JPH08263350A - Information management system and method - Google Patents

Abstract

PURPOSE: To provide an information management system which does not need any device of large processing capability by shortening the response time for retrieval and updating of the managed information and also attaining the distributed processing of information. CONSTITUTION: Plural information processors 31 perform the distributed management of information on an entire system. Then each of these processors 31 stores its own managing information in a semiconductor memory 59 and retrieves and updates the information through the memory 59. The information stored in the memory 59 is periorically copied to a magnetic disk 41. An updating history of information is acquired by a magnetic disk 39 every time the information of the memory 59 is updated. If the processor 31 has a fault, the information of the memory 59 is corrected based on the backup information stored in the disk 41 and the updating history stored in the disk 39 after the fault of the processor 31 is eliminated.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an improvement of an information management system suitable for inquiries such as futures trading information, stock price information and horse racing result information.

[0002]

2. Description of the Related Art Conventionally, as this kind of information retrieval system, for example, a system called a host centralized system having a host centralized device (a host computer; hereinafter the same) as shown in FIG. Are known.

In the system described above, a wide area network (WAN: abbreviation for wide area network) 7 is provided from M search terminals 3.
Based on the information search request transmitted via the network or the update information transmitted from the N update terminals 5 via the network 7, the host centralized apparatus 1 centrally executes the corresponding processing. Composed. The host centralized apparatus 1 has a magnetic disk 9 for storing search information and a magnetic disk 1 for storing updated information in order to intensively execute the above processing.
1, a user information backup tape 13 for backing up information in the magnetic disk 9, a magnetic disk 15 for storing new search information, and an update terminal 17 similar to the update terminal 5 are directly connected. The host concentrator 1
A file information access processing unit 19, an editing program 21, and an updating program 2 having various functions as described in detail later.
3, and a file restoration processing unit 25.

Next, the processing operation at the time of information retrieval in the above system will be described. First, when an information retrieval request is transmitted from the retrieval terminal 3 to the host centralized apparatus 1, the editing program 21 is activated, and the retrieval information storage magnetic disk 9 (for example, a database is retained via the access processing unit 19). Read the information in. Then, the read information is edited and the edited information is transmitted to the search terminal 3 that has transmitted the information search request. The transmitted information is displayed on the display unit of the search terminal 3.

Next, the processing operation at the time of updating information in the above system will be described. First, when the update information is transmitted from the update terminal 5 (or 17) to the host centralized apparatus 1, the update program 23 is activated and the update program 23
Requests the access processing unit 19 to update the information in the magnetic disk 9. In response to this update request, the access processing unit 19 updates the information in the magnetic disk 9 and, in preparation for recovery when a failure occurs in the magnetic disk 9, updates the contents of the update to the magnetic disk 11 (journal). ) Inside. Then, the update content is notified to the update program 23 as the processing result. The above-mentioned update contents are the update terminal 5 (or 1
7) will be notified. Further, all information on the magnetic disk 9 is periodically copied to the backup tape 17.

Further, the processing operation when a failure occurs in the magnetic disk 9 in the above system will be described. When a failure occurs in the magnetic disk 9, the file recovery processing unit 25 is activated after the failure is removed, and the contents of the backup tape 13 are copied to the magnetic disk 15. After that, the update information of the magnetic disk 11 is overwritten on the magnetic disk 15 via the access processing unit 19.

[0007]

In the above system, the information (for example, database) to be searched and updated is stored in the magnetic disk 9. Therefore, there is a problem in that it takes a long time to respond because the magnetic disk device must be accessed in both cases of information retrieval and updating.

Further, in the above-mentioned conventional system, there is a problem that the host centralized unit 1 having a large processing capacity is required because the host centralized unit 1 performs all the processes of information retrieval, update and failure recovery.

Therefore, it is an object of the present invention to use a high response information storage device such as a semiconductor memory,
An object of the present invention is to provide an information management system that does not require a host centralized device having a large processing capacity by shortening the response time at the time of access such as information retrieval and updating and enabling distributed processing.

[0010]

An information management system according to the present invention comprises a plurality of information processing devices for managing the resources of the entire system in a distributed manner. Here, "resource" means
In this specification, the database to be searched and updated,
It means information such as files and tables or a collection of information.

Each information processing device holds a local resource managed by the semiconductor memory. Then, an access operation for searching or updating is performed with respect to the local resource on the semiconductor memory.

Further, each information processing apparatus has a first non-volatile storage medium for intermittently acquiring and retaining backup information of the semiconductor memory and a local non-volatile storage medium for ensuring the reliability of local resources on the semiconductor memory. A second non-volatile storage medium that acquires and retains a history of updates made to the resources, and information in the first and second non-volatile storage media is used for the local resources in the semiconductor memory. And a memory recovery means for performing a failure recovery process.

The present invention also provides an information management method executed by the above method.

[0014]

According to the method of the present invention, the resources in the system are distributed and managed by a plurality of information processing devices. Each information processing device holds each local resource in the semiconductor memory and performs an access operation such as a search or an update with respect to this semiconductor memory. Therefore, access response time is short and high throughput can be obtained.

Each information processing apparatus intermittently acquires a backup of local resources on the semiconductor memory in the first nonvolatile storage medium (for example, a magnetic disk). Also, every time the local resource is updated, the update history is updated to the second
To a non-volatile storage medium. Then, when a failure occurs, after the failure is removed, the backup information and the update history on the nonvolatile storage medium are used to restore the local resource on the semiconductor memory. Therefore, the reliability of local resources on the semiconductor memory is ensured.

Further, since resources are distributed and managed by a plurality of information processing devices, the influence of a failure occurring in a certain information processing device is limited to the local range of the information processing device, and the reliability of the system is improved. To do. Further, when the traffic increases, if the information processing device is added accordingly, it can be dealt with without making a special change of the application, and the load on each information processing device does not increase.

In the preferred embodiment, an update information transmission device is connected to a plurality of information processing devices via a network. This update information transmission device issues update information for each local resource to each information processing device. Each information processing device updates each local resource based on the update information. However, when the update fails due to a failure or the like, the update information transmitting apparatus acquires and holds the update information related to the failed update in the non-volatile storage medium in the own apparatus. Then, in each information processing device, in the recovery process after the failure removal, not only the backup information and the update history in the own device but also the unexecuted update information in the update information transmitting device are used to locally The physical resources. This allows
The reliability of decentralized managed resources is guaranteed reliably.

[0018]

Embodiments of the present invention will now be described in detail with reference to the drawings.

FIG. 2 is a block diagram showing the structure of a portion that functions during normal operation in one embodiment of the information management system of the present invention.

In this system, a plurality of information processing devices 3
1 is a network (WAN or LAN: local area net)
It is connected to the update information transmission device 33 via a work abbreviation 35.

Each information processing device 31 has an initial table file 37 storing initial values of information (hereinafter referred to as resources) to be searched and updated, such as a database.
After-update information storage magnetic disk 39 for storing resource update history information (information showing the updated image of each updated record), two-side rescue basic information storage magnetic disk 41, and The search terminal 43 is directly connected. In addition, each information processing device 31 includes an access processing unit 51 having various functions as described in detail later, an update program 53, an editing program 55, a basic relief information acquisition processing unit 57, and a semiconductor memory table 59.

The plurality of information processing devices 31 manage the overall resources of the system in a distributed manner. In other words, each information processing device 31 holds local resources managed by the information processing device 31 on the semiconductor memory table 59 during system operation, and performs access operations such as search and update on the semiconductor memory table 59.

The update information transmitting device 33 includes an update terminal 45.
And an untransmitted information storage magnetic disk 47 are connected. The update information transmission device 33 includes an update information transmission program 61 and a file information access processing unit 63 having various functions as will be described in detail later. The update information transmission device 33 functions to order each information processing device 31 to update each local resource.

Next, the processing operation of resource search and update in the above system will be described with reference to FIG. In the following, a search and update process for a local resource in one information processing device 31 shown in the uppermost part of FIG. 2 will be described as an example.

First, the processing operation at the time of retrieval will be described.

At the start of operation of the system, the information processing device 3
In 1, the resources in the initial table file 37 are copied to the semiconductor memory table 59 by the access processing unit 51.

When an information search request is issued from the search terminal 43 during the operation period, the edit program 5 is sent from the search terminal 43.
Based on the search condition input in 5, the following sequence for information search is executed.

That is, the editing program 55 is activated, and the information corresponding to the above search conditions is read from the semiconductor memory table 59 via the access processing section 51. Then, the read information is edited and the edited information is transmitted to the search terminal 43 that has issued the information search request. The transmitted information is displayed on the display unit of the search terminal 43. As a result, a series of processing operations at the time of information retrieval is completed.

Next, the processing operation at the time of updating information will be described.

First, when the update information is issued from the update terminal 45, the update information transmission program 61 of the update information transmission device 33 is started by this, and the above-mentioned update is sent from the program 61 to the information processing device 31 via the network 35. Information is sent.

When the above-mentioned update information is received by the information processing device 31, the update program 53 is activated thereby. Then, the program 53 updates the contents of the semiconductor memory table 59 via the access processing unit 51, and stores the updated information image (update history information) in the magnetic disk 39.

When the updating process is completed, the updating program 53 notifies the updating information transmitting device 33 via the network 35 that the contents of the semiconductor memory table 59 have been correctly updated. As a result, a series of processing operations at the time of updating information is completed.

Here, the contents of the semiconductor memory table 59 are not correctly updated despite the fact that the update information transmission program 61 has sent the above-mentioned update information to the information processing device 31 due to some abnormality in the system. Then, the following processing operation is performed.

That is, when the update information is not reflected in the semiconductor memory table 59 due to an abnormality in the network 35 or the information processing device 31, etc., the program 61 of the update information transmitting device 33 causes the information processing device 33 to process the information. Since the device 31 does not receive the notification that the update has been successful, the abnormality is recognized, and the name and update information of the information processing device 31 in which the abnormality has occurred (image format after update)
Is stored in the untransmitted information storage magnetic disk 47 via the file access processing unit 63.

In the present system, the basic rescue information acquisition processing section 57 is activated at a fixed time interval (or at any time depending on the processing status) asynchronously with the information search and information update in the semiconductor memory table 59. Magnetic disk 41 for two surfaces
Of these, in principle, the resources in the semiconductor memory table 59 are copied to the magnetic disk 41 on the older (earlier in time) side, that is, the backup information of the resources is acquired.

In this backup acquisition, if the previous backup information is completely stored in both magnetic disks 41 on the two sides, the magnetic disk 41 on the older side is selected as the backup acquisition destination. Note that this selection is performed by the processing unit 57 by the magnetic disk 41 on each surface.
It can be determined by referring to the backup acquisition time (time stamp) recorded in.

On the other hand, while the previous backup is being acquired, the previous backup information is incomplete on the magnetic disk 41 on any one side due to the occurrence of some abnormality in the system. If it is stored only in this case, the backup information of this time is acquired in the magnetic disk 41 on the side of the incomplete storage state. Whether the previous backup information is complete or incomplete can be determined by checking the trailer information of the magnetic disk 41 (for example, checking whether the time stamp times stamped at the head and tail of the magnetic disk 41 are the same). .

In this way, the basic rescue information acquisition processing section 57 leaves the contents of the magnetic disk 41 on the side where the backup information is completely stored until the end and the backup acquisition time is newer, New backup information is acquired on the magnetic disk 41 on the other surface.

FIG. 3 is a block diagram showing the structure of a portion that functions in the recovery processing in the event of an abnormality in this embodiment.

The information processing apparatus 31 has a memory information restoration processing section 65 and an update information collection program 67 which have various functions as will be described in detail later. The update information transmission device 33 has an untransmitted information transmission application program 69 having a function to be described in detail later.

Next, a process of restoring the semiconductor memory table 59 when an abnormality occurs in the information processing device 31 will be described with reference to FIG.

This restoration process is performed until just before the system becomes abnormal (hereinafter referred to as the first restoration process).
And a process of recovering information from the time when the abnormality occurs in the system to the present (hereinafter referred to as a second recovery process).

First, the first restoration process will be described.

In the information processing device 31, when it is recognized that the system has been started in an abnormal state in the initial setting process at startup, the recovery processing unit 65 automatically starts and By performing the processing described, the semiconductor memory table 59 is restored to the state immediately before the information processing device 31 becomes abnormal.

First, the restoration processing section 65 selects one of the two basic magnetic disk 41 for storing the rescue basic information which is used alternately.
The magnetic disk 41 in which the information in the semiconductor memory table 59 is completely stored to the end and the latest information is stored is specified. This specification is possible by checking the trailer information of the magnetic disk 41 on each surface.

After the magnetic disk 41 is specified as described above, the restoration processing section 65 copies the complete and latest backup information stored in the magnetic disk 41 to the semiconductor memory table 59. Then, after this copying, the recovery processing unit 65 reads the above-mentioned update history information from the updated information magnetic disk 39, and updates the information in the semiconductor memory table 59 based on the update history information. This completes the first recovery process.

Next, the second recovery process will be described.

When the above-mentioned first restoration process is completed, the update information collection program 67 is started. Then, the program 67 requests the untransmitted information transmission application program 69 to transmit the untransmitted information via the network 35. Upon receiving this transmission request, the program 69
The untransmitted information (format of updated image) of the corresponding information processing device 31 is transferred from the untransmitted information storage magnetic disk 47 via the file information access processing unit 63 to the program 6
Send to 7. Upon receiving the untransmitted information, the program 67 reflects the update of the untransmitted information in the semiconductor memory table 59 via the access processing unit 51. This completes the second recovery process.

According to the present embodiment as described above, unlike the conventional system, the resources are stored in the semiconductor memory table 59 built in each information processing apparatus 31, and the access for searching or updating is stored therein. Since it is decided to perform the above, it is possible to greatly reduce the access time for search and update, so that high response and high throughput can be expected.

Further, in this embodiment, since the resources of the entire system are managed by a plurality of information processing devices in a distributed manner, even if the traffic increases rapidly in the future, the information processing devices 31 can simply be added accordingly. , Can be easily handled without substantial changes to the application or complicated tuning.

Further, if the information processing devices 31 are added in accordance with the increase in traffic, the processing amount of each information processing device 31 does not change so much, and the load on each information processing device 31 does not increase.

Since a plurality of information processing devices 31 manage resources in a distributed manner, even if there is an information processing device 31 that is down, the range of influence of the information processing device 31 is a local resource of that information processing device 31 only. Therefore, the reliability of the system is greatly improved.

Furthermore, since the restoration process may be performed only for the local resource of the information processing device 31 in which the abnormality has occurred, the restoration of the resource can be performed quickly.

It should be noted that the above-mentioned contents are only related to one embodiment of the present invention, and it does not mean that the present invention is limited to the above-mentioned contents.

For example, in the above embodiment, the update information transmitting device is connected to the information processing device via the network, and the search terminal is directly connected to each information processing device, but it is not always necessary to have such a configuration. Alternatively, for example, the present invention can be applied to a system in which an update information transmission device is provided in each information processing device and a search terminal is connected to a network.

[0056]

As described above, according to the present invention,
By shortening the response time of access such as information search and update, and enabling distributed processing,
It is possible to provide an information management system that does not require a host centralized device having a large processing capacity.

[Brief description of drawings]

FIG. 1 is a block diagram showing a conventional information management system.

FIG. 2 is a block diagram showing a configuration of a part related to processing during normal operation of the information management system according to the embodiment of the present invention.

FIG. 3 is a block diagram showing a configuration of a part related to a failure recovery process of the embodiment.

[Explanation of symbols]

 31 information processing device 33 update information transmission device 35 network 37 initial table file 39 magnetic disk for storing information after update 41 magnetic disk for storing basic information for rescue 43 search terminal 45 update terminal 47 magnetic disk for storing untransmitted information 51 access processing unit 53 Update program 55 Edit program 57 Basic relief information acquisition processing unit 59 Semiconductor memory table 61 Update information transmission program 63 File information access processing unit 65 Memory information restoration processing unit 67 Update information collection program 69 Untransmitted information transmission application program

Claims (3)

[Claims] 1. A semiconductor memory, comprising: a plurality of information processing devices for managing the resources of the entire system in a distributed manner, wherein each information processing device holds a local resource managed by each of the information processing devices; A memory access unit for searching and updating a local resource; a first non-volatile storage medium for intermittently acquiring and retaining backup information of the local resource in the semiconductor memory; A second non-volatile storage medium for acquiring and retaining a history of updates made to the local memory in the semiconductor memory using information in the first and second non-volatile storage media. An information management system, comprising: a memory recovery unit that performs a failure recovery process. 2. The information management system according to claim 1, wherein the update information transmission is communicable with the plurality of information processing devices and issues update information for each local resource to each information processing device. Further comprising a device, wherein the update information transmitting device has a third non-volatile storage medium for acquiring and retaining update information related to the failed update when the update in the individual information processing device fails, The memory restoration means of the update information processing device is the first
In addition to the information in the second non-volatile storage medium, the update information in the third non-volatile storage medium of the update information transmitting device is also used to perform a failure recovery process on the local resource in the semiconductor memory. An information management system characterized by performing. 3. A method for distributed management of resources by a plurality of information processing devices, the process of holding a local resource managed by each information processing device in a semiconductor memory of each information processing device, A step of retrieving and updating the local resource; a step of intermittently acquiring and retaining backup information of the local resource in the semiconductor memory in the first non-volatile storage medium; Second history of updates made
Of the non-volatile storage medium and holding the same, and a step of performing a failure recovery process on a local resource in the semiconductor memory using the information in the first and second non-volatile storage media. An information management method characterized by the following.