier 901 is discharged from the storage medium loading / unloading device 422 (23), and the recovery process ends (24).

(2-2) Specific Example of Fault Medium Logical Identification Process A specific example of the fault medium logical identification process (18) is shown below.

The first method is a method of writing data having incorrect ECC information to all blocks of the defective medium.

In general, in the storage medium, the ECC information is a unit of a bit string of data to which a redundant bit string of a certain length is added so that the division result of a certain generator polynomial Q becomes 0 (hereinafter referred to as ECC unit). Called)). When the data is read, it is confirmed that the division result with the generator polynomial is 0 in ECC units, and if it is not 0, it is regarded as an error.

Therefore, in the first method, first, the failure medium logical identification processing unit 203 holds a block which constitutes an ECC unit, in which the data bit string is all 0s and the redundant bit string is a bit string in which the result of the generating polynomial is not 0. . In the process (18), the data stored in the failed medium logical identification processing unit 203 is written from the main control unit 201. Processing (18)
Is repeated up to the last block, the failure medium 9
Since an error occurs when reading all 00 data, the library device cannot recognize the medium again,
It can be immediately determined that the medium is a failure medium.

As a second method, a simplified version of the first method is used. When the storage medium is recognized in the library apparatus, only the block holding the management information of the storage medium to be accessed first is provided with incorrect ECC information. There is also a method to rewrite the data you have. In this case, the logic identification process can be completed earlier than destroying all the blocks.

(2-3) Concrete Example of Faulty Medium Physical Identification Processing A concrete example of the faulty medium physical identification processing (22) is shown below.

In the first method, the faulty medium physical identification processing device 421 creates a seal in which information on the medium and the medium is clearly a faulty medium, and as a process (22), this is used as a fault identifier. 901 is a method of pasting on the faulty medium 900. This method is applicable to any portable medium.

The second method is the failure medium physical identification device 42.
1 is a method for realizing the processing (22) of adding the failure identifier 901 by directly printing the same content as described above on the failure medium 900. This method can be applied to a directly printable portable medium such as a DVD.

The third method is the failure medium physical identification device 42.
1 is a method of making the failure medium 900 unusable as a storage medium by breaking or bending it. This method can be applied to a portable medium that cannot be used as a storage medium by breaking / bending the medium.

With the above configuration and operation, according to the present invention, the failed medium is ejected after the failed medium can be identified logically and physically during the failed medium recovery process. You can prevent it from being used. Even if it is used by mistake, the library device can be immediately determined to be a failed medium. Therefore, the smooth operation of the library device becomes possible.

Further, in the above embodiment, the library device to which the concept of the data arrangement in which the redundant data is created by the parity of the disk array is applied is described, but the present invention performs the appropriate redundant data creating means and the recovery process. For example,
It is not restricted to the above embodiment.

[0035]

The physically identifiable process of the present invention allows the user to visually identify that the medium is defective. As a result, there is an effect of preventing the user from using the failed medium again. Further, by the logically identifiable processing, even if the failed medium is mistakenly input, the library device can immediately identify the failed medium. As a result, it is possible to prevent the possibility of multiple failures caused by the operation of the library apparatus using the faulty medium after the faulty medium is thrown in.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a library device according to an embodiment.

FIG. 2 is a configuration diagram of a fault medium logical identification means in the embodiment.

[Explanation of symbols]

100: host device, 200: library control unit, 20
1: main control unit, 202: storage medium failure detection means, 2
03: failure medium logical identification processing means, 204: host input / output means, 205: library control unit internal bus, 210:
Storage medium conveyance control means, 221-225 data transfer device input / output means, 300: library group, 301: library group 1, 302: library group 2, 30
3: Library group 3, 304: Library group 4, 305: Library group 5, 311: Data transfer device, 321: Storage medium storage means, 400: Storage medium transport unit, 410: Storage medium transport means, 420: Storage medium input / Ejecting means, 421: physical processor for physical identification of failed medium, 422: storage medium inserting / ejecting device, 500: host interface, 600: storage medium transport control interface, 701: data transfer device interface, 80
0: storage medium, 900: failed medium, 901: failed medium identifier

─────────────────────────────────────────────────── ─── Continuation of the front page (72) Takashi Oeda, Takashi Oeda 1099, Ozenji, Aso-ku, Kawasaki-shi, Kanagawa Hitachi, Ltd. System Development Laboratory (72) Akira Yamamoto, 1099, Ozenji, Aso-ku, Kawasaki-shi, Kanagawa Hitachi, Ltd. (56) Reference JP-A-7-201162 (JP, A) JP-A-7-244563 (JP, A) JP-A-7-319634 (JP, A) JP-A-62-235551 ( JP, A) JP 5-24316 (JP, A) JP 63-306635 (JP, A) JP 1-183832 (JP, A) JP 55-157043 (JP, A) (58 ) Fields surveyed (Int.Cl. ⁷ , DB name) G06F 3/06-3/08 G11B 20/10-20/18

Claims (6)

(57) [Claims] 1. A library control unit for controlling access to a plurality of library groups composed of a plurality of storage media, and a storage medium transport unit for transporting the plurality of storage media, wherein the library control unit comprises: A storage medium failure detection means for detecting a failure of the plurality of storage media, and a main control unit for controlling data read / write control for the plurality of library groups and conveyance control of the plurality of storage media; For the storage medium determined to be defective by the detection means, the main control unit writes data that causes a data error in a predetermined location of a storage area of the storage medium determined to be defective, and the storage medium is conveyed. Control unit to eject the storage medium judged to be defective, in which the data causing the data error is written. Characteristic portable media library device. 2. The portable medium library device according to claim 1, wherein the predetermined location is all blocks included in the storage medium determined to be in failure. 3. The portable medium according to claim 1, wherein the predetermined location is a block in which information for managing the storage medium is stored in the storage medium determined to be the failure. Library device. 4. have a failure medium physical identification processor for adding said storage medium conveying unit further fault identifier is the storage medium determining that the failure, the storage medium conveying unit, the malfunction determination has been stored The portable medium according to claim 1, 2 or 3, wherein when the medium is discharged to the outside, the fault identifier is added to the storage medium that has the fault medium identification processing unit and is determined to have the fault. Library device. Wherein said fault identifier is a portable medium library apparatus according to claim 4, wherein the malfunction determination is a storage medium is a seal information is described to indicate that it is faulty. Wherein said fault identifier is the failure judged to be a storage medium indicating that a fault, according to claim 4, wherein it is a character printed on the storage medium determining that the fault Portable media library device.