JP3501944B2 - Database management device and recording medium on which the program is recorded - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a database management device capable of making the contents of data the same as that of the other party by communication, and in particular, a database management device capable of reducing the amount of data communicated or processed during synchronization processing, and The present invention relates to a recording medium on which the program is recorded.

[0002]

2. Description of the Related Art Conventionally, information processing devices have been widely spread, and in recent years, for example, high-performance information processing devices such as desktop personal computers and information processing devices having excellent portability such as personal digital assistants. It is not unusual for individuals to own multiple information processing devices. Here, when managing personal information such as a schedule or an address book by these plural information processing devices, it is required that the databases provided in any of the information processing devices have the same data. To be done. Therefore, for example, in the case of managing personal information between a personal computer and a portable information terminal, for example, a personal database is managed by a personal computer, and the data of the master database is transferred to a portable information processing device database (portable database). I copy and carry my personal information with me.

As described above, when the same data is managed by a plurality of database management devices, when the data is changed by a certain database management device, all the database management devices also need to update the databases. Here, if each database management device can always communicate,
If the updated database management device notifies another database management device of the data change, the data change can be immediately reflected. However, each database management device cannot always communicate, for example, when the database management device is a portable information terminal. In this case, when the data is changed, the data change cannot be reflected in the database of another database management device. Therefore,
In order to maintain the sameness of the databases, a process for making the contents of the database the same as that of the other party, that is, a synchronization process is required when each database management device becomes communicable.

Explaining in the above example, when data is changed in the portable information terminal, synchronization processing is performed when the portable information terminal is moved back to the home or office, and the portable database of the portable information terminal and the master on the personal computer. The data will be the same with the database.

Here, download / upload can be given as an example of the conventional synchronization processing. In particular,
When the data is changed on the personal computer, the contents of the master database are copied to the mobile information terminal (download). On the other hand, when the data is changed on the mobile information terminal, the contents of the mobile database are copied (uploaded) to the personal computer.

[0006] Another conventional example is synchronization processing using a union. In this synchronization processing, when the two database management devices can communicate with each other, the contents of both databases are compared, and a new set is created as a union set containing all the information added to each database. To be done. After that, the created database is stored in both database management devices.

For example, both databases initially contain data "A", and one database management device X
When the data of “B” is added in the above, and the data of “C” is added in the other database managing apparatus Y, the database managing apparatus X sends the content “A + B” held by itself to the database managing apparatus Y. To do. On the other hand, the database management device Y creates a union “A + B + C” of the received “A + B” and the content “A + C” held by itself,
Store as your own database. Further, the database management device Y determines that the contents of the database “A + B +
"C" is sent back to the database management device X. As a result, the database management device X sets the contents of its own database to the same “A + B + C” as the database management device Y.
Can be updated.

[0008]

However, in the conventional database management apparatus which performs the above-mentioned synchronization processing, the amount of data (communication amount) transmitted between both database management apparatuses and each database management in the synchronization processing. This causes a problem that the amount of data calculated by the device is extremely large.

Specifically, in downloading / uploading, all the contents of one database are transmitted to the other database management device regardless of the number of changed data. For example, in the case of a database having 1000 data, even if one data is changed, 10
00 data is transmitted. As a result, the amount of communication transmitted between both database management devices increases. At the same time, in the database management device on the receiving side, the contents of the database are replaced by the database on the transmitting side, so that the data change on the receiving side is lost.

Further, in the synchronous processing using the union, one database is transmitted to the other in order to create the union, and the created union is sent back as a new database. As described above, in the synchronization processing, all the contents of the database have to be reciprocated between the two database management devices, which further increases the communication amount during the synchronization processing.

In addition, when the data is deleted in only one side, the data remains in the other side database. Therefore, the new database contains the data again and the data is not deleted. As a result, the amount of memory required to store the database increases and the amount of communication during synchronization processing further increases.

In order to solve these problems, for example,
Japanese Patent Laid-Open No. 9-6693 describes, as another example of the synchronization process, a database management device in which update information indicating whether or not there is a change is added to all data included in the database. Specifically, each database management device stores, for each data included in the database, ID information for identifying the data and update information indicating no change / correction / addition / deletion. When the data is changed, the update information corresponding to the data is updated. Furthermore, at the time of synchronization processing, each database management device notifies the other party of ID information and update information of all data. On the other hand, the counterpart database management device searches the data corresponding to the data in its own database based on the received ID information and compares the update information of both. Furthermore, when only one is changed, the unchanged data is replaced with the changing data. If both have been changed, the other party's data is added to its own database, and its own data is added to the other party's database. As a result, when the data is not changed in any of the database management devices, the sending of the data can be omitted, and the communication amount at the time of synchronization processing can be reduced.

However, in the above configuration, when the data is corrected by both sides, each is added to the other database, and the number of data contained in the database increases. As a result, the amount of memory required to store the database increases and the amount of communication also increases. Further, in order to suppress the increase in the number of data, it is necessary to delete unnecessary data after the synchronization processing. However, this operation requires extracting data corrected on the one hand and data corrected on the other hand from the database after synchronization processing, comparing both, and specifying the data to be deleted, which is very troublesome. Takes.
Furthermore, since one database management device is usually operated, it is necessary to reflect the deletion of data in the other database management device, which further increases the communication volume between both database management devices.

The present invention has been made in view of the above problems, and an object thereof is not to increase the scale of a database even when a plurality of corresponding data are independently updated. It is to realize a database management device capable of performing synchronous processing with a small amount of effort and a small amount of communication data, and a recording medium in which the program is recorded.

[0015]

In order to solve the above-mentioned problems, a database management apparatus according to a first aspect of the present invention includes a plurality of database management apparatuses and a database stored in the database management apparatus that can communicate with each other. Between,
Provided for configuring a database system having a synchronization processing means for matching the contents of corresponding data,
A database management device having at least one database is characterized by taking the following means.

That is, a state in which, for each of its own data corresponding to the data in the other party's database, state information indicating a change in the data is stored during the asynchronous period between the previous synchronization processing and the next synchronization processing. For the information storage means and the set of data corresponding to each other during the synchronization processing,
Collision detection means for detecting whether or not there is a plurality of changed data in the data set based on the respective state information, and in the case of a plurality of collisions, before transmitting all the contents of the changed plurality of data, Priority data determined by its own priority side determination means and the synchronization processing means that causes the synchronization processing means to replace the content of the non-priority side data with the content of the priority side data. Side or opponent
And a detection unit that detects a mismatch with the designated priority side . Note that the data change is a process that may affect other databases, and indicates at least one type of data content update, addition or deletion, or data position movement in the database, for example.

In the above structure, the status information is provided in at least the data of the own database that is associated with the data of the other database, and is updated in response to an operation on the data during the asynchronous period. It

When the synchronization processing is instructed, the collision detection means receives its own status information and the other party's status information before transmitting the respective data contents in order to match the contents of the associated data sets. Based on the state information, it is determined whether or not a data collision has occurred in the data set. In addition, regarding what kind of change, the collision of data can be detected depends on the change indicated by the state information.
At least, it is possible to detect a collision related to a data change of the type indicated by the state information.

Here, the state information indicates the change of data during the asynchronous period, and the amount of data is smaller than the content of the data. Therefore, it is possible to detect data collision with a smaller amount of data than in the case of transmitting and receiving data contents.

When a data collision is detected, the priority side determination means sends the data contents to the database management device determined by the synchronization processing, for example, the user or a higher-level application before transmitting the entire content of the data. Inquiry result of
Alternatively, priority data is determined from the data set based on a part of the data content. Further, the synchronization processing means replaces the data content on the non-priority side with the data content on the priority side.

In the above configuration, in the event of data collision, prioritized data is determined before the entire contents of the data are transmitted,
The entire data content is unidirectionally transmitted from the database management device on the priority side to the database management device on the non-priority side. Therefore, compared with the case where the priority side is decided and the data contents are replaced after transmitting all the data contents,
The amount of data transmitted and processed between the database management devices can be reduced. Furthermore, since the priority side is determined at the time of data collision, the number of data included in each database does not increase even when the data is updated by a plurality of database management devices. As a result, it is possible to reduce the amount of memory required to store the database and the amount of data transmitted and processed during the synchronization process.

Here, a plurality of database management devices are better
If there is a front-side decision means, it is decided by each
The priority side may not match. In this case, synchronous processing
The means is the priority side determined by a database management device
When the synchronization process is performed based on the
The content of the data determined as the priority side by the management device is undesired.
Will disappear.

On the other hand, the above database management device
Is the priority side determined by one's own priority side determination means, and
Equipped with detection means to detect inconsistencies with the designated priority side
It is characterized by having.

According to this structure, the detecting means is provided on each priority side.
Is detected, the above undesired data contents are deleted.
Loss can be prevented. Therefore, for example, each database
Prompt the management device to re-determine the priority side, or
Appropriate processing such as saving all data
it can.

Further, in the database management apparatus according to a second aspect of the present invention, in the configuration of the first aspect of the invention, the priority side determining means is only a part of each of the changed contents of the plurality of data. Is characterized by determining the priority side.

According to the above configuration, the priority side determining means determines the priority side based on a part of the data contents such as the last update time. Therefore, it is possible to reduce the amount of data to be transmitted and processed in the synchronization processing, as compared with the case of transmitting the entire content of data.

On the other hand, in the database management apparatus according to the third aspect of the present invention, in the configuration of the first aspect of the invention, when there is a plurality of changed data, which data has priority. The user or higher
It is characterized by inquiring to the application .

In the above-mentioned configuration, when data collides, the priority side determining means, for example, information indicating the state of the collision, such as its own data content and status information of each, or
Such as to notify the information indicating the occurrence of a collision, the application and the user of the higher-level, query the data should be given priority,
The priority side is decided based on the answer. Here, the information notified at the time of inquiry is often smaller than the entire data content, and in particular, it is not necessary to newly transmit its own data content and each status information from another database management device. Therefore, it is possible to reduce the amount of data transmitted and processed during the synchronization processing. In addition, the priority side can be determined more accurately because the instructions of the user and the higher-level application can be reflected in the determination of the priority side.

In addition to the above configuration, when there are a plurality of changed data, a priority side information storage means for storing priority side information indicating which data is prioritized is provided, and the priority side determination means is provided. The priority side may be determined based on the priority side information . Note that the priority side information may be set by an instruction from the user or a higher-level application as long as it is stored by the time the priority side is determined. It may be set based on.

According to the above configuration, the priority side determining means determines the priority side based on the priority side information.
It is possible to reduce the amount of data transmitted and processed during the synchronization processing.

According to a fourth aspect of the present invention , in the database management apparatus according to the third aspect of the present invention, the priority side determining means determines which data has a higher priority when there are a plurality of changed data . User or higher level application
® was Awa query to down, the preferential is determined based on the answers, and the preferential, based on the preferential designated by the other party, it instructs the preferential to the synchronization processing means, and the query It is characterized by disconnecting the communication with the other party once between instructions.

According to the above configuration, the communication is disconnected between the inquiry and the instruction on the priority side. So, for example,
Even if it takes a long time for the database management device to make an inquiry and get a response, such as when the user is out of the seat or when the priority side is determined by a complicated calculation, it is necessary for the synchronization processing. The total communication time can be kept constant. As a result, it is possible to reduce communication costs and power consumption.

In addition to the above-mentioned configuration, it is provided with a subset setting means for setting a subset in its own database, and the state information storage means carries out the previous synchronization processing to the subset shared with the other database. The state information may be stored for the data included at the point of .

In the above configuration, since the sharing range can be specified independently of the size of the database, it is possible to omit the synchronization processing of data that does not need to be shared without newly creating a dedicated sharing database. Therefore, it is possible to reduce the amount of data transmitted and processed during the synchronization process without increasing the amount of memory required to store the database.

In addition to the above configuration, the synchronization processing means performs a synchronization process on the data included in the shared range of each database, and the database management device stores the shared range of its own database and the other database. A sharing setting means for associating with the sharing range may be provided. The shared range may be a subset of the database or the entire database. Further, a range existing before the sharing setting, such as the entire database or an already existing subset, may be designated as the sharing range, or a new range may be created by creating a subset.

According to the above configuration, the sharing of the data in the sharing range can be designated only by associating the sharing ranges, so that it is possible to designate the data to be shared with less operation than in the case of associating each data.

Further, in addition to the above configuration, the synchronization processing means performs a synchronization process of data included in the shared range of each database, and the database management device associates the shared ranges associated with each other. A sharing canceling means for canceling may be provided. The sharing canceling unit may cancel only the association and leave the subset within the sharing range or the data within the sharing range, or may delete the subset or the data within the sharing range.

According to the above configuration, the association between the shared ranges can be canceled when the subsequent synchronization processing becomes unnecessary.
Therefore, it is not necessary to monitor the status of the data within the shared range to update the status information, and it is possible to suppress the amount of memory required to store the status information and the amount of calculation for updating the status information.

In order to solve the above-mentioned problems, a recording medium according to a fifth aspect of the present invention is a recording medium in which a program for managing the same contents of data corresponding to each other among a plurality of databases is recorded. The state information storage means, the collision detection means , and the priority side determination means according to claim 1.
And, as a detecting means , a program for operating a computer is recorded.

When the computer reads the recording medium having the above configuration and executes the program, the computer
It operates as the database management device according to claim 1.
Therefore, similarly to the first aspect, it is possible to reduce the amount of memory required to store the database and the amount of data transmitted and processed during the synchronization processing.

[0041]

BEST MODE FOR CARRYING OUT THE INVENTION [First Embodiment] The following will describe one embodiment of the present invention with reference to FIGS. 1 to 61. That is, the database system according to the present embodiment is limited to a plurality of data management devices that manage individual information and the like by a plurality of database management devices and have a limited period during which they can communicate with each other and can update data independently. It is a system suitably used when it is required to keep the contents of corresponding data identical between the databases. Although various system configurations are conceivable, the present embodiment will be described by taking a database system including database management devices, which are equal to each other, as in the case of a mobile information terminal, as an example.

Specifically, for example, as shown in FIG.
The database system 1 according to the present embodiment is an information processing device (database management device) that stores a database.
Each of the information processing devices 2 includes a communication device 3 for communicating with the other information processing device 2, and a display device 4 and an input device 5 for processing input / output with the user. And are connected.

The communication device 3 is, for example, an interface such as a LAN interface or an RS232C interface, a modem, an infrared transmitter / receiver, or a radio transmitter / receiver. Via at least one of various communication media, such as a network, infrared or wireless,
The communication device 3 of the other party can communicate with each other. The display device 4 is a CRT display, a liquid crystal display, or the like, and displays various information such as the contents of the database of the information processing device 2 or the operation result on the database.
It can be displayed according to an instruction from the information processing device 2. Further, the input device 5 is a keyboard, a mouse, a pen, a touch panel arranged on the display device 4, or the like,
The user's instruction can be input to the information processing device 2.

On the other hand, as shown in FIG. 1, the information processing device 2 according to this embodiment is connected to the communication device 3 and communicates with the other information processing device 2 via the communication device 3. The user operation unit 2 which is connected to the display unit 4 and the input device 5 and serves as an interface with the user.
2, the database management unit 23 that stores the database DB, the database processing unit 24 that controls the database management unit 23, the communication unit 21 and the database management unit 23 according to instructions from the user operation unit 22, A synchronization unit 25 that performs processing related to synchronization with another database DB is provided. Each of the above-mentioned members may be realized by hardware, or may be a functional block realized by the CPU executing a predetermined program. In the latter case, each member can be realized only by distributing a recording medium in which the program is stored and executing the recording medium on a computer that can read the recording medium.
Distribution is easy. The database management unit 23
Corresponds to the database, the state information storage means, and the priority side information storage means described in the claims.

In the present embodiment, as shown in FIG. 3, the database DB is managed in units of a tip T indicating one piece of data and a bag F which is a population of the tips. A tip management unit 31 and a bag management unit (subset setting means) 32 for storing a database are provided. Further, the database processing unit 24 includes a tip update processing unit 41, a tip addition processing unit 42, a tip deletion processing unit 43, etc. that respectively update / add / delete the tip T. The database processing unit 24 also includes members related to other operations of the database DB, such as searching and displaying the tip T or changing the inclusion relationship with the bag F. As a result, a tip T that satisfies a desired condition, such as a tip T having a predetermined content, is searched to generate the bag F, or a desired tip T is selected from the displayed tips T to select a desired tip T. It can be inserted into the F bag.

Further, the synchronizing section 25 includes a synchronizing processing section 51 for performing the synchronization processing of the tip T and a sharing setting processing section 52 for performing a sharing setting described later prior to the synchronization processing.
Further, for example, a sharing cancellation processing unit 53 that cancels the sharing setting when synchronization processing becomes unnecessary is provided. The synchronization processing unit 51 corresponds to the synchronization processing unit, the collision detection unit, the priority side determination unit, and the detection unit described in the claims, and the sharing setting processing unit 52.
The sharing cancellation processing section 53 corresponds to the sharing setting means and the sharing cancellation means, respectively. Further, the configurations of the database processing unit 24 and the synchronization unit 25 are closely related to the data stored in the database management unit 23, and the operation based on the data has a great feature. Therefore, hereinafter, only the configuration of the database management unit 23 will be described, and the configurations of the database processing unit 24 and the synchronization unit 25 will be described in detail together with the operation.

Specifically, the tip management section 31 is
Data for one case is managed by a data structure called a tip. Each tip T is a tip I for uniquely identifying itself from other tips T in the database DB.
D and an attribute list LA in which data for one case is stored. In the attribute list LA, as shown in FIG. 4, the attribute name NA indicating the type of the attribute A and the tip T
The combination with the attribute value VA indicating the value of the attribute A in the data stored in is stored in a list.

For example, the case of storing the address book in the database DB will be described as an example.
It corresponds to one address data. If each address data is composed of three types of data, that is, data indicating a name, data indicating an address, and data indicating a telephone number, each tip T
Has _three attributes A _{1 to} A _{3 of} name, address and telephone number.
Attribute list LA of each tip T defined as a set of
As shown in FIG. 4, the combination of the attribute name NA and the attribute value VA is expressed as a list arranged in order. For example, in the attribute list LA of the first tip, a combination of an attribute name NA ₁ of “name” and an attribute value VA _{1 of} “○ yama × husband” is stored as the attribute A ₁ of the name. Similarly, a combination of "address" and "Tokyo city, Tokyo ...", and "phone number" and "0123 -..."
And a combination of are stored.

The bag management unit 32 stores the bag F included in the database DB and the inclusion relationship between the tips T and bag F. As shown in FIG. 5, one information processing device 2 has a plurality of database DBs.
_{The case where 1} , DB ₂ , and DB ₃ are stored will be described as an example. The relationship of the information processing device 2∋database DB∋float F∋tip T is established. The bag management unit 3
2 stores this relationship as a tree structure S as shown in FIG.

In the tree structure S, like the tip T ₂ , it may be included in a single bag (F ₁ ),
Like the tip T ₁ , it may be included in a plurality of bags (F ₁ · F ₂ ). As a result, as shown in FIG.
The tip T ₁ included in the plurality of bags F ₁ and F ₂ can be expressed.

If the tree structure S can be stored,
The entire tree structure S may be stored, or the tree structure S may be divided into a plurality of lists and a list of database DBs included in the information processing device 2 and a list of bag F for each database DB may be stored. It may be stored as a list of tips T for each bag F. In the present embodiment, as in the case of the tip T, an ID for uniquely identifying each information processing device 2, each database DB, and each bag F is assigned, so that each list above
It is stored as a list of the ID. When stored as each of the above lists, the tip T included in a certain bag F is immediately found by referring to the list. On the contrary, the bag F including the tip T is the worst case. , I need to search all the lists. therefore,
When the search time is required to be shortened, the bag F including the tip T for each tip T, apart from the above-mentioned lists.
It is better to have a list of.

Database system 1 according to the present embodiment
Then, as shown in FIG. 7, the range (shared range) that is the target of the synchronization processing is determined in units of the F bags F, and whether or not a certain tip T is shared between other database DBs is shared in advance. Then the set bag (shared bag S
It is determined by whether or not it belongs to F).

Here, normally, access to another information processing device 2 via the communication device 3 or the like is extremely slow as compared with access within the information processing device 2 of its own, and the time required for the synchronization processing increases. It tends to be. Therefore, there is a strong demand for shortening the required time, and it is desired to reduce the amount of data transmitted between the information processing devices 2a and 2b. Since the access to the information processing device 2 is relatively fast, even if the processing during the period between the previous synchronization processing and the next synchronization processing (during the asynchronous period) increases, the processing time is Not too long.

In order to reduce the amount of data, the bag management unit 32 according to the present embodiment stores a dirty bit list LD consisting of dirty bits D indicating the state of each tip T during the asynchronous period. . Dirty
In the bit list LD, the order of the dirty bits D is managed so that the dirty bit D of the tip T corresponding to each other in the self and the opponent is the same.
As will be described later, the correspondence relationship between tips T is also shown. Further, the pocket management unit 32 stores a sync limit Lim indicating the number of tips T already associated with the previous synchronization process, and the tip T corresponding to the dirty bit D is shared by the partner. It can be determined whether or not it is already associated with the tip T in the SF.

For example, as shown in FIG. 8, in the previous synchronization processing, three tips (T _{1a to} T _3a and T _{1b to} T _3b ) are associated between both shared bags SF _a and SF _b . In this case, "3" is stored as the sync limit Lim in both the information processing devices 2a and 2b. Furthermore, as shown in FIG. 9, the dirty bit list LD _a of the information processing apparatus 2a, the dirty bits D _1a to D _6a corresponding to the tip T _1a through T _6a is stored in this order . Similarly, the dirty bit of the information processing device 2b
In the list LD _b , dirty bits D _{1b to} D _3b , D _7b and D _8b corresponding to the respective tips T _{1b to} T _3b , T _7b and T _8b are stored in this order. In this case, since the sync limit Lim is “3”, the tips T _{4a to} T _6a in which the dirty bit D is stored after the third bit,
And, T _7b through T _8b are not associated with the tip T in the shared bag SF of the other party. Note that FIG. 9 shows an example of a method of storing the dirty bit list LD in which a combination of a dirty bit D and a tip ID corresponding to the dirty bit D is sequentially stored.

Hereinafter, for convenience of explanation, the order of the dirty bits D in the dirty bit list LD is referred to as a sync tag, and is represented as a number in [] in the drawing. Also, Dirty Bit D and Tip T, etc.
Corresponding data are denoted with the same subscript, for example, dirty bit D _1a and tip T _1a . Further, the information stored in one of the information processing devices 2a is attached with _a subscript _a such as, for example, tip T _1a to be distinguished from the information stored in the other information processing device 2b (subscript _b is attached). . The subscripts are omitted when it is not necessary to specify the correspondence or when the names are collectively referred to.

The dirty bit D is created corresponding to each of the tips T existing in the shared bag SF even at least once during the asynchronous period, and it is determined whether the corresponding tip T is updated during the asynchronous period. It has an update parameter that indicates and a movement parameter that indicates whether or not the tip T has moved inside and outside the shared bag SF during the asynchronous period. The update parameter takes a value of "m" indicating update or "p" indicating maintenance, and the value of the movement parameter is "X" indicating stillness, "I" indicating insertion, or "I" indicating ejection. It is either "O". More specifically, "X" is
It has been shown that it has continued to exist in the shared bag SF since the last synchronization process, and "I" has been put in the shared bag SF since the last synchronization process or has been returned. . Similarly, “O” has been extracted from the shared bag SF since the previous synchronization processing, or
It shows that it was inserted once and then removed again. The movement parameter is stored, for example, as shown in FIG. 10, as a combination of an addition bit BI indicating addition to the shared bag SF and a deletion bit BO indicating deletion from the shared bag SF. Further, for example, when it is determined as a result of the synchronization processing that the tip T exists outside the shared bag SF, and thereafter, when it is unnecessary to monitor the state of the tip T, both bits BI and BO are set to ON. Hereinafter, this state is referred to as invalid "K" regardless of the value of the update parameter. Therefore, the states of the tip T which can be identified from the dirty bit D are the static maintenance "Xp", the static update "Xm", and the insertion maintenance "I".
There are seven: p ”, insert update“ Im ”, discharge maintenance“ Op ”, discharge update“ Om ”, and invalid“ K ”.

By the way, even if the same shared bag SF is used, if the shared bag SF of the other party is different, the asynchronous period with each is usually different from each other. Therefore,
In the present embodiment, the dirty bit list LD and the sync limit Lim are provided for each combination of the shared bag SF of itself and the shared bag SF of the other party.

For example, as shown in FIG. 11, the shared bag SF _a has three different shared bags SF _b , SF _c , S.
When it corresponds to F _d , as shown in FIG. 12, the corresponding dirty bit list LD _ab and LD _ac
・ LD _ad and sink limit Lim _ab・ Lim
_ac · Lim _ad is provided. As shown in FIGS. 11 and 12, each dirty bit list LD may be in the same order as the dirty bit list LD of the other party, and the dirty bit list (LD _ab
The order or value of the dirty bit D or the sync limit Lim may be different between LD _ac ). For example, in the dirty bit list LD _ab , the dirty bit D _2ab (whose value is Im) indicating the tip T _2a is stored second, but the dirty bit
In the bit list LD _ac , dirty bit D
_2ac (value is Xp) is stored first. However, dirty bit list LD of shared bag SF _{b
In b} , the dirty bit D _2b indicating the tip T2 _b is stored second, and the dirty bit list L
In D _c , the dirty bit D _2c is stored first.

In the bag management unit 32 according to this embodiment,
In addition to each of the above lists, a sharing partner list LS indicating a sharing partner of the sharing bag SF is provided for each sharing bag SF. By referring to this list LS, it is possible to easily search all the sharing partners of a certain shared bag SF, so that, for example, a change such as addition, deletion, or content update is made to the tip T in the shared bag SF. In this case, all the dirty bits D corresponding to the tip T can be updated in a shorter time.

The operation of the database system 1 having the above configuration will be described with reference to FIGS. 13 to 61. That is, as roughly shown in FIG. 13, the sharing setting is performed prior to the actual synchronization processing, and for example, the setting of the shared bag SF of itself and the other party, the dirty bit list LD, and the sync limit Lim are set. The processing required for the synchronization processing, such as the generation of (S1), is performed (S1). At this stage, as shown in FIG. 14, both sharing bags SF _a · SF _b
It is only registered in both the information processing devices 2a and 2b that the range is a shared range, and the contents of both shared bags do not match until the synchronization process is performed.

After that, the dirty bit list LD is updated during the asynchronous period based on the operation to the database DB of itself, regardless of whether the both information processing devices 2a and 2b can communicate. (S2). Further, during a period in which both the information processing devices 2a and 2b can communicate, a synchronization process is performed, and as shown in FIG.
Aligned with reference to bit list LD and sync limit Lim, and dirty bit list L
D and sync limit Lim are also updated (S
3). When the communication path between the information processing devices 2a and 2b is intermittently connected, the shared bag whose consistency is maintained by the intentional synchronization process as in S3 is very useful.

Since the tip T can be put in a plurality of the pockets F, when the synchronization processing is executed in any of the sharing settings, the tip T belonging to the cloth F which is not related to the synchronization processing is changed. It may be done. However, the tip T is not inserted into or removed from the bag F other than the shared bag SF regardless of which sharing setting is performed.

The processes of S2 and S3 are repeated as necessary, and when the subsequent synchronization process becomes unnecessary, S
In 4, the sharing setting is canceled. Although FIG. 13 describes that step S2 is first performed after step S1, but if both steps S2 and S3 are alternately repeated, the order of both steps S2 and S3 may be exchanged. .

Here, the sharing cancellation basically only cancels the correspondence between the sharing bags SF, and the sharing bag S
The contents of F do not necessarily match each other. Therefore, when the contents of the shared bag SF are made to coincide with each other, the synchronization process may be performed immediately before the cancellation of sharing, or automatically according to a user's instruction.

In the above description, the sharing setting and the synchronizing process are separately described, but the synchronizing process can be performed simultaneously with the sharing setting. In this case, for example, the contents of both shared bag SFs match at the end of sharing setting. However, in the present embodiment, the sharing setting and the synchronization process are separated to reduce the time required for the sharing setting. Accordingly, at the time of sharing setting, both information processing devices 2 confirm with each user whether sharing setting is possible, and at the time of synchronization processing, when only the user of one information processing device 2 operates, The total operation time by both users can be shortened. As a result,
Both improvement of safety and reduction of operation time can be realized.

In the following, as an example of the processing of S1,
The sharing setting of the bag fusion type model will be described in detail. As shown in FIG. 14, the sharing settings of the fusion type are as follows.
A balloon F _{a that is} present in advance in both information processing devices 2a and 2b
- a shared settings between F _b, as shown in FIG. 16, when both the information processing devices 2a, 2b is communicable, in S11, the shared setting processing unit 52 provided on one of the information processing apparatus 2a , And sends a sharing setting request to the other information processing device 2b. To the request includes a machine ID showing an information processing apparatus 2a, a bag ID showing the bag F _a of the information processing apparatus 2a, bag F of an information processing apparatus 2b to the desired share
_It includes a bag ID indicating _b . In the following, the side that sends the sharing setting request is referred to as the sharing setting activation side, and the side that receives the sharing setting request is referred to as the sharing setting passive side.

Further, the sharing setting processing unit 52 on the sharing setting starting side registers the same information as the sharing setting request as the sharing information in S12. Further, the sharing setting processing unit 52 causes the sharing partner list L corresponding to the own bag F _a.
Create S and register the opponent's bag ID. As a result, the bag F _a becomes the shared bag SF _a . In addition, a dirty bit list LD and a sync limit Lim corresponding to the combination of the shared bag SF _a and the partner bag F _b are created.

On the other hand, the information processing device 2b on the sharing setting passive side
Then, when the sharing setting processing unit 52 receives the sharing setting request in S13, the sharing information is registered based on the sharing setting request in S14. The shared information includes a machine ID indicating its own information processing device 2b, a bag ID to be shared by the information processing device 2a, a bag ID to be shared by the information processing device 2b, and determined at the time of sharing setting. Options are included. Furthermore, sharing setting processing unit 52, similarly to the sharing setting initiator, the shared contact list LS that corresponds to his bag F _b, and registers the bag ID of the other party, to generate a shared bag SF _b, itself The shared scrounger SF _b and the other's shared scrounger SF _a
Dirty bit list L corresponding to the combination with
Create D and Sync Limit Lim.

Database system 1 according to the present embodiment
Then, various types of information processing devices 2 may be used, for example, whether the priority side determination rule or the download process is necessary in the synchronization process of S3, the necessity of copying at the time of collision, the model adopted in the sharing cancellation of S4, or the like. Options (see below)
If the option is selectable, the option is set at the stage of sharing setting, and the sharing setting request also includes information indicating the option. Thereby, the setting time of each option can be limited to the sharing setting time.

Here, when each information processing device 2 can set an option at an arbitrary time, the priority side determination rule or the like is changed according to the operation on one of the information processing devices 2, and the other information processing device 2 is processed. There is a risk that the user of the device 2 may make undesired settings. However, in the above configuration, the option is changed only at the time of sharing setting at which each information processing device 2 can communicate, so that the above-mentioned undesired setting can be rejected.

Although FIG. 16 does not show the step of authenticating the sharing partner, whether or not the sharing setting passive side makes the sharing setting according to the contents of the sharing setting request such as the option of the sharing setting. May be determined. In this case, the information processing device 2 on the sharing setting passive side may store, for example, a receivable sharing setting request in advance, and determine the necessity based on the stored sharing setting request.

The initial values of the dirty bit list LD and the sync limit Lim are as follows:
_It depends on the correspondence of the tip T included in a.SF _b . For example, when sharing settings, both sharing bags SF _a · S
When the tips T in F _b are determined to be different tips T, the initial value of the sync limit Lim becomes 0, and the dirty bit D corresponding to each tip T is set to the insertion update “Im”. It Further, for example, in the case of associating the tip T with reference to the content of the tip T, the sync limit Lim is the number of the tips T that have been successfully associated, and the dirty bit D corresponding to the tip T is It is stored so that the ranks in both dirty bit lists LD are the same. It should be noted that if the contents of both tips T are the same, the value of the dirty bit D becomes the still maintenance "Xp", and if it is not clear whether they are the same or not, it is set to the still update "Xm".

The model of sharing setting is not limited to the bag fusion type model, but one of the sharing bags S at the time of sharing setting like the bag taking-in model shown in FIG.
F may be newly created, or both information processing devices 2a and 2b may be created as in the new bag creation model shown in FIG.
Alternatively, a new shared bag SF may be created. In these cases as well, a process substantially similar to the process shown in FIG. 16 is performed, and the shared information, the dirty bit list LD and the sync limit Lim, the shared information, etc. are registered in each of them. However, a new sharing bag SF is added to the sharing setting passive side.
When a new item is created, the bag ID of the shared bag SF is returned to the sharing setting activation side when the shared information is registered.

Next, the process of S2 shown in FIG. 13, that is, the process during the asynchronous period will be described in detail. During the asynchronous period, the operation to the tip T, the dirty bit list LD and the sync limit L at the time of the operation are performed.
Based on im, the dirty bit D corresponding to tip T is updated as shown in FIG.

Specifically, when the content of the tip T is instructed to be updated, the tip update processing unit 41 (see FIG. 1) receives the instruction from the user operation unit 22 and gives the instruction in S21 shown in FIG. Accordingly, the content of tip T is updated. Further, in S22, the update parameter is set to "m" for the dirty bit D corresponding to the tip T. The movement parameter is kept as it is.

The tip T is a plurality of shared bag SFs.
, And each shared bag SF may be shared with a plurality of opponents. As a result, as a whole of the information processing device 2, there may be a plurality of dirty bits D corresponding to the tip T. Therefore, all the dirty bits D corresponding to the tip T are searched and each is updated. For searching the corresponding dirty bit D, for example, all the dirty bit lists LD may be referred to, or in order to shorten the search time, the dirty bit D corresponding to each tip T may be searched.
The list may be created in advance, and the search may be performed by referring to the list.

In any case, in the present embodiment, the tip T which belongs to the shared bag SF even once in the asynchronous period.
, The dirty bit D is created, the operation to the tip T is monitored, and the dirty bit D is updated. Therefore, even if the tip T is not present in the shared bag SF at the present time due to, for example, the movement of the tip T, the dirty bit D is surely updated.

On the other hand, a new tip T or a tip T is created.
When the tip T is inserted into the shared bag SF due to the movement of the tag or the like, as shown in FIG. 21, the tip addition processing unit 42 (see FIG. 1) causes the bag management unit 42 to respond to the instruction from the user operation unit 22. 32 to control the tip T
Belong to the shared bag SF (S31). Further, in S32, the tip addition processing unit 42 determines whether or not the dirty bit list LD indicating the shared bag SF includes the dirty bit D corresponding to the tip T.

Here, when the dirty bit D is included, it indicates that the content update of the tip T is already monitored, for example, when the tip T in the shared bag SF is taken in and out. Therefore, the tip addition processing unit 42 sets the movement parameter of the dirty bit D to "I" in S33. On the other hand, when it is not included, for example, when it is not included in the shared bag SF once, when a tip T is newly created, and the like, it is determined whether the content of the tip T is updated during the asynchronous period. Indicates that it cannot be identified. Therefore, in S34, the tip addition processing unit 42, for example, the dirty bit D indicating the tip T with a value of “Im” in the order after the sync limit Lim, such as the end of the dirty bit list LD. To add. When there are a plurality of dirty bit lists LD corresponding to the shared bag SF, the processes of S32 to S34 are performed for all dirty bit lists LD.

Further, when the tip T is discharged from the shared bag SF by, for example, deleting or moving the tip T, as shown in FIG.
3 (see FIG. 1) controls the bag management unit 32 according to an instruction from the user operation unit 22 to remove the tip T from the shared bag SF (S41). Here, in the present embodiment, the dirty bit D is held for the tip T included in the shared bag SF even once in the asynchronous period. Therefore, the tip addition processing unit 42 determines S42.
In, all the dirty
In the bit list LD, the movement parameter of the dirty bit D corresponding to the tip T is changed to "O". The update parameter is kept as it is.

For example, at the end of the synchronization process, FIG.
As shown in FIG. 5, all the tips T in both shared pockets SF _a and SF _b are in a one-to-one correspondence with each other, and the contents of the corresponding tips T are set to be the same. Therefore, as shown in FIG. 24, the sink limit Li
m is the same as the length of the dirty bit list LD (the number of dirty bits D), and both dirty bits
In the list LD, the corresponding dirty bit D _a
D _b is stored so as to have the same rank. Further, the value of each dirty bit D is set to "Xp".

Furthermore, during the asynchronous period, both information processing devices 2
Then, when the respective databases DB are operated independently, the respective processing units 41 to 43 of the respective information processing devices 2 update the respective dirty bit lists LD. As a result, during the asynchronous period, the value of dirty bit D always indicates the state of the corresponding tip T. As a result, when the respective databases DB are independently updated, both dirty bit lists LD do not match at the time immediately before the next synchronization process, as shown in FIG. 25, for example.

However, both dirty bit lists LD
In, the rank indicated by the sync limit Lim and the dirty bits D before that are not deleted even if the corresponding tip T is deleted, and the same order is maintained during the asynchronous period. As a result, the ranks of the dirty bits D indicating the tips T corresponding to each other in both shared pockets SF _a and SF _b such as the dirty bits D _3a and D _3b are kept the same.

The dirty bit D after the sync limit Lim is newly added to the shared bag SF during the asynchronous period.
Shows the tip T added to the tip T. For example, even if the dirty bit D is stored in the same order such as the dirty bit D ₂₂ D ₂₇ , the tip T of each dirty bit D ₂₂ D ₂₇ is stored.
Does not correspond. These sink limit lim
The subsequent dirty bit D indicates “Im” or “Om”, and “Om” indicates that the tip T newly inserted in the shared bag SF does not belong to the shared bag SF at present. It is shown that. When the number of added tips T is different, both dirty bit lists LD
Are also different in length.

In this state, if a user's instruction or a sensor for detecting communication is instructed to perform a synchronization process,
The synchronization process shown in S3 of FIG. 13 is performed. More specifically, as shown in FIG. 26, in S51, the synchronization processing unit 51 of each information processing device 2 registers the dirty bit D after the sync limit Lim in each dirty bit list LD. Of these, the dirty bit D whose movement parameter is "O" is deleted. As a result, as shown in FIG. 27, both dirty bits
In the list LD, the value of the dirty bit D after the sync limit Lim is only “Im”.

The dirty bit D having the movement parameter "O" is not included in the shared tip SF of the corresponding tip T at the time of the previous synchronization processing and the current synchronization processing. Therefore, the content of the tip T does not need to be sent to the partner regardless of the state of the shared bag SF of the partner. As a result, by deleting these tips T, both information processing devices 2
The number of dirty bits D transmitted between a and 2b can be reduced.

When the compression of the dirty bit list LD is completed in the above S51, both synchronization processing units 5
In S52, 1 transmits and receives the dirty bit D after the sync limit Lim and generates the dirty bit D corresponding to the dirty bit D sent from the other party. Specifically, one of the information processing apparatus 2b is a own dirty bit list LD _b, and sends the number Y _b of dirty bits D after the sink limit Lim _b, the other information processing equipment 2a, in their own dirty bit list LD _a, sink limit Lim _a
Immediately after, the dirty bit D with the value "Op" is inserted by the number Y _b received. Furthermore, both information processing devices 2
a-2b increases the respective sink limit Lim only the Y _b. As a result, as shown in FIG. 28, the dirty bit D corresponding to the tip T added by the information processing device 2b is also created in the information processing device 2a. Similarly, the dirty bit D corresponding to the tip T added by the information processing device 2a is also added to the information processing device 2b, and the sync limit Lim is adjusted.

As a result, as shown in FIG. 29, the lengths of both dirty bit lists LD _a and LD _b become equal to each other, and the dirty bits D and D of the same rank are placed.
Indicates the corresponding tip T. Also, both sync limits Lim are dirty bit list L
There is no dirty bit D after the sync limit Lim, which is the same as the length of D. At this stage, the dirty bit list LD and the sync limit Lim may be updated, and the contents of the tip T need not be transmitted.

In the above description, the case of transmitting / receiving the number Y _a (Y _b ) has been described as an example. However, if the dirty bit D of the corresponding tip T can be set in the same order, the information to be transmitted / received is the number Y. _{It is} not limited to _a (Y _b ). Also, in the above, the dirty bit D having a value of “Op” was added,
Similar effects can be obtained if the priority is lower than "m".

When the association of the dirty bit D is completed in the above S52, the two synchronization processing units 51, in S53, based on the dirty bit D transmitted between the two information processing devices 2a and 2b. It is determined whether the tip T corresponding to the dirty bit D is arranged in the shared bag SF. If both dirty bits D _a
When the movement parameters of D _b are different, for example, the priority rule shown in FIG. 30 is used to determine whether or not the tip T _b is arranged in the shared bag SF _b . Note that FIG.
Then, the tip of the arrow has priority.

Specifically, the dirty bit D in the dirty bit list LD is sequentially transmitted from one information processing device 2a to the other information processing device 2b. In the other information processing device 2b, the received dirty bit D
_a and his own dirty bit list LD _b
Compare dirty bit D _{b in the} same order as it was received. The synchronization processing unit 5 of the information processing device 2b
31, the combination of movement parameters is (XX), (XI), (II), (IK) as shown in FIG.
In either case of (1−O) and (I−O), it is determined that the tip T of the dirty bit D is arranged in the shared bag SF, and both tips T corresponding to the dirty bit D are set in the shared bag SF. Belong to. Further, the movement parameters of itself and the other party are set to "X".

On the contrary, the combinations of movement parameters are (X-O), (O-K), (O-O) and (K-).
In either case of K), the synchronization processing unit 51
It is determined that the tip T is arranged outside the shared bag SF, and, for example, the tip T of itself and the opponent is ejected from the shared bag SF and each dirty bit D is set to the invalid state “K”. . If both the information processing devices 2a and 2b are operating normally, the above combination does not become (X-K). Therefore, in this case, it is determined that an error has occurred, and error processing is performed, for example, asking for a user's instruction.

The discharge of the tip T of the opponent and the setting of the dirty bit D are performed by the own dirty bit D.
May be sent to the other party before the change, and may be processed by the other party's synchronization processing unit 51, or the result of its own judgment may be transmitted to the other party. In either case, the dirty bit D and the tip T of the other party are specified in the order in which the dirty bit D or the determination result is sent, and the information for specifying each is not sent in particular. Therefore,
The amount of data transmitted at the time of inside / outside determination is extremely small.
As a result, as shown in FIG. 32, both dirty bits
The movement parameters of all dirty bits D in the list LD are set to "X" or "K".

Note that the tip T whose movement parameter is not "X", that is, the tip T judged to be outside the shared bag SF and the tip T having an error are not subject to the tip synchronization processing of S54 and S55 below. And
The update parameter priority determination and tip T content transmission are not performed.

Subsequently, with respect to the tip T (the tip T having the movement parameter "X") which is determined to be within the shared pocket SF in the balloon synchronization of S53, the synchronization processing unit 51 determines in step S54 that both dirty Bit D
Are compared with each other to determine the priority of each corresponding tip T. The priority of Tip T is shown in FIG.
, The update "m" has priority over the maintenance "p", as determined by the update parameters. As a result, whether or not the contents of the tip T should be transmitted and the direction of the transmission are determined.

Specifically, the synchronization processing section 51
As shown in FIG. 34, both dirty bits D _a D
_b is compared, and the combination of update parameters is (pp)
In the case of, it is determined that the contents of the tip T need not be sent. When the combination is (pm), the side with the update parameter "m" is determined to be the priority side, and the content of the tip T is transferred from the priority side information processing apparatus 2 to the non-priority side information processing apparatus 2. Determine that should be sent. Therefore, when both dirty bit lists LD have the values shown in FIG. 32,
As shown in FIG. 35, the priority side is determined. In S54, when the combination is (m−m), that is, when the tip T is updated by both, the tip T
And the priority side is determined by a predetermined priority rule, as will be described later.

When the priority of the tip T is judged in S54 and it is judged that the contents of the tip T need to be sent, the both synchronization processing units 51 of the tip T on the priority side are judged in S55. The content is sent to the non-priority side, and the content of the tip T on the non-priority side is replaced with the content on the priority side. As a result, since the contents of both tips T match, the update parameter of each dirty bit D is set to "p". As a result, as shown in FIG.
All dirty bits D in the dirty bit list LD are set to either "Xp" or "K".

Further, when the tip synchronization is completed in S54 and S55, both dirty bit lists LD are compressed in S56, and the dirty bit D having the value "K" is used for each dirty bit list LD.
It is deleted from the bit list LD. For example, in FIG. 36, the seven dirty bits D at the fifth, sixth, tenth, eleventh, nineteenth, and twentieth positions are deleted from the respective dirty bit lists LD. As a result, as shown in FIG. 37, all the dirty bits D become “Xp”, and the sync limit Lim is updated according to the number of dirty bits D. In the example of FIG. 37, the order of the dirty bits D having the value “Xp” is not changed, but if the corresponding dirty bits D are the same in both dirty bit lists LD, You may change the order.

By the synchronization processing of S51 to S56, the number of tips T in both shared bag SFs and the contents of the corresponding tips T become the same. As a result, even when the database DB is operated independently on one of the information processing devices 2, if the synchronization processing is performed at the time when communication with the other information processing device 2 becomes possible, each shared bag SF SF. The contents of can be matched. Therefore, even when the database system 1 is configured with the information processing device 2 in which the periods in which communication is possible with each other is intermittent,
Data integrity can be maintained.

Further, in the above synchronization processing, in S55,
Before sending the contents of Tip T, S53 and S54
At both information processing devices 2a and 2b, dirty bit D
Are compared, and whether or not the content needs to be sent and the transmission direction are set to Tip T.
It is decided for each. Therefore, the tip T, which is found to be unnecessary to send the content at this point, can omit the transmission of the content. Here, in this embodiment, the update of the content of the tip T having the dirty bit D is monitored.
Therefore, not only the tip T that has not been moved or updated, but is present in the shared bag SF at both the previous synchronous processing time point and the current synchronous processing time point, and is updated during the asynchronous period. For the tip T that does not exist, it is possible to omit sending the content.

Further, in this embodiment, since the correspondence relation of the tip T is stored as the rank of the dirty bit D, it is possible to store the ID indicating the tip T of the opponent.
Can store correspondence. In addition, since the corresponding tip T can be specified without sending the ID, the amount of data transmitted between the information processing devices 2a and 2b and the dirty bit when the dirty bit D is sent and received. The amount of memory required to store D can be reduced.

In the present embodiment, the dirty bit D is created for the tip T included in the shared bag SF even once in the asynchronous period, but the present invention is not limited to this. At the end of the previous synchronization process, the tip T included in the shared bag SF is
If the bit D is created, it is possible to omit the content transmission of the tip T, which has only been moved, as in the present embodiment. However, like the present embodiment, when the dirty bit D is also created for the tip T currently included in the shared bag SF, the speed at which the tip T newly inserted during the asynchronous period is searched for is high. Improve and
By extracting only the tips T of which the movement parameter of the dirty bit D is “X” or “I”, the list of the tips T included in the shared bag SF in the tree structure S shown in FIG. 6 can be created. As a result, the tip T
The list of can be deleted and the amount of memory can be reduced.

Note that, for example, each processing such as addition of the tip T may be performed by the synchronization processing section 51 of any of the information processing devices 2. For example, one of the synchronization processing units 51 may perform all the processing, or both of them may share the processing.

In the following, the operation in the case where the collision of the tip T is detected in S54 will be described in more detail. That is, in the present embodiment, when a collision is detected, any one of the following three models is adopted as a main body for resolving the collision. Although one model may be adopted, the information processing device 2 according to the present embodiment
Uses the model designated as an option during the sharing setting shown in S1 of FIG. 13 among the plurality of usable models.

The first model (discussion model) is shown in FIG.
As shown in FIG. 8, the sinks of both information processing devices 2a and 2b
The manager 61 manages each upper application 6
The information processing device 2a and 2b independently determines the priority side. In the second model (selection solution model), as shown in FIG. 39, only the sink manager 61 of both information processing devices 2a and 2b inquires the higher-level application 62 to determine the priority side. To do. Furthermore, the third model (automatic solution model)
As shown in FIG. 40, both sync managers 61 select the priority side based on a predetermined priority rule without inquiring the upper application 62. The sync manager 61 is a part of the synchronization processing unit 51 shown in FIG. The upper application 62 is an application that refers to data in the database, such as an application that analyzes data in the database, or an application that inquires the user of the information processing device 2 about the priority side. Then, the priority side can reply.

In the selection solution model, the method of selecting one of the information processing devices 2 is the same as the one-sided selection model of the automatic solution model to be described later, such as the presence or absence of the higher-level application 62, the start side of the sharing setting, and the like. Can be adopted. However, like the model showing the method of determining the priority side, if the selected information processing device 2 is different, there is a possibility that the undesired one may be selected.
The method of selecting is preferably set when sharing is set.

The discussion model further includes a negotiation resolution type model for resolving a collision during synchronization processing in which a collision is detected, and a collision detection and collision resolution communication once according to the timing of resolving the collision. It is divided into an application-solving model that disconnects. Similarly, the selection solution model is divided into a client immediate model and a client delay model. As shown in FIG. 41, the negotiation resolution type model and the client immediate type model are
It is classified into an immediate solution model that is in the same synchronization process as the detection, and the application resolution model and the client delay model are classified into a delay solution model in which the resolution time is in the synchronization process after the collision is detected.

Here, in the case where the negotiation solution type model is adopted among the immediate solution models, for example, as shown in FIG. 42, in S61, the sink manager 61 of one of the information processing devices 2a has its tip The information necessary for determining the priority side among the contents of the above and the information necessary for determining the priority side among the contents of the tip T of the other party are presented to the upper application 62. Similarly, in S62, the information indicating the tip T of the user and the tip T of the partner is presented to the higher-level application 62 in the information processing device 2b of the partner. Further, in S63 and S64, the upper application 62 of each of the information processing devices 2a and 2b independently determines the priority side.

When the priority sides are determined by both sides, the sync manager 61 determines in S65 whether or not the priority sides are the same. If both match (S6 above)
If YES in 5), it is determined in S66 whether or not the priority side is the own device. In the case of the own device, the sink manager 61 sends the contents of its own tip T to the partner information processing device 2 in S67,
Replace the contents of Tip T of the opponent. On the other hand, if it is not the own device, in S68, the content of the tip T of the partner is received from the information processing device 2 of the partner,
Replace the contents of your own tip T. As a result, the content of the tip T sets that have collided becomes the same.

In S65, if the priority sides of both do not match, the sink manager 61 of the information processing device 2a, in S69, notifies the host application 62 of its own in addition to the contents presented in S61. Furthermore, the priority side decided by the other party is presented. Similarly, in S70, the sink manager 61 of the information processing device 2b presents the presentation content of S62 and the priority side determined by the information processing device 2a to its own upper application 62. After that, the processes from S63 onward are repeated, and the priority side is reconsidered by each upper application 62.

In the case where the client immediate model is adopted, for example, as shown in FIG. 43, in one information processing device 2, as in S61 of FIG. Is displayed on the upper application 62 (S81). Also, S82
At, the upper application 62 determines the priority side based on these pieces of information. Further, when the priority side is the own device (YES in S83), the contents of the tip T of the partner is replaced with the contents of the tip T of the partner in S84, and the partner device is prioritized (No in S83).
In the case of), in S85, the content of the tip T of the opponent is replaced with the content of the tip T of the opponent. As a result, the content of the tip T sets that have collided becomes the same.

On the other hand, when the application resolution type model is adopted among the delay resolution type models, for example, as shown in FIG. 44, in S91, it is determined whether or not the contents of both tips T match. The case where both match is the case where the collision of the tip T is detected in the previous synchronization process, and the solutions independently proposed by both match. In this case, in S95, the sync manager 61 of each of the information processing devices 2a and 2b sets the update parameter of its own dirty bit D corresponding to the tip T to the maintenance "p" and completes the processing. .

On the other hand, if the contents of both tips T do not match (No in S91), each sync manager 61 exchanges the contents of tip T with each other in S92, and then S93. Then, for example, by disconnecting the communication or determining the collision of another tip T, the synchronization processing regarding the tip T is interrupted once. Further, while the communication is disconnected, the sync manager 61 of both the information processing devices 2a and 2b responds to the instruction of the respective upper application 62 in S94, independently of the sync manager 61 of the other party, to perform the conflict resolution process. Then, the process is completed.

In the following, based on FIG. 45, the above S94
The process will be described in more detail. That is, the sync manager 61, in S111, in S91,
The content of the tip T received from the other party and the content of the tip T of its own are presented to the host application 62 of its own, and an instruction of which is prioritized is received from the host application 62.

When the upper application 62 determines that the content of the tip T of the partner device is prioritized (S1 above).
If YES in 12), the sync manager 61
In S113, the content of the tip T of the user is replaced with the content of the tip T of the other party, and in S114, the update parameter of the dirty bit D corresponding to the tip T is set to the update "m", and the processing is executed. Complete. On the contrary, when it is determined that the content of the tip T of the own device is prioritized (No in S112 above), the sync manager 6
In step S114, the update parameter of the dirty bit D corresponding to the tip T is set to update "m" without performing the process of step S113.

Here, when both the information processing devices 2a and 2b perform the processes of S111 to S114, the update parameter of the dirty bit D is updated in S114 regardless of which is determined to be the priority side. Since it is set to "m", it is once determined to be a collision in the next synchronization processing. However, if both priority sides match, the contents of both tips T match, so the determination in S91 shown in FIG. 44 is YES.

For example, when the information processing apparatus 2a determines that the partner apparatus is the priority side and the information processing apparatus 2b determines that the own apparatus is the priority side, the information processing apparatus 2a is Since the judgment in S112 is YES,
The content of the tip T of the information processing device 2a is replaced with the content of the tip T of the information processing device 2b. Further, in the information processing device 2b, since the determination in S112 is No, the content of the tip T of the information processing device 2b is not changed.

As described above, when both priority sides match, the contents of both tips T match at the time of the next synchronization processing, and the determination in S91 is YES. As a result,
The processing of S95 is performed in both the information processing devices 2a and 2b, and the update parameter of the dirty bit D corresponding to the tip T becomes the maintenance "p", and the collision is completely resolved. .

If the two priority sides do not match, the contents of both tips T do not match at the time of the next synchronization processing. Therefore, in the next synchronization process, the processes after S92 are repeated again, and the respective upper applications 62 are requested to resolve the conflict again.

In the case of the client delay type model,
The information processing device 2 that makes an inquiry to the higher-level application 62 performs the process shown in FIG. That is, S1
At 21, the sync manager 61 transfers and receives the contents of the tip T of the other party to itself. Furthermore, S1
At 22, the update parameter of the dirty bit D of the opponent corresponding to the tip T is changed to the maintenance “p”.

Further, in S123 and S124, similar to S93 and S111 shown in FIG. 45,
The priority side is determined after the interruption of the synchronization process. If it is determined that the partner device is the priority side (YES in S125), the sync manager 61 determines in S126 that the above S1.
The contents of the tip T of the partner device received at 21 are replaced with the contents of the tip T of the own device. Furthermore, S127
Then, the update parameter of its own dirty bit D corresponding to the tip T is set to the maintenance "p" and the processing is completed.

On the contrary, when the own device is determined to be the priority side (No in S125), the sync manager 61 determines that the dirty manager corresponding to the tip T concerned.
The update parameter of bit D is set to update "m" (S1
28). Here, in S122, the update parameter of the dirty bit D corresponding to the tip T is set to the maintenance “p” in the partner device. Therefore, when the own device is determined to be the priority side (the update parameter is updated "m")
In this case), the content of the tip T of the partner device is replaced with the content of the tip T of the own device in the next synchronization process.

When the above delay solution model is adopted,
After a tip T collision is detected in a certain synchronization process,
Since the contents of the tip T are matched in the synchronization process from the next time onward, it is not necessary to determine the priority side during a single synchronization process. As a result, for example, the upper application 6
Even when the sync manager 61 takes a long time from the inquiry to the priority side until the answer is obtained, as in the case where 2 receives the instruction from the user, the total communication time is set to a fixed time. Therefore, the communication cost and power consumption can be reduced. In the above description, the case where all the contents of both tips T are presented has been described as an example, but if the priority side can be determined, the information to be presented is the occurrence of a collision,
The same effect can be obtained even if the content of the tip T is part.

On the other hand, the automatic solution type model is a sequential decision model for deciding the priority side for each set of the tips T that have collided,
At the start of the synchronization processing, one information processing device 2 can be roughly divided into a one-side selection model in which the information processing device 2 is set as the priority side in advance. An example of the former is a method of determining the priority side based on a part of the content of the tip T, such as the last update time. In this case, for example, as shown in FIG.
31, the sync manager 61 shown in FIG.
Acquires the last update time t1 of the tip T to be processed in its own information processing device 2. In addition, S13
In 2, the last update time t2 of the tip T of the opponent corresponding to the tip T is acquired. Further, in S133, the last update time t1 of the user is the last update time t2 of the other party.
Is newer than. Own last update time t
If 1 is newer, the contents of the tip T of the opponent are replaced with the contents of the tip T of the opponent in S134. On the contrary, if the last time t2 of the other party is newer,
In S135, the content of the tip T of the opponent is replaced with the content of the tip T of the opponent. As a result, the tip T updated later can be prioritized. For example, in FIG.
In the example shown in FIG. 8, the last update time of the tip T _8a of the information processing device 2a is newer than the tip T _8b of the information processing device 2b. Therefore, the contents of tip T _8a replace the contents of tip T _8b . The criterion for determining the priority side is not limited to the last update time, but the tip T
The same effect can be obtained if it is a part of the contents of.

Further, in the one-selection model among the automatic solution type models, for example, as shown in FIG. 49, in S141, it is determined whether or not the priority side determined by the start of the synchronization process is the own device. To be done. If the own device is prioritized, the content of the tip T of the other party is replaced with the content of the tip T of the other party in S142, and if the other device is prioritized, the content of the tip T of the other party is used in S143. The contents of tip T of are replaced. Thereby, the tip T on the predetermined side can be prioritized.

In this embodiment, the priority side referred to in S141 is determined by one of the following three models. The first model is a model in which the priority side is set in advance during the synchronization setting shown in S1 of FIG. 13, is stored as priority side information, and is determined based on the priority side information. Any method can be used for the priority side.
For example, it may be set according to an instruction from the user or the upper application 62.

The second model is a model of the first model that determines the priority side based on the side that has activated the sharing setting. It is divided into the passive side priority model. The sharing setting starting side priority model is preferably used, for example, when the database system 1 including the information processing devices 2 ... Always wants to give priority to the client / server type client side, and the starting side of the synchronization process and the contents of the tip T. Regardless of the above, the sharing setting activation side is always prioritized. On the other hand, the shared setting passive side model is preferably used, for example, when the server side of the database system 1 is always desired to be prioritized, and the shared setting passive side model is always prioritized. In either case, since the activation of the sharing setting and the specification of the priority side can be designated by the same operation, the operation can be simplified as compared with the case where the respective operations are designated by different operations.

Further, the third model is a model for deciding the priority side on the basis of the side activating the synchronization processing in S3 of FIG. 13, and includes the synchronization processing activation side priority model and the synchronization processing passive side priority model. Can be divided into These models are used, for example, when the information processing devices 2 are equal and the former wants to prioritize its own device, and the latter when it wants to prioritize the other device. In either case, since the activation of the synchronization processing and the identification of the priority side can be designated by the same operation, the operation can be simplified as compared with the case where the respective operations are designated by different operations. Furthermore, the priority side can be changed depending on which information processing device 2 starts the synchronization process at the time of each synchronization process.

By the way, in the above, when the tip T is changed by only one information processing device 2, the change is reflected in the other information processing device 2, but the present invention is not limited to this. For example, as shown in FIG. 50, the information processing device 2
In accordance with the contents of the shared bag SF _a on the priority side (a), the shared bag S on the information processing device 2b on the non-priority side is shared.
The contents of F _b may be updated. In this case, the synchronization process is
It becomes a download (upload). If the information processing device 2 can select whether or not to download,
The selection is performed in the same manner as the model of the priority side determination in S of FIG.
It is better to specify it as an option at the time of sharing setting shown in 1.

Specifically, as shown in FIG. 51, S15
In 1, the dirty bit list LD on the priority side
Is compressed similarly to S51 shown in FIG. further,
In S152, the tip T added during the asynchronous period is deleted from the shared bag SF on the non-priority side. Also,
In S153, the dirty bit D corresponding to the tip T added on the priority side is added to the dirty bit list LD on the non-priority side, as in S52.

In S154, the synchronization processing unit 51 shown in FIG. 1 makes an internal / external decision based on the dirty bit D on the priority side, and sets the movement parameter of the dirty bit D to "X" in both dirty bit lists LD. Or “K”, and the tip T is moved according to the inside / outside determination. That is, when the movement parameter of the dirty bit D on the priority side is "X" or "I", the inside of the shared bag SF is shown. Therefore, the tip T is left in the shared bag SF and the movement parameter of both dirty bits D corresponding to the tip T is set to “X”. Also, the dirty bit D on the priority side
If the movement parameter of "K" is "K" or "O", it means that the tip T is outside the shared hat SF and the dirty bit D corresponding to the tip T is moved. The movement parameter of is set to "K".

Further, in S155, the contents of the tip T on the non-priority side are replaced with the contents of the tip T on the priority side based on the update parameters of both dirty bits D. Specifically, the update parameter on the priority side is updated "m".
In the case of, the content of the tip T is sent to the non-priority side. Further, even when the update parameter on the priority side is the maintenance “p”, if the update parameter on the non-priority side is the update “m”, the content of the tip T is similarly replaced on the non-priority side. If both update parameters are maintained “p”, the contents of tip T are not transferred.

When the content synchronization of the tip T is completed in S155, both dirty bit lists LD are compressed in S156, as in S56 of FIG. As a result, by the time when the synchronization process is started, the contents of the non-priority side shared bag SF can be updated in accordance with the predetermined contents of the priority side shared bag SF.

Further, as one of the options that can be set by the sharing setting processing section 52, as shown in FIG. 52, the tips T that have collided are regarded as different tips T, and their own tips T are shared by the other party. A model added to the bag SF can also be adopted. As with the necessity of downloading, whether or not to adopt the model is also specified as an option at the time of sharing setting.

[0136] Specifically, as shown in S161 of FIG. 53, when the tip T _{4a-T 4b} collide in dirty bit list LD _b partner, its dirty
The dirty bit D in the bit list LD _a
A dirty bit D _4ab having a lower priority than its own dirty bit D _4a , such as “Op”, is newly created in the same order as that of _4a . Similarly, in S162, in its own dirty bit list LD _a ,
Create a new dirty bit D _4ba in the same rank as the opponent's dirty bit D _4b . Further, based on these dirty bits D, in S163, the tip T _4a
The contents of is transmitted to the information processing apparatus 2b, a new tip T _4ab corresponding to the tips T _4a is added to the shared bag SF _b. Further, in S164, a new tip T _4ba corresponding to the tip T _4b is added to the shared bag SF _a . Thereby, when the tip T collides, all the changed contents of the tip T can be saved.

Note that, in FIGS. 52 and 53, an example in which copying is performed only when a collision of the tip T occurs has been described. However, as another option, as shown in FIG. 54, at least one of the corresponding tips T is set. A model that copies when it is changed can also be adopted. In this case, the set of tips T that has not been changed is not changed at all, but the tip T is copied when at least one of the sets of tips T is changed. As a result, a copy is also created for the tips T _2a , T _2b and T _3a , T _3b , only one of which has been changed, and the entire contents of the changed tip T can be saved.

The processing during the asynchronous period shown in S2 of FIG. 13 and the synchronous processing shown in S3 are alternately repeated.
After that, when the synchronization process becomes unnecessary, in S4,
Sharing settings are canceled. As an example of cancellation of sharing, a model in which the bag F and the tip T remain on both information processing devices 2a and 2b (tip copying type model) will be described. As shown in FIG. 55, the model is provided on one of the information processing devices 2a. The sharing cancellation processing unit 53, in S171,
The sharing cancellation request is sent to the sharing cancellation processing unit 53 of the other information processing device 2b. The sharing cancellation request includes, for example, information indicating the shared pocket SF of the partner such as a combination of the own pocket ID and the ID of the partner, and information indicating the shared pocket SF of the partner corresponding to the shared balloon SF. Has been.

Further, in S172, the sharing cancellation processing unit 53 on the sharing cancellation starting side, for example, the dirty bit list LD and the sync limit Lim corresponding to the above combination, and the partner registered in the sharing partner list LS. The shared information provided at the time of sharing setting, such as the bag ID, is deleted. On the other hand, in the information processing device 2b on the sharing cancellation passive side, the sharing cancellation processing unit 53 deletes the sharing information stored therein based on the received sharing cancellation request (S173, S174). In addition, in FIG. 55,
Although the authentication of the sharing partner is not shown, the sharing cancellation passive side may determine whether or not to cancel the sharing in accordance with the content of the sharing cancellation request, similarly to the sharing setting.

As a result, as shown in FIG. 56, the shared bag SF shared between the two information processing devices 2a and 2b.
_{The a} · SF _b is the non-shared bag F _a · F _b . In the sharing cancellation of the tip copy type model, when the sharing cancellation is instructed, the sharing setting between both sharing bags SF _a and SF _b is only canceled, and the tip T included in each of them is set to the both sides. Remains on the seams F _a and F _b .

The sharing cancellation model is not limited to the tip copy type model, and various models can be applied. For example, when the sharing setting is canceled, as shown in FIG. 57, a model in which a tip is left only in one of the bags F and a tip is deleted in the other bag F (a model in which the tip is unevenly distributed) is adopted. As shown in FIG. 58, the tip extinction type model, that is, both of the bag F
A model in which the tip T is deleted may be adopted. Further, in the case of the tip copy type model or the tip uneven distribution type model, when sharing cancellation is instructed, it is possible to perform sharing processing and then cancel sharing. In this case, even if the synchronization process is not instructed, the tip T
The integrity of can be maintained. Further, in the case of removing the tip T as in the tip uneven distribution model and the tip disappearance model, as shown in FIG. 59 and FIG. Good. In any case, the information processing device 2
If a plurality of sharing cancellation models can be selected, it is better to set the sharing cancellation model as an option at the time of sharing setting shown in S1 of FIG. 13, as in the model selection for determining the priority side.

By the way, in the above, S51, that is,
During the synchronization processing shown in S3 of FIG. 13, the dirty bit D after the sync limit Lim was compressed, but this processing only refers to the dirty bit D stored in its own information processing device 2. Therefore, it can be performed during the asynchronous period of S2.

In this case, the tip deletion processing section 43 (see FIG. 1) is operated as shown in FIG.
Instead of 1, the process shown in FIG. 61 is performed. That is, S1
At 81, it is determined whether or not the dirty bit D corresponding to the tip T is after the sync limit Lim. If it is, then at S182, the dirty bit D is deleted from the dirty bit list LD. It
On the other hand, if it is not after, the movement parameter of the dirty bit D is changed to "O" in S183. As a result, the processing of S51 is not required during the synchronization processing, and the required time can be shortened.

When compression is performed during the asynchronous period, the tip T having a very short period existing in the shared bag SF, for example, when the tip T is inserted by an erroneous operation, is used.
Dirty Bit D associated with
It can be deleted from the list LD immediately. As a result, the length of the dirty bit list LD can be suppressed and the amount of memory required for storage can be reduced.

In FIG. 61, the case where the dirty bit list LD is compressed when the tip T is discharged has been described, but the dirty bit list LD may be compressed at any time during the asynchronous period. In this case, for example, due to an instruction from a higher-level application or the like, a dirty file that is located after the sync limit Lim and has a movement parameter of “O” is ejected.
Bit D is deleted from the dirty bit list LD. As a result, the dirty bits D of a plurality of tips T can be deleted from the dirty bit list LD by one processing.

Second Embodiment By the way, in the first embodiment, each information processing device 2 ...
The description has been given by taking as an example the case in which it is possible to create a bag F that is a subset of, and each is equal. On the other hand, in the present embodiment, as shown in FIG.
A case where 1 includes an information processing device (database management device) 12 that cannot set a subset will be described.

The database system 11 is preferably used when it is composed of information processing devices that are not equal to each other, such as a client-server type database system. The information processing device 2 is realized, for example, as a desktop computer that operates as a server, and the information processing device 12 has a database DB that can be stored in a larger scale than the information processing device 2, such as a mobile information terminal. Realized with a small device.

The information processing device 12 has substantially the same structure as the information processing device 2 shown in FIG. 1, but the database management unit 23, as shown in FIG.
B is stored as one bag F, and as shown in FIG. 62, the bag F, which is the entire database DB of the information processing device 12, is shared with the bag F of the information processing device 2.

Even in the database system 11 having the above-described configuration, the dirty bit D is compared between the information processing devices 2 and 12 before the content transmission of the tip T during the synchronization processing, and it is determined whether the content transmission is necessary or not. The transmission direction is determined. Here, in the information processing device 2, as in the first embodiment,
Based on the dirty bit D, the tip T that has only been moved can be determined. As a result, the content transmission of the tip T can be reliably suppressed, and the amount of data transmitted during the synchronization processing can be reduced.

Further, also in this embodiment, since the correspondence relation of the tips T is stored in the order of the dirty bits D in both the dirty bit lists LD, the state of the tips T can be grasped with a smaller memory amount. At the same time, it is possible to reduce the amount of data communicated during the synchronization processing.

As an example of using the database system 11 having the above configuration, for example, the information processing device 2 installed in a company or the like and serving as a server collectively manages the address information required by all the sales persons, and becomes a portable information terminal. Information processing equipment 12
Can be given to each salesman.
Here, the combination of address information required for each sales person is different from each other, and there are many cases in which a plurality of sales persons have common address information. Therefore, in the information processing device 2, by storing all the address information in the database DB and managing the combination of the address information required for each sales person as the bag F, the information processing device 2 can
For example, unlike the case where the address information for each sales person is stored in another database DB, each address information can be managed without impairing the consistency. Also, the information processing device 1
2 only needs to be able to store one bag F of the information processing device 2, so that the storage capacity can be reduced compared to the case where the entire address information is stored.

In the first and second embodiments described above, the case where the tip T indicates an individual address has been described as an example, but the data indicated by the tip T is naturally free depending on the use of the database. Can be set to. Further, the case has been described as an example where the tips T have the same type of attribute A ... And the tips T having the same sharing partner are collected in the shared bag SF, but the present invention is not limited to this. For example, the shared bag SF may be provided for each combination of the sharing partner and the application. Taking the case where there are applications of address book, electronic book and mail address database as an example, the tip T used in the address book has attributes indicating a name, a telephone number, a mail address, an address and a place of work. ing. The tip T used in the telephone directory has attributes indicating a name and a telephone number, and the tip T of the mail address database is
Has attributes indicating a name and a mail address.
Further, each tip T is collected in the bag F indicating the application used. For example, the tip T used in the telephone book also belongs to the bag F, which indicates the address book.
In this case, if you want to update the entire address book, you can set sharing for the bag F that shows the address book.
When it is desired to change only the bag F of the phone book, the sharing may be set for the bag F and synchronization processing may be performed. In this way, if the bag F is assigned to each application to be used, synchronous processing can be performed for each application.

In the first and second embodiments, the correspondence is stored as the order of dirty bits D, but the invention is not limited to this. For example, the corresponding tip T may be specified by information indicating the tip T of the other party such as the tip ID. However, in this case, it is necessary to send not only the dirty bit D but also the tip ID of the other party at the time of synchronization processing, which may increase the amount of data to be communicated.

Furthermore, the data amount of the tip ID of the opponent increases as the scale of the database DB of the opponent increases. As a result, in particular, as in the second embodiment,
If the information processing device 2 and the information processing device 12 are not compatible, the information processing device 12 is strongly requested to reduce the memory amount.
In the above, the amount of memory required to store the correspondence becomes larger than that specified in the order of the dirty bit D.

On the other hand, in the first and second embodiments, the correspondence relationship between the tips T is that the tips T themselves.
Are stored in the order of information related to. Therefore, the correspondence can be stored regardless of the scale of the partner database DB and the storage method of the tip T. As a result, it is possible to reduce the amount of memory required to store the correspondence relationship and to reduce the amount of data transmitted during the synchronization processing, as compared with the case where the information indicating the tip T of the other party is specified.

In the first and second embodiments, the order of the dirty bits D is the dirty bit.
Although stored as a list LD, it is not limited to this. For example, a list indicating the order may be provided separately from the dirty bit list LD, and the dirty bit D and information indicating the order may be stored for each tip T. However, even if the dirty bit D and the order correspond to the same tip T, the dirty bit D and the order are provided for each combination of the own shared bag SF and the partner shared bag SF, and the number of shared partners is the database DB. Often unpredictable at the time of creation. Therefore, it is better to provide the dirty bit D and the order separately from the content itself of the tip T. Therefore, if each of them is stored separately, information indicating the corresponding tip T is needed separately, and the required memory amount increases. Therefore, by storing the correspondence relationship as the order of storing the dirty bit D, the required memory amount can be further reduced.

Further, in each of the above-mentioned embodiments, the dirty bit D is provided for each combination of the own shared bag SF and the shared bag SF of the other party, but if the asynchronous periods are the same, a plurality of shared bag SFs are provided. Dirty between SF
The bit list LD (dirty bit D) can be shared.

In each of the above embodiments, for example, the case where each information processing device 2 (12) is provided with the synchronization processing section 51 has been described. However, each information such as the synchronization processing section 51 and the sharing setting processing section 52 is described. Processing equipment 2
The member that operates only during the period (12) in which communication is possible may be provided in either one. Further, for example, in the synchronization processing unit 51, a portion that detects a collision of the tip T and a portion that determines the priority side may be arranged in different information processing devices 2 (12). If the above members can be formed at the time when communication is possible, the same effect can be obtained.

[0159]

As described above, the database management apparatus according to the first aspect of the present invention stores the status information indicating the data change during the asynchronous period for each of its own data corresponding to the data in the other party's database. State information storage means and a collision detection means for detecting whether or not there is a plurality of changed data in the data set, based on the respective state information, for a set of data corresponding to each other during synchronization processing. In the case of a plurality of data, before transmitting all the contents of the changed plurality of data, it is determined which data has priority, and the synchronization processing means determines the contents of the data on the priority side of the non-priority side. The priority determining means that replaces the contents of the data and the priority determining means determined by its own priority determining means
Detection means to detect inconsistency with the priority side specified by hand
A configuration that includes and.

In the above configuration, the data collision such as the status information can be detected with a smaller amount of data than the content of the data, and the prioritized data is determined before transmitting the entire content of the data, and The contents replace the contents of the non-priority side data. Therefore, even if the data collide, it is possible to prevent the scale expansion of each database. As a result, it is possible to reduce the amount of memory required to store the database and the amount of data transmitted and processed during the synchronization process.

Further, the detecting means detects the disagreement on the priority side.
Since it detects, you can prevent the loss of undesired data contents
It Therefore, for example, priority is given to each database management device.
To encourage the re-determination of the
It has the effect of enabling appropriate processing such as existing
It

As described above, in the database management apparatus according to the second aspect of the present invention, in the configuration of the first aspect of the present invention, the priority side determining means is one of the plurality of changed data contents. In this configuration, only a part of the above is compared to determine the priority side.

Therefore, it is possible to reduce the amount of data to be transmitted and processed at the time of synchronous processing, as compared with the case of transmitting the entire contents of data.

As described above, in the database management apparatus according to the third aspect of the present invention, in the configuration of the first aspect of the invention, the priority-side determining means determines which data is changed when there are a plurality of changed data. User or above
It is a structure to inquire the application of the highest rank .

Therefore, it is possible to reduce the amount of data transmitted and processed at the time of synchronization processing and to determine the priority side more accurately.

As described above, in the database management apparatus according to the fourth aspect of the present invention , in the configuration of the third aspect of the invention, the priority-side determining means determines which data is changed when there are a plurality of changed data. User or higher priority
The application is inquired , the priority side is determined based on the answer, and the priority side is instructed to the synchronization processing means based on the priority side and the priority side designated by the partner, and The communication with the other party is once disconnected between the inquiry and the instruction.

According to the above configuration, the communication is disconnected between the inquiry and the instruction on the priority side. Therefore, even if it takes a long time for the inquiry to be answered,
The total communication time required for synchronization processing can be kept constant. As a result, it is possible to reduce the cost and power consumption required for communication.

As described above, the recording medium according to the fifth aspect of the present invention has a program for operating a computer as the state information storage means, the collision detection means , the priority side determination means and the detection means according to the first aspect. This is the recorded configuration.

Therefore, by causing the computer to execute the above-mentioned program, it is possible to reduce the amount of memory required for storing the database and the amount of data transmitted and processed during the synchronization processing, as in the first aspect. Produce an effect.

[Brief description of drawings]

FIG. 1 shows an embodiment of the present invention and is a block diagram showing a main configuration of an information processing device.

FIG. 2 is a block diagram showing an example of a database system constructed using the information processing device.

FIG. 3 is an explanatory diagram showing a relationship between a tip indicating one piece of data and a balloon that is a population thereof in the information processing device.

FIG. 4 is an explanatory diagram showing a method of storing tips in the information processing device.

FIG. 5 is an explanatory diagram showing a relationship among a tip, a bag, and a database in the information processing device.

FIG. 6 is a diagram illustrating a method of storing tips in the information processing device, and is an explanatory diagram illustrating a tree structure including tips, bags, and a database.

FIG. 7 is an explanatory diagram showing a relationship between a shared bag that is a shared range of an information processing device and a shared bag of another information processing device in the database system.

FIG. 8 is an explanatory diagram showing a tip correspondence and a correspondence storage method in the database system.

FIG. 9 is an explanatory diagram showing a dirty bit list stored in each information processing device in the database system.

FIG. 10 is an explanatory diagram showing an example of storage of movement parameters in dirty bits forming the dirty bit list.

FIG. 11 is an explanatory diagram showing a correspondence relationship of tips when a shared bag of its own is shared with a plurality of shared bags.

FIG. 12 shows the dirty bit
It is explanatory drawing which shows a list.

FIG. 13 is a flowchart showing a schematic operation of the database system.

FIG. 14 is an explanatory diagram showing sharing settings of a balloon fusion type model in the database system.

FIG. 15 is an explanatory diagram showing a synchronization process in the database system.

FIG. 16 is a flowchart illustrating in detail the operation at the time of sharing setting in the database system.

FIG. 17 is a diagram showing a modified example of the database system, and is an explanatory diagram showing sharing settings of the bag capture type model.

FIG. 18 is a diagram showing another modification of the database system, and is an explanatory diagram showing sharing settings for a new balloon model.

FIG. 19 is an explanatory diagram showing changes in dirty bits according to various operations during an asynchronous period in the database system.

FIG. 20 is a flow chart for explaining the details of the operation during the asynchronous period of the database system, and showing a tip content updating process.

FIG. 21 is a flowchart for explaining the operation of the above database system during an asynchronous period in detail and showing a tip insertion process.

FIG. 22 is a flowchart for explaining the operation during the asynchronous period of the database system in detail, and is a flowchart showing a tip discharging process.

FIG. 23 is an explanatory diagram showing a tip correspondence relationship immediately after the end of the synchronization processing in the database system.

FIG. 24 is an explanatory diagram showing a dirty bit list immediately after the end of the synchronization processing in the database system.

FIG. 25 is an explanatory diagram showing a dirty bit list immediately before the next synchronization process in the database system.

FIG. 26 is a flowchart illustrating in detail the operation during synchronization processing in the database system.

FIG. 27 is a diagram for explaining a state during synchronization processing in the above database system.
It is explanatory drawing which shows a bit list.

FIG. 28 is a diagram for explaining another state during synchronization processing in the database system, in which the dirty bit added by another information processing device is its own dirty bit.
It is explanatory drawing which shows the state added to the bit list.

FIG. 29 is a view for explaining still another state during synchronization processing in the database system, in which the dirty bit added by itself is added to the dirty bit list of another information processing device. FIG.

[Fig. 30] Fig. 30 is a diagram for explaining the synchronization processing of the database system, and is an explanatory diagram showing the priority of the dirty bit movement parameter.

FIG. 31 is a diagram for explaining the synchronization processing of the database system, and is an explanatory diagram showing changes in dirty bits.

FIG. 32 is an explanatory diagram for explaining the synchronization processing of the database system, showing the dirty bit list at the end of the inside / outside determination synchronization processing.

FIG. 33 is a diagram for explaining the synchronization processing of the database system, and is an explanatory diagram showing the priority of the update parameter of the dirty bit.

[Fig. 34] Fig. 34 is a diagram for explaining the synchronization processing of the database system, and is an explanatory diagram illustrating changes in the update parameter of the dirty bit.

FIG. 35 is an explanatory diagram for explaining the synchronization processing of the database system, showing the priority side determination by the dirty bit.

FIG. 36 is an explanatory diagram for explaining the synchronization processing of the database system, showing the dirty bit list after the data content synchronization processing.

FIG. 37 is an explanatory diagram for explaining the synchronization processing of the database system, showing the dirty bit list further compressed after the data content synchronization processing.

FIG. 38 is an explanatory diagram illustrating a discussion model for explaining a subject that resolves a tip collision in the database system.

[Fig. 39] Fig. 39 is a diagram illustrating a subject that resolves a tip conflict in the database system, and is an explanatory diagram illustrating a selection resolution model.

FIG. 40 is an explanatory diagram illustrating an automatic resolution model for explaining a subject that resolves a tip conflict in the database system.

FIG. 41 is an explanatory diagram showing a relationship between a tip collision resolution subject and a resolution time in the database system.

FIG. 42 is a flow chart showing an operation at the time of tip conflict resolution in the database system and showing a negotiation resolution model.

FIG. 43 is a flow chart showing an operation when resolving a tip conflict in the database system and showing a client immediate model.

FIG. 44 is a flowchart showing an operation at the time of tip conflict resolution in the database system, and is a flowchart showing an application resolution model.

FIG. 45 is a flowchart illustrating FIG. 44 in more detail.

FIG. 46 is a flowchart showing an operation at the time of tip conflict resolution in the database system, and is a flowchart showing a client delay type model.

FIG. 47 is a flowchart showing an operation at the time of tip conflict resolution in the database system, and is a flowchart showing an automatic resolution type model based on the last update time.

FIG. 48 is an explanatory diagram showing an update status of each tip in the database system.

FIG. 49 is a flowchart showing an operation at the time of tip conflict resolution in the database system, showing a one-side selection model of the automatic resolution model.

FIG. 50 is an explanatory diagram showing downloading in the database system.

FIG. 51 is a flowchart showing an operation at the time of download in the database system.

FIG. 52 is an explanatory diagram showing a case where the association is changed using a copy at the time of a tip collision in the database system.

FIG. 53 is a flowchart showing an operation at the time of changing association using copying in the database system.

FIG. 54 is an explanatory diagram showing a case where the association is changed by using a copy when changing the tip in the database system.

[Fig. 55] Fig. 55 is a flowchart describing in detail the sharing cancellation operation in the database system.

FIG. 56 is an explanatory diagram showing cancellation of sharing of a tip copy type model in the database system.

[Fig. 57] Fig. 57 is a diagram illustrating a modification of the database system and is an explanatory diagram illustrating cancellation of sharing of a tip ubiquitous model.

FIG. 58 is a diagram showing still another modification of the database system, and is an explanatory diagram showing cancellation of sharing of the tip extinction type model.

FIG. 59 is a view showing still another modified example of the above database system, which is a tip uneven distribution type model, and
It is explanatory drawing which shows the cancellation | release of sharing in the case of also erasing an empty bag.

FIG. 60 is a view showing still another modified example of the above database system, which is a tip disappearance model, and
It is explanatory drawing which shows the cancellation | release of sharing in the case of also erasing an empty bag.

[Fig. 61] Fig. 61 is a flowchart showing the operation of compressing the dirty bit list during the asynchronous period, showing still another modification of the database system.

FIG. 62 is a diagram showing another embodiment of the present invention, and is an explanatory diagram showing the relationship among tips, bags and databases shared among information processing devices in a database system.

FIG. 63 is an explanatory diagram showing a method of storing tips in an information processing device that processes the entire database as one bag in the database system.

[Explanation of symbols]

1/11 Database system 2/12 Information processing equipment (database management device) 23 Database management unit (database, state information storage means, priority side information storage means) 32 Float management unit (subset setting means) 51 Synchronization processing unit (synchronization Processing unit, collision detection unit, priority side determination unit, detection unit) 52 sharing setting processing unit (sharing setting unit) 53 sharing cancellation processing unit (sharing cancellation unit)

─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Tama Kamakura 22-22 Nagaike-cho, Abeno-ku, Osaka-shi, Osaka Inside Sharp Corporation (72) Inventor Fusayuki Fujita 22-22 Nagaike-cho, Abeno-ku, Osaka-shi, Osaka Sharp (56) References JP-A-9-134304 (JP, A) JP-A-9-128280 (JP, A) JP-A-8-328931 (JP, A) JP-A-8-241236 (JP, A) JP-A-11-327987 (JP, A) JP-A-12-40025 (JP, A) Shiro Iba Two outsiders, must-have PIM software for 11 product work groups, including Outlook 98 Japanese version, Hello! PC, Japan, SoftBank, May 8, 1998, Volume 5, No. 9, p. 162 ―― 181 (especially 178 ―― 181) Masahiro Nakamura, “Anywhere Office” software environment is set up, Nikkei Electronics, Japan, Nikkei BP, October 10, 1994, No. 619, p. 137 ―― 141 Hiroyuki Hasegawa, Personal Computer LAN Beginning with Windows95, Japan, AI Publishing, April 22, 1996, 1st edition, p. 218 ―― 226 Rei Watanabe, Make troublesome file management easier, Nikkei MAC, Nikkei BP, Japan, February 17, 1994, No. 11, p. 124 ―― 129 Masahiro Nakamura, Asynchronous updating of replicated files is the core technology of distributed database, Nikkei Electronics, Japan, Nikkei BP, June 6, 1994, No. 609, p. 101 ―― 110 (58 ) Fields surveyed (Int.Cl. ⁷ , DB name) G06F 12/00 G06F 13/00

Claims (5)

(57) [Claims] 1. To configure a database system having a plurality of database management devices and a synchronization processing means for matching the contents of corresponding data between databases stored in the database management devices which can communicate with each other. In a database management device that is provided and has at least one database, for each of its own data corresponding to the data in the other database, change of data during the asynchronous period between the previous synchronization processing and the next synchronization processing. Whether or not there is a plurality of data changed in the data set based on the respective state information for the set of state information storage means for storing the state information indicating And a collision detection means for detecting all of the contents of a plurality of changed data in the case of a plurality of To, to determine whether to give priority to any data, to the synchronization processing means, a preferential determining means for creating replace the contents of the data of the non-priority side by the contents of the data preferential, is its preferential determining means The determined priority side and the other party's finger
A database management device comprising: a detection unit that detects a mismatch with the determined priority side . 2. The database management according to claim 1, wherein the priority side determining means determines the priority side by comparing only a part of each of the changed plurality of data contents. apparatus. Wherein said preferential determining means, when the modified data of the plurality, a preference of any of the data, the user
The database management apparatus according to claim 1 , further comprising inquiring an upper application . 4. The modified data is changed by the priority side determining means.
If there is more than one, the user can select which data has priority.
Or inquire to the upper application and answer
The priority side is determined based on the
Priority is given to the synchronization processing means based on the determined priority side.
In addition to instructing the other side, communication between
The database according to claim 3, wherein the database is cut once.
Management device. 5. Data corresponding to each other among a plurality of databases
A program for managing the same contents of the data is recorded.
In the recorded medium, the state information storage means, the collision detection means, the priority,
Operates a computer as a side determination means and detection means
It is characterized in that the program for making it recorded
Recording medium.