JPH09330263A - Data archive system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make it possible to take out and carry data without any data copying and transfer processing time while maintaining cost advantages by distributing data to a corresponding storage medium according to the attribute of the data. SOLUTION: Client computers 300 are connected to a data archive server 100 by a network 400. To the data archive server 100, an automatic storage device library 200 is connected which has a recording medium insertion and ejection slot 210. Software of the data archive server 100 consists principally of data distribution control based upon the attribute of archived data, state management about whether the medium is dismounted, archive management for access control over the medium, and library control which controls an automatic storage device library 200 according to the indication of archive management software.

Description

Detailed Description of the Invention

[0001]

TECHNICAL FIELD The present invention relates to a data archive system.

[0002]

2. Description of the Related Art Hitherto, as a data archiving method of a computer system, there are a method in which an archiving device is provided for each computer and the method is performed individually, and a method in which a data archive server is used and intensively performed.

The former method is a method in which a storage device such as a magnetic tape device is connected to each computer and data on the computer is stored in the connected magnetic tape device. This method has the advantage that when carrying the stored data, it is only necessary to bring the magnetic tape storing the necessary data, so that the processing time for carrying the data is short, but the data for archiving must be stored for each computer. There is a disadvantage that the cost is high because a storage device needs to be installed. Further, since the management of the storage medium storing the data is not automated, there is also a problem that the data is easily lost, and it takes time to retrieve and retrieve necessary data.

The latter method of processing data intensively by a data archive server is a method of connecting a plurality of (client) computers to a data archive server via a network and storing data intensively in a storage device managed by the data archive server. It is. This method can reduce the cost required for data archiving per client. Therefore, it is possible to use an automatic storage device library as shown in Japanese Patent Publication No. 7-27442. By using such an automatic storage device library, it is possible to automate the management of the storage medium and solve the problems of the former method such as loss of data and time required for data retrieval / search. In other words, the client computer can hold online the data that was too large in capacity and could only be held offline.

[0005]

The data archive server using the automatic storage device library has such advantages, but on the other hand, when trying to use the stored data, it is necessary to copy the data via the network. Generally, data transfer via a network is slow, and it takes time to transfer a large amount of data. For example, using the currently popular Ethernet (10 Mbits / sec),
It takes at least 800 seconds or more to transfer data to a magnetic disk device (1 Gbyte) generally used in personal computers. Actually, it takes several times as much time as the efficiency of the network and the processing time other than the network are required. In particular, when data is used from a remote place, even if the network is connected, it is usually much slower than the local network, and sometimes the network is not connected. Therefore, the data is carried on a removable medium (magnetic tape, optical disk, etc.).

However, if a magnetic tape device or the like is attached to each client for carrying data, the cost merit of the centralized data archive is lost.

Further, in order to take out the medium of the automatic library device of the data archive server and bring it, it is necessary to control the distribution of the data suitable for taking out to each medium in advance.

Further, while the medium is taken out, the data on the medium becomes offline and inaccessible. When the medium storing the data commonly accessed by each client computer is removed, the data is removed. Processing such as reading, updating, and adding will not be possible until the medium is set back.

It is a first object of the present invention to provide a system which can take out archived data and carry it without any data copy or transfer processing time while maintaining the cost advantage of the centralized data archive server.

A second object of the present invention is to provide data distribution control in consideration of removal of a recording medium in a data archive server using an automatic library device.

A third object of the present invention is to provide a system in which a data archive server using an automatic library apparatus can process data on the medium as usual even if the recording medium is removed. It is in.

[0012]

In order to solve the above-mentioned problems, the present invention relates to a data archive server using an automatic library device, which handles data according to a means for removing a designated medium from a library and an attribute of data to be archived. Means for distributing the data to the storage medium, a means for reproducing the data of the removed medium by adding and recording a redundant code when archiving the data, and a method for updating or adding data to the removed medium. A means is provided for temporarily storing the data and reflecting the updated / added data on the recording medium when the medium is set in the automatic library device again.

The data is stored in advance so that it is convenient for removing the medium and taking out the data by means for allocating the data to the corresponding storage medium according to the attribute of the data to be archived. For example, the data may be stored in a different medium for each owner.

With this arrangement, when taking out the data owned by the owner, the means for removing the designated medium from the library designates the medium storing the own data and removes it from the library device. Can be brought out. At this time, data copy processing and data transfer processing via the network are not required.

Further, when data is archived, a redundant code is added and recorded, and the data of the removed medium is reproduced by a means for reproducing the data of the removed medium. Is read, the requested data is calculated and reproduced from the redundant code and the data of the medium which is not removed on the data archive server. This allows each client to read data without being aware that the medium has been removed.

Further, the data archive server is provided with means for temporarily storing data updates and additions to the removed medium and reflecting the updated / additional data on the recording medium when the medium is set again in the automatic library apparatus. By this, each client can update / add data without being aware that the medium has been removed.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is an explanatory view of the present invention. A plurality of client computers 300 are connected to the data archive server 100 via a network 400. A recording medium insertion / removal port 21 is provided in the data archive server 100.
An automatic storage device library 200 having 0 is connected.

FIG. 2 is a block diagram of the automatic storage device library 200. What is the data archive server 100?
Connection is made via a host interface 251 such as CSI (Small Computer System Interface). The library controller 250 interprets a command sent from the data archive server 100 via the host interface 251, and drives the drive device 220 or the medium changer 2.
40 is controlled. The drive device 220 is a drive device of a removable medium type storage device such as a magnetic tape drive device or an optical disk drive device, and is connected to the library controller 250 via a drive interface 252. Drive interface 252 is typically SC
SI. The medium changer 230 automatically supplies the medium to the drive device 220. The medium changer 230 stores the recording medium 241 stored in the medium rack 240 by the library controller 250 via the medium changer interface 253.
Drive device 22 according to the command sent to 30
0 or set in the recording medium insertion / removal port 210.

FIG. 3 is a block diagram of the software of the data archive server 100. Archive management 500 and operating system 501, file system 502, and archive management software instructions that perform data distribution control according to the attributes of the data to be archived, state management such as whether or not a medium is removed, and access control to the medium The library control 503 that controls the automatic storage device library 200 according to the above is a main component.

First, the archive process when the medium is not removed but is in the automatic storage device library 200 will be described with reference to the flowchart of FIG.

A data archive request from a client arrives at the archive management software of the data archive server via the network.

The archive management software searches the data storage medium according to the attribute such as the owner of the requested archive data. The user sets in advance which attribute is used to manage the medium. In this embodiment, it is assumed that the storage medium is simply set to correspond one-to-one with the attribute of the owner.

Next, a storage medium for storing redundant data, which will be described later with reference to FIG. 9, is searched. The redundant data storage medium uniquely corresponds to the data storage medium.

Then, according to the search result, the data archive server issues a command to the automatic storage device library to set the storage medium of the data and the corresponding redundant data in the drive. The library controller interprets the command, and the medium changer sets the medium in the drive device.

The archive data sent from the client via the network is written on the data medium.
Further, the redundant data is also updated so as to maintain the relationship of the formula of FIG.

This redundant data update is used later to regenerate the data when the medium of the redundant data group is removed.

Next, the archive processing when the medium is removed will be described.

Up to the point where the medium is searched, it is the same as when there is no medium. When the archive management software detects that the medium has been removed as a result of the search, only the redundant data storage medium of the medium is set in the drive device, and the update data is temporarily stored in the storage device (main storage of the data archive server or secondary storage). It is recorded on the memory or in the memory device in the automatic memory device library). When the medium is present, the redundant data is upset and updated.

For example, when there is a request from a client to remove the data medium of a particular owner. The archive management software of the data archive server searches the medium corresponding to the owner, and instructs the automatic storage device library to move the medium to the medium removal / insertion slot. Thus, the medium ejected to the removal port by the medium changer can be carried by the user. If you have a drive device of the medium before taking out,
The data can be read from the taken-out medium (the owner's data is archived). Moreover,
No data copying or data transfer over the network is required to carry.

The processing when a client issues a request to read the data of the removed medium to the archive management software of the data archive server will be described with reference to FIG.

The archive management software searches the data storage medium and detects that the medium has been removed.
Next, the remaining storage media of the same redundant data group are searched and a command is sent to the automatic storage device library to set them in the drive.

The data to be read takes the exclusive-OR of all the rest of the same redundant data group D1 = D2
It can be calculated by + D3 + P (+; exclusive OR). The calculation result is sent to the client via the network and the processing ends.

FIG. 8 shows a process of inserting the taken-out medium into the automatic storage device library.

First, the medium is inserted from the medium insertion / removal port of the automatic storage device library.

By detecting that the medium has been inserted, it is checked which data owner's medium has been inserted. The fact that the medium has returned to the automatic storage device library is registered in the medium management table of the archive management software. The archive management software reflects data update / addition on the medium while the medium is being taken out.

FIG. 9 shows data distribution according to attributes, a method of constructing a redundant data group, and a method of calculating redundant data when constructing one redundant data group with four media.

The redundant data on the medium for redundant data is calculated by taking the exclusive OR of all the data on the remaining media of the redundant data group. This method is the same as the method of calculating redundant data in the disk array device.
When this redundant data is used, even if any one of the four media forming the redundant data group is lost, the data can be reproduced by calculation. The data reproducing method is the calculation shown in FIG. Here, the reproducing method when the medium D1 is taken out is shown.

As described above, since only one of the redundant data groups can be reproduced, a medium which is likely to be removed at the same time, for example, four media belonging to the same attribute A in FIG. It may be configured to belong to the data group. Then, even if all four media belonging to attribute A are removed at the same time, all data can be reproduced.

When the second medium is removed from the same redundant data group, the data cannot be reproduced. In this case, 1) Do not allow the second one to be removed, 2)
If the second data is removed, the data archive server terminates the data read request from the client without error without reproducing the data. 3) Use redundant data with a stronger code such as Reed-Solomon code. A method of making it possible to reproduce even if it is wet is considered.

[0041]

According to the present invention, data can be provided while providing a highly reliable archive means which is much cheaper than a system in which each computer on the network is equipped with a data archive device and which has less loss of a storage medium. The data classified by the data attribute can be taken out at high speed and carried without copying or data transfer processing via the network.

Further, it is possible to perform processing such as reading, updating, and adding on the fetched data without the influence of the fetched data.

[Brief description of drawings]

FIG. 1 is an explanatory diagram of a data archive system of the present invention.

FIG. 2 is a block diagram of an automatic storage device library.

FIG. 3 is a block diagram of software of a data archive server.

FIG. 4 is a flowchart of archive processing when there is a medium.

FIG. 5 is a flowchart of archive processing when a medium is removed.

FIG. 6 is a flowchart of medium removal processing.

FIG. 7 is a flowchart of a data reading process when the medium is removed.

FIG. 8 is a flowchart of medium insertion processing.

FIG. 9 is an explanatory diagram of data distribution by attributes and redundancy groups.

[Explanation of symbols]

100 ... Data archive server, 200 ... Automatic storage device library, 210 ... Recording medium insertion / removal port.

Claims (2)

[Claims] 1. A data archive system in which a data archive server having a plurality of client computers and an automatic storage device library using a removable medium is connected via a network, and means for removing a designated storage medium from the library device and a file attribute. And a means for storing the data in the corresponding storage medium according to the above, and temporarily storing the update and addition of the data to the removed storage medium, and when the recording medium is set again in the automatic storage device library, Although it is a centrally managed data archive server that has means for reflecting updated and additional data on the recording medium, it retrieves archived data without the need for data copy processing,
A data archive system, which can be carried and can be updated and added while the data is being taken out. 2. A data archive system according to claim 1, wherein the data is read from the removed storage medium by means of reproducing the data of the removed medium by adding a redundant code.