JP2018169791A - Information management method and information processing system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable data recovery by roll forward when a failure occurs and not disturb system operations even during continuing of the failure, without increasing the cost.SOLUTION: A management server 20 distributes, to a distribution destination system 30, update information including at least information capable of identifying update contents and information indicating source of the update when specific data update occurs. When the distribution destination system 30 updates the specific data in the self-system, it records the update information as a log and sends the update information to the management server 20. When the management server 20 distributes the update information, the distribution destination system 30 records the update information as a log. The management server 20 restores a database 40 by roll forward with the log in the distribution destination system as journal information.SELECTED DRAWING: Figure 6

Description

  The present invention relates to an information management method and an information processing system for performing data recovery by roll forward when a failure occurs.

  There is an integrated ID management server as a system that centrally manages and distributes ID (identification) information for authentication and authorization of business system users (see, for example, Patent Document 1). The delivery destination system is, for example, groupware, an SSO (Single Sign On) system, or a business application. The integrated ID management server basically becomes a master of ID information, and when the ID information is changed, the changed information is distributed to each system.

  However, there are cases where an update that has occurred in the distribution destination system needs to be input to the integrated ID management server. For example, it corresponds to a user password change in Active Directory (registered trademark). In such a system configuration, when a failure occurs in the integrated ID management server and the backup data at a certain point is restored, there is a problem of how to synchronize and update data with each delivery destination system. is there.

JP 2011-76430 A Special table 2009-508191 Japanese Patent No. 2609625

  There is roll forward as a data recovery method. Roll forward is a method of restoring data using a backup file and a journal file when a failure occurs in a database system or the like.

  In order to perform recovery by roll forward in a system including the integrated ID management server, there is a method in which the internal database of the integrated ID management server always keeps logging update information as a journal file. However, with this method, it is not possible to log updates that occur on the delivery destination system from when a failure occurs in the integrated ID management server until recovery.

  Concerning the consistency of data on each system at the time of return, if there is an update to be input to the integrated ID management server in the distribution destination system, the existence check of the integrated ID management server is performed before the input, The mechanism of two-phase commit that inputs only when it is doing can be considered. However, with this mechanism, an update operation cannot be performed on each system of the distribution destination while the failure of the integrated ID management server occurs.

  Patent Documents 2 and 3 describe techniques for reproducing a replica of a system by performing continuous data shadowing. The system can be automatically recovered by activating the replica in the event of a failure. However, this method requires a separate machine for activating the replica. Therefore, there is a problem that costs are increased accordingly.

  An object of the present invention is to perform roll-forward data recovery when a failure occurs without increasing the cost, and to prevent troubles in the operation of the system even if the failure continues.

  An information management method according to the present invention is an information management method executed in an information processing system including a management server that manages specific data using a database and a distribution destination system that receives distribution of the specific data, When an update of specific data occurs, the management server distributes update information including at least information capable of specifying the update contents and information indicating the source of the update to the distribution destination system. When update of specific data occurs in the system, the update information is recorded as a log, and the update information is sent to the management server. When the distribution destination system distributes the update information from the management server, the update information Is recorded as a log, and the management server restores the database by roll-forward using the log in the distribution destination system as journal information And wherein the door

  An information processing system according to the present invention is an information processing system that includes a management server that manages specific data using a database, and a distribution destination system that receives distribution of the specific data. When an update occurs, update information including at least information that can identify the update content and information indicating the source of the update is distributed to the distribution destination system, and the distribution destination system updates specific data in its own system. Includes update information management means for recording the update information in a log file as a log and sending the update information to the management server when the update information is distributed from the management server. The update information is recorded as a log in the log file, and the management server jars the log recorded in the log file in the distribution destination system. As Le information, and performing database recovery with a roll-forward.

  According to the present invention, it is possible to perform data recovery by roll-forward when a failure occurs without increasing the cost, and to prevent the operation of the system from being hindered even if the failure continues.

It is a block diagram which shows the structural example of the system containing a database and an integrated ID management server. It is explanatory drawing which shows an example of the information output to a log file. It is explanatory drawing which shows an example of the recording method of the log file in a receiving agent. It is explanatory drawing for demonstrating the unprocessed of the subsequent update. It is a flowchart which shows the recovery process by roll-forward. It is a block diagram which shows the principal part of the information processing system by this invention.

  Hereinafter, an embodiment of an information processing system according to the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing a configuration example of an information processing system including a database and an integrated ID management server.

  The information processing system shown in FIG. 1 includes a personnel system 100, a management system 110, a business system 120, and a client terminal 131 (for example, a personal computer). In addition, although one business system 120 is shown in FIG. 1, a plurality of business systems may be included. 1 shows one client terminal 131, but there are a plurality of client terminals in the system.

  At least the management system 110 and the business system 120 are communicably connected via a communication network (for example, an intranet such as an in-house LAN (Local Area Network)).

  The personnel system 100 includes a personnel database 101 in which employee information is stored.

  The management system 110 includes an integrated ID management server 111 that centrally manages and distributes ID information. The integrated ID management server 111 includes a recovery module 112 that manages roll forward. “Centralized management of ID information” is to collectively manage ID information (ID or ID and password) of a plurality of users (users, for example, employees) of the business system 120.

  Further, the management system 110 includes an LDAP (Light Weight Directory Access Protocol) server 113 having a database 114 for storing directory information as a repository for storing ID information. Update information of ID information from the integrated ID management server 111 is reflected in the database 114 of the LDAP server 113 by the LDAP request 115.

  The administrator console 141 is connected to the integrated ID management server 111 so that data can be transmitted and received. Maintenance information, periodic data updates from the personnel database 101, and the like are transmitted to the integrated ID management server 111 to the integrated ID management server 111 in accordance with the operation of the administrator on the administrator console 141.

  The business system 120 includes a reception agent 121 that is notified of data update in the integrated ID management server 111, a log file 126 that records update information from the integrated ID management server 111 or the client terminal 131, a business application interface (I / F) 127 and business application 128 are included.

  The reception agent 121 includes a data conversion module 122, a data transmission module 123, a data conversion table 124, and a log output module 125. Data relating to the business distributed to the business system 120 is converted into a form suitable for the distribution destination by the data conversion module 122 and then supplied to the business application I / F 127. Therefore, the distributed data is reflected in the business application 128 via the business application I / F 127.

  The log output module 125 outputs the distributed data (for example, update information of ID information) to the log file 126. In addition to the route from the integrated ID management server 111, there is a route from the client terminal 131 as a delivery route for distribution data to the business system 120.

  The client terminal 131 is used by a user. In response to a user operation on the client terminal 131, for example, a query 132 is sent to the business system 120. In the business system 120, the reception agent 121 is arranged so that the query 132 can be acquired before the business application I / F 127 (in Active Directory, a program loaded in the Local Security Authority). When the reception agent 121 acquires the query 132, the data conversion module 122 converts the query 132 into a form that can be taken into the integrated ID management server 111 with reference to the data conversion table 124. Then, the data transmission module 123 transmits a query 132 to the integrated ID management server 111.

  The update information included in the query 132 is output from the log output module 125 to the log file 126. Thereafter, the query 132 is reflected in the business application 128 via the business application I / F 127.

  FIG. 2 is an explanatory diagram illustrating an example of information output to the log file 126. In the example illustrated in FIG. 2, one record includes an update occurrence time (not a distribution time), update source server host information, and update destination server host information. In the example shown in FIG. 2, the update content is represented in a LDAP data interchange format (LDIF) format that is an LDAP data exchange format.

  When the information illustrated in FIG. 2 is created, the reception agent 121 sets the update source server (a server in which an update of ID information or the like has occurred) and an update destination server (a server to which an update of ID information or the like is notified) as follows: Judgment is based on these criteria.

  The reception agent 121 is an update that occurs in the integrated ID management server 111 (including a case where an update that occurred in a certain cooperation destination system is pulled back via the reception agent 121), and is a distribution destination system (in this embodiment). , The update source server host name and the update destination server host name are both set as the host name of the integrated ID management server 111. When the update source is the integrated ID management server 111 and the update is a distribution target update to the distribution destination system, the reception agent 121 updates the update source server host name as the host name of the integrated ID management server 111. The destination server host name is the host name of the delivery destination system.

  In the case of an update that occurred in the distribution destination system, the reception agent 121 uses both the update source server host name and the update destination server host name as the host name of the distribution destination system.

  FIG. 3 is an explanatory diagram showing an example of how the log file is recorded in the receiving agent. In FIG. 3, in addition to the integrated ID management server 111 (indicated by IDM (IDentity Management) in FIG. 3), in general, a plurality of distribution destination systems are assumed. Are shown as SYS1 and SYS2.). In each of SYS1 and SYS2, the reception agent 121 is configured as shown in FIG.

  FIG. 3 illustrates that the update (mod1, mod2, mod3) has occurred at each of the update times t1, t2, and t3. mod1 is an update that occurs in the integrated ID management server 111 and that the distribution destination systems are SYS1 and SYS2. mod2 occurs in the integrated ID management server 111, and the distribution destination system is an update only for SYS2. mod3 is an update that occurred in SYS1. For SYS2, mod3 update corresponds to an update that occurred in the integrated ID management server 111 (when it is pulled back via the reception agent 121).

  Therefore, according to the above criteria, the contents of the log files 126 of SYS1 and SYS2 are as illustrated in FIG.

  It is conceivable that the same attribute value is changed at a short interval in a plurality of delivery destination systems (SYS1 and SYS2 in the example shown in FIG. 3). In this case, in consideration of the update occurrence time, if the update occurrence time of the subsequent update is chronologically earlier than the immediately preceding update occurrence time, the subsequent agent is controlled not to be processed by the reception agent 121.

  FIG. 4 is an explanatory diagram for explaining the unprocessed update of the subsequent update, taking the update of the userPassword attribute value as an example. In the example shown in FIG. 4, in SYS1, the userPassword attribute value is updated for the password b at time t1, and then the userPassword attribute value is updated for the password c at time t2 in SYS2.

  The receiving agent 121 in SYS1 records the update of the userPassword attribute value at time t1 in the log file 126, but the update information is transmitted to SYS2 via the integrated ID management server 111. In SYS2, it is assumed that the arrival time (delivery time) of the update information is later than time t2.

  In that case, after the update at time t2 is recorded in the log file 126 in SYS2, the update information at time t1 arrives. The receiving agent 121 of SYS2 does not record the update information at time t1 in the log file 126.

  FIG. 5 is a flowchart showing recovery processing by roll forward in the system shown in FIG.

  Assume that a failure occurs in the integrated ID management server 111 and the database 114 is restored with backup data at a certain past time. The integrated ID management server 111 communicates with the reception agent 121 of the business system 120 that is the distribution destination of information (reception agents of all business systems when there are a plurality of business systems) at the time of starting the service after recovery. And the receiving agent is notified of its last update time (the time when the integrated ID management server 111 in the restored database 114 last updated the data) (step S101).

  The reception agent 121 (the reception agent of all business systems when there are a plurality of business systems) is a record generated after the last update time of the integrated ID management server 111 from the record recorded in its own log file 126. And the extracted record is sent as an LDAP request to the integrated ID management server (step S102).

  The integrated ID management server 111 sorts the records acquired from the receiving agent 121 (the receiving agents of all business systems when there are a plurality of business systems) in order of update time (step S103). Then, the integrated ID management server 111 performs roll forward recovery by sequentially inputting the sorted records to the LDAP server 113 (step S104).

  As described above, in this embodiment, update information generated in the integrated ID management server 111 (including update information generated in a business system different from the business system to which update information is distributed) is included in each business system. The distribution information is distributed to the reception agent, and the reception agent records the distributed update information in the log file 126 as a record. In addition, the reception agent in each business system records update information generated in its own system as a record in the log file 126 and distributes it to the integrated ID management server 111.

  By executing such control, the integrated ID management server 111 can utilize the log record held by each receiving agent as journal information when recovering by roll forward.

  Therefore, the integrated ID management server 111 does not require a backup server for holding at least journal information. That is, it is possible to perform roll-forward data recovery when a failure occurs while suppressing an increase in cost.

  Also, the receiving agent in each business system records update information in the system illustrated in FIG. 1 as needed. Therefore, even after the failure occurs in the integrated ID management server 111 and until the service is restored, the data from the restored last update time to the latest time is distributed even if the service of each business system that is the distribution destination is continued. Roll forward recovery can be performed so that consistency is achieved between the previous system (business system) and the integrated ID management server 111. In other words, roll-forward data recovery can be performed when a failure occurs while suppressing an increase in cost, and operation of the business system can be prevented from being disturbed even if the failure continues.

  In the above embodiment, a system including the management system 110 including the integrated ID management server 111 that centrally manages and distributes the ID information and the business system 120 including the reception agent 121 is taken as an example. The target is not limited to a system having such a configuration.

  Any system in which the system or apparatus having the reception agent 121 having the function of recording update information in the log file 126 in the above embodiment and the system or apparatus having the update information distribution function in the above embodiment are linked. The present invention can also be applied to systems other than the information processing system of the above embodiment.

  FIG. 6 is a block diagram showing the main part of the information processing system according to the present invention. The information processing system 10 illustrated in FIG. 5 includes a management server 20 (for example, an integrated ID management server) that manages specific data using the database 40 (for example, the database 114 of the LDAP server 113), and distribution of specific data. Receiving the distribution destination system 30 (for example, the business system 120), the management server 20 includes at least information that can specify the update contents and information that indicates the source of the update when the update of specific data occurs. Is distributed to the distribution destination system 30, and the distribution destination system 30 records the update information as a log in the log file 32 (for example, the log file 126) when the update of specific data occurs in its own system. In addition, update information management means 31 (for example, the reception agent 121) that transmits update information to the management server 20 The update information management means 31 records the update information as a log in the log file 32 when the update information is distributed from the management server 20, and the management server 20 records it in the log file 32 in the distribution destination system 30. Restore the database by roll-forward using the recorded log as journal information.

DESCRIPTION OF SYMBOLS 10 Information processing system 20 Management server 30 Delivery destination system 31 Update information management means 32 Log file 40 Database 100 Personnel system 101 Personnel database 110 Management system 111 Integrated ID management server 112 Recovery module 113 LDAP server 114 Database 115 LDAP request 120 Business system 121 Receiving agent 122 Data conversion module 123 Data transmission module 124 Data conversion table 125 Log output module 126 Log file 127 Business application I / F
128 Business application 131 Client terminal 132 Query 141 Administrator console

Claims (8)

An information management method executed in an information processing system including a management server that manages specific data using a database, and a distribution destination system that receives distribution of the specific data,
When the management server is updated, when the update of the specific data occurs, the management server distributes update information including at least information capable of specifying the update content and information indicating the source of the update to the distribution destination system,
When the delivery destination system updates the specific data in its own system, it records update information as a log, and sends the update information to the management server,
When the delivery destination system delivers update information from the management server, the update information is recorded as a log,
The information management method, wherein the management server restores the database by roll forward using a log in the distribution destination system as journal information. The information management method according to claim 1, wherein the specific data is ID information. The information management according to claim 1 or 2, wherein when the specific data is updated in the local system, the distribution destination system converts the data into a format handled by the management server and then sends the update information to the management server. Method. The update information includes time information indicating the time when the update occurred,
When update information is distributed from the management server, if the update information that occurred before the time indicated by the time information included in the update information is recorded as a log, The information management method according to claim 1, wherein update information is not recorded. An information processing system including a management server that manages specific data using a database, and a distribution destination system that receives distribution of the specific data,
The management server distributes, to the distribution destination system, update information including at least information capable of specifying update contents and information indicating an update generation source when an update of the specific data occurs.
The delivery destination system includes update information management means for recording update information in a log file as a log when update of the specific data occurs in the own system, and sending update information to the management server,
The update information management means records the update information as a log in the log file when the update information is distributed from the management server,
The information processing system, wherein the management server restores the database by roll-forward using a log recorded in the log file in the distribution destination system as journal information. The specific data is ID information;
The information processing system according to claim 5, wherein the management server is an integrated ID management server that centrally manages ID information. 6. The update information management means in the distribution destination system sends the update information to the management server after conversion into a format handled by the management server when the specific data is updated in the own system. Item 7. The information processing system according to Item 6. The update information includes time information indicating the time when the update occurred,
The update information management means in the distribution destination system, when update information is distributed from the management server, when update information generated before the time indicated by the time information included in the update information is recorded in the log file The information processing system according to any one of claims 5 to 7, wherein the distributed update information is not recorded in a log file.

Images