JP5452516B2 - Distributed processing apparatus and distributed processing method - Google Patents

Description

  The present invention relates to a technique of a distributed processing apparatus and a distributed processing method.

  In the transaction processing system, when a transaction request is transmitted from the client to the processing server, the processing server performs the processing and returns the result as a response signal to the client. And it can respond to the increase in the number of clients by setting it as the distributed processing system in which a some processing server cooperates. When the load balancer of the distributed processing system accepts requests issued from clients as a single gateway, the load balancer distributes those requests to a plurality of processing servers.

  As a method for distributing the processing requested by the load balancer, there is consistent hashing (Non-patent Document 1). Consistent hashing has a simple interface that accepts only data storage requests and data acquisition requests. By looking at the identifier of the data requested by the received signal, any of the servers manages the target data. And the storage request and the acquisition request are distributed to this server.

D. Karger, E. Lehman, T. Leighton, M. Levine, D. Lewin, and R. Panigrahy, “Consistent Hashing and Random Trees: Distributed Caching Protocols for Relieving Hot Spots on the World Wide Web,” in Proceedings of the 29th ACM Symposium on Theory of Computing (STOC'97), May 1997, pp.654-663.

  A transaction handled by the transaction processing system may be related to a plurality of transaction states. Here, “related” between transactions means that, for example, B2BUA (Back-to-Back User Agent) in SIP (Session Initiation Protocol) terminates two calls by the application server and always sets these as a pair. It refers to a form that performs processing.

  B2BUA is an intermediary between two clients when a SIP session is connected between two clients and the two cannot be connected directly due to differences in the supported SIP methods. By providing another device (B2BUA), B2BUA is used as an adapter so that two clients can connect a session.

Thus, by distributing a plurality of related transactions to the same processing server, it is possible to shorten the processing waiting time between the plurality of transactions.
On the other hand, the relationship between a plurality of transactions may change over time. For example, when a two-party call between the user A and the user B by the B2BUA connection and a two-party call between the user C and the user D have been established, the calls are not disconnected, and AC, BD, respectively. The case of switching to a two-way call in between.
However, in the conventional transaction distribution processing, such efficient distribution considering the relationship between a plurality of transactions is not performed, and the transaction processing efficiency is lowered.

  Therefore, the main object of the present invention is to realize an efficient transaction distribution process in consideration of the relationship between a plurality of transactions.

  In order to solve the above-mentioned problems, the present invention provides a distributed processing apparatus in which a plurality of processing servers that process transaction request signals from clients exist and the request signals are distributed to the processing servers. The distributed processing device is a plurality of the processing servers, a data management device that stores transaction data for each of the processing servers, and a distribution destination of the request signal from the plurality of processing servers 1 A distribution server that selects a single processing server, and the storage unit of the distributed processing device includes a plurality of related transactions grouped and an allocation ID that indicates an allocation destination of data of these transactions The first correspondence data indicating the correspondence between the data management device and the data management device which is the placement destination of the data corresponding to the placement ID Second correspondence data indicating the correspondence between the device ID of the processing server connected to the location ID and the placement ID is stored, respectively, and the distribution server searches the placement ID from the received request signal. From the second correspondence data, the device ID of the corresponding processing server is specified as a distribution destination of the request signal, the request signal is transferred to the specified processing server, and the processing server When receiving a request signal, executing a transaction process, and performing a configuration change of a plurality of related transactions grouped by the execution, the arrangement ID and the transaction associated with the first correspondence data Corresponding to the second correspondence data from the changed arrangement ID of the correspondence. To the data management device, wherein connected to the processing server being, said from the data management device connected to the processing server before the change, wherein the moving data of the transaction.

  As a result, one allocation ID is assigned in units of a plurality of related transactions, and by changing the configuration of the plurality of related transactions, efficient transaction distribution processing considering the relationship between the plurality of transactions is performed. Can be realized.

  In the present invention, the plurality of related transactions are call connections between the plurality of clients terminated by B2BUA (Back-to-Back User Agent) in SIP (Session Initiation Protocol), and the configuration of the transaction is changed. Is a change of a speaker of the client when a call process is performed between the clients by the call connection.

  As a result, it is possible to replace the other party in a two-party call without disconnecting the call.

  In the present invention, the processing server receives the request signal, executes a transaction process, and accesses the data in the data management device different from the data management device connected to the processing server by executing the processing. The corresponding processing server is specified from the storage unit of the distributed processing device using the placement ID as a key, and the data is transmitted to the data management device connected to the specified processing server. It is characterized by performing access to.

  Thereby, the processing server can realize access to data that the processing server is not in charge of.

  According to the present invention, in the storage unit of the distributed processing device, as the correspondence information between the placement ID and the device ID, one device ID is assigned to a range indicating a range from the start point to the end point of the placement ID. When range-corresponding data to be associated is stored and a new processing server is added to the distributed processing device, one of the device IDs in the range-corresponding data is selected, and the selected range When the range of the corresponding data is divided, one of the divided ranges is associated with the device ID of the new processing server, and when the processing server is deleted from within the distributed processing device, The apparatus ID of the processing server to be deleted is selected, and the range of the selected data corresponding to the range is integrated with the range of the other processing server.

  As a result, even when processing servers are added, bottlenecks such as waiting for access processing between processing servers do not occur, and the performance of the transaction processing system can be improved.

  According to the present invention, it is possible to realize an efficient transaction distribution process in consideration of the relationship between a plurality of transactions.

It is a block diagram which shows the transaction processing system regarding one Embodiment of this invention. It is a block diagram which shows the apparatus structural example in the transaction processing system regarding one Embodiment of this invention. It is a block diagram which shows the detail of each server of the transaction processing system regarding one Embodiment of this invention. It is a flowchart which shows the transaction distribution process in the distribution server regarding one Embodiment of this invention. It is a flowchart which shows the execution process of the transaction in the processing server regarding one Embodiment of this invention. It is a block diagram which shows the transaction processing system in a 1st comparative example.

  Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a configuration diagram showing a transaction processing system. The transaction processing system is configured by connecting a client 1, a load balancer 2, and a distributed system 9 via a network. The distributed system 9 includes a distribution server 4, a processing server 5, and a data management device 6. The number of devices constituting the transaction processing system in FIG. 1 is not limited to the number illustrated in FIG. 1 and may be an arbitrary number.

  1 is configured as a computer having a CPU (Central Processing Unit), a memory, a hard disk (storage means), and a network interface. The computer reads the data into the memory. Each processing unit is operated by executing the program.

  As an example of a program that operates in a transaction processing system, there is distributed processing middleware for realizing an application server that performs call (session) processing such as a telephone. The distributed processing middleware prepares an application separately from the distributed processing middleware, and allows the application to realize service processing without being aware of the distribution of devices. By using such distributed processing middleware, it is possible to perform large-scale processing due to the effect of the number of multiple servers while making it look like a single application server from the outside.

The client 1 requests the processing server 5 to process a transaction. In response to the request from the client 1, the load balancer 2 selects one of the plurality of distribution servers 4 in the distributed system 9 and requests the selected distribution server 4 to perform processing.
Here, since the client 1 sees the distributed system 9 as a system that operates as a single device via the gateway called the load balancer 2, the device configuration in the distributed system 9 can be flexibly changed (addition of devices). Or delete).

The distribution server 4 receives a request from the load balancer 2, selects one of the plurality of processing servers 5, and requests the selected processing server 5 for processing. Here, in the selection processing of the processing server 5, the location of transaction data managed in the data management device 6 is referred to. First, the processing server 5 and the data management device 6 are connected at a ratio of 1: 1, and a certain data management device 6 stores data of a certain transaction. Therefore, it is necessary for the distribution server 4 to select the processing server 5 that accommodates the data management device 6 that manages the data of the requested transaction.
The processing server 5 receives the request from the distribution server 4 and processes the transaction request by accessing the transaction data in the data management device 6.

FIG. 2 is a configuration diagram illustrating an example of a device configuration in the transaction processing system. The difference from FIG. 1 is that each device (distribution server 4, processing server 5, and data management device 6) in the distributed system 9 is accommodated in the housing of one server device 8.
In general, communication between devices (communication via a network cable) has a communication speed slower than intra-device communication (communication via an internal bus) due to communication header processing or the like. Therefore, by storing a group of devices having a high communication frequency in one physical device called the server device 8 (distributed processing device), the frequency of in-device communication can be increased and the communication efficiency can be improved.

FIG. 3 is a configuration diagram showing details of each server of the transaction processing system.
The distribution server 4 includes a transaction distribution unit 41, a data arrangement table 42 (second correspondence data), and an arrangement search unit 43.
The processing server 5 includes a transaction processing unit 51, an arrangement ID assignment unit 52, a set change unit 53, another management data acquisition unit 54, and a transaction management table 55 (first correspondence data).

The transaction distribution unit 41 refers to the data arrangement table 42 to determine the processing server 5 that processes the transaction request from the load balancer 2 and transmits the request to the transaction processing unit 51 of the processing server 5.
The transaction processing unit 51 receives the transaction request, accesses the transaction data in the data management device 6, and processes the transaction request.

The transaction processing unit 51 assigns an arrangement ID when a change in the arrangement of transaction data in the data management device 6 occurs (new data creation, existing data movement, existing data deletion, etc.) in the course of processing a transaction request. The unit 52 is requested to perform allocation processing for allocation IDs allocated to transaction data.
In the course of processing a transaction request, the transaction processing unit 51 requests the set changing unit 53 to change the correspondence when the correspondence between the transaction data and the arrangement ID changes. .
In the process of processing a transaction request, the transaction processing unit 51 requests the other management data acquisition unit 54 to access the data when accessing the data management device 6 under the processing server 5 of the other device. To do.

Table 1 is an example of the data arrangement table 42. The data arrangement table 42 is a table that associates, for each ID of the processing server 5, a value range (showing a range from the start point ID to the end point ID) of the arrangement ID that the processing server 5 takes charge of.
The ID of the processing server 5 may be the same value as the arrangement ID that is the starting point of the arrangement ID in charge, or may be a random value.
An arrangement ID is an ID assigned to data of one or more transactions, and transaction data to which the same arrangement ID is assigned is stored in the same one data management device 6. For example, the transaction data to which the allocation ID “100” is assigned is stored in the data management device 6 under the processing server 5 “P1”.
The transaction data is data necessary when processing a transaction, and includes, for example, a call identifier, caller, receiver, and state transition in call processing.
The transaction state transition is a parameter indicating a call state such as busy or ringing.

The initial value of the correspondence data between the processing server ID and the arrangement ID in the data arrangement table 42 is given in advance by an administrator or the like. By setting the range of the placement ID so that there is no overlap or omission between the processing servers 5, one processing server 5 can be obtained from an arbitrary placement ID.
In addition, it is desirable that the value range of the arrangement ID in the data arrangement table 42 is registered in advance in the data arrangement table 42 for the arrangement IDs that have not yet been assigned to any transaction data (unissued). This eliminates the need to update the data placement table 42 without depending on the number of placement ID issuance and deletion processes.

Table 2 is a table illustrating the data arrangement table 42 when one server P5 is added to the data arrangement table 42 of Table 1.
Along with the addition of the server P5, a value range of one arrangement ID is selected from the value ranges of the arrangement IDs each of which the four existing processing servers 5 are in charge of, and the selected value range is divided into the server P5. Also assign.
In Table 2, the value range (600 to 999) of the server P4 in charge of the largest data amount (D = 400) is selected from among the data amounts handled by each processing server before the division, and the selected before the division. The range is divided into two (600 to 799, 800 to 999). Thereby, the data amount for each processing server is made uniform.
Similarly, when the number of processing servers decreases (for example, when a failure occurs in the processing server P3), the processing servers in charge (P2, P4) in charge of before and after the value range (400 to 599) that the processing server P3 to be decreased is responsible for. Of the two selected units and the processing server with the smaller amount of data currently being handled (in the case of Table 2, the data amount is the same, P2), the value range assigned to the processing server to be reduced Are integrated (200 to 599).

As described above, when the number of processing servers 5 with respect to the data placement table 42 is increased or decreased, the data placement table 42 needs to be updated. However, the amount of change in the range of the placement ID that is the update range is the responsibility of the processing server 5 for one unit. Just the amount. For example, when the number of processing servers 5 is increased to five, a change amount of 1/5 is sufficient, and when the number of processing servers 5 is increased to N, a change amount of 1 / N is sufficient.
Since the change of the processing server 5 in charge is accompanied by the movement of data in the data management device 6 accommodated by the processing server 5, the load of the data movement process is reduced by reducing the amount of change of the data arrangement table 42. be able to.

When the transaction distribution unit 41 receives a transaction processing request from the load balancer 2, the transaction distribution unit 41 determines the processing server 5 in charge of the transaction processing request from the data arrangement table 42 using the arrangement ID as a search key. When the transaction processing request does not include an arrangement ID, the arrangement ID may be acquired from the transaction management table 55 (Table 3) described later using the transaction ID as a search key.
That is, the transaction distribution unit 41 specifies the processing server 5 instead of the data manager by using a logic similar to the specification of the data manager of the consistent hashing.
Then, the transaction distribution unit 41 notifies the transaction processing unit 51 of the determined processing server 5 of the transaction processing request.

  The placement ID assigning unit 52 assigns one placement ID to one or more transactions and stores the result in the transaction management table 55. When a plurality of transactions are grouped and one arrangement ID is assigned, the plurality of transactions assigned the same arrangement ID are transactions that are related to each other. Examples of the “related transaction” include two call processes connected by B2BUA, but the type of transaction is not limited to call processing, and the number of transactions is not limited to two.

Table 3 is a table illustrating the transaction management table 55. The transaction management table 55 is a table that associates, for each placement ID, the ID of one or more transactions to which the placement ID is assigned.
For example, in the records of the arrangement ID “0” and the arrangement ID “1”, two transactions connected by B2BUA are grouped and correspond to one arrangement ID.
The record with the arrangement ID “2” corresponds to one very simple transaction “T4” corresponding to a SIP proxy in SIP.

Table 4 is a table illustrating the transaction management table 55 (change of arrangement ID = 0, 1).
The set change unit 53 reflects the change in the transaction management table 55 when the relevance between the plurality of transactions is changed during transaction processing, and based on the updated contents of the transaction management table 55, It controls the movement process of the transaction data in the data management device 6.

In Table 4, when the transaction “T0, T1” with the arrangement ID “0” and the transaction “T2, T3” with the arrangement ID “1” are before the change, the set changing unit 53 sets T1 and T2 respectively. By moving between records, the transaction management table 55 is updated as the transaction “T0, T2” with the arrangement ID “0” and the transaction “T1, T3” with the arrangement ID “1”.
Note that such a process for updating the transaction management table 55 occurs, for example, when a speaker is replaced in B2BUA.

  The other management data acquisition unit 54 receives an instruction from the transaction processing unit 51 and performs data access to the data management device 6 accommodated in the other processing server 5. Note that the acquisition instruction of data managed by the other device from the transaction processing unit 51 is, for example, confirming whether each transaction state of the replacement B2BUA is in a call establishment state when performing speaker replacement in B2BUA. Notified when an access to another B2BUA state occurs.

First, the other management data acquisition unit 54 requests the allocation server 4 to identify the processing server 5 in charge of the allocation ID by notifying the allocation search unit 43 of the allocation ID notified from the transaction processing unit 51. .
The arrangement search unit 43 refers to the data arrangement table 42 and identifies the processing server 5 from the arrangement ID.
The other management data acquisition unit 54 acquires the data received the instruction from the transaction processing unit 51 from the data management device 6 accommodated in the processing server 5 which is the specified other device. Thereby, even if a plurality of processing servers 5 do not directly share (connect) one data management device 6, it is possible to acquire data that is not handled by itself.

FIG. 4 is a flowchart showing transaction distribution processing executed by the transaction distribution unit 41 of the distribution server 4.
In S101, a signal is received from the load balancer 2.
In S102, it is determined whether or not an arrangement ID is included in the signal of S101. If Yes in S102, the process proceeds to S104, and if No in S102, the process proceeds to S103.
In S103, the allocation ID allocation unit 52 is requested to allocate an allocation ID newly allocated to the signal in S101, and the process proceeds to S104.
In S104, with reference to the data arrangement table 42, the processing server 5 in charge of the arrangement ID is specified from the arrangement ID assigned to the signal in S101.
In S105, the signal received in S101 is transferred to the processing server 5 identified in S104.

  FIG. 5 is a flowchart showing transaction execution processing in the processing server 5.

In S201, the processing server 5 receives the signal of S105.
In S202, the transaction processing unit 51 executes transaction processing while appropriately accessing the data of the transaction corresponding to the signal in S201 with respect to the data management apparatus 6.
When an access to data managed by another device occurs during the process of S202 (S203, Yes), the transaction processing unit 51 accesses the data managed by the other device via the other management data acquisition unit 54. Execute (S204) and proceed to S205. On the other hand, when no access occurs (S203, No), the process proceeds to S205.

  When a transaction that constitutes a transaction group to which one arrangement ID in the transaction management table 55 is assigned during the process of S202 occurs (S205, Yes), the transaction processing unit 51 passes the set change unit 53. Then, the transaction management table 55 is updated (S206), and the arrangement ID embedded in the response signal to the signal received in S201 is changed to the changed arrangement ID in S206 (S207), and the process proceeds to S211. On the other hand, when no transaction change occurs (S205, No), the process proceeds to S211.

In S211, the transaction processing unit 51 ends the transaction process executed in S202.
In S212, a response signal to the received signal in S201 is transmitted to the client 1 based on the results processed in S202 to S211.

  According to the present embodiment described above, the set change unit 53 updates the distribution of transactions to the transaction management table 55 in accordance with the change in the relationship between the transactions, whereby the relationship between the transactions. Efficient transaction distribution processing considering the above can be realized.

Furthermore, according to the present embodiment, the content of the data placement table 42 is configured to allocate a range of placement IDs for each processing server 5 in order to make the change processing of the data placement table 42 as small as possible. Hereinafter, in order to explain the effect of the configuration contents of the data arrangement table 42, comparison with three comparative examples (first comparative example to third comparative example) is performed.
(First Comparative Example) A configuration without a data arrangement table.
(Second Comparative Example) A configuration in which the arrangement ID is associated with the processing server 5 every time an arrangement ID to be used is issued as the data arrangement table.
(3rd comparative example) The structure matched with the process server 5 based on the value of the result of the remainder calculation which uses arrangement ID as an input parameter as a data arrangement table.

  FIG. 6 is a configuration diagram showing the transaction processing system in the first comparative example. This first comparative example differs from the transaction processing system of FIG. 1 in that it does not have a distribution server 4 and in that a plurality of processing servers share one data management device.

  When a load balancer distributes a transaction request to a processing server, the distribution processing server can access data from the same data management apparatus regardless of which processing server is distributed, so the data allocation table 42 is not necessary. .

  However, in a configuration in which data management devices are shared, when a plurality of related transaction processes occur on a plurality of processing servers, a wasteful time for waiting for the respective data access between the plurality of processing servers occurs as a bottleneck. Therefore, even if the number of processing servers is increased, it is difficult to improve the performance of the entire transaction processing system.

Table 5 is a table illustrating a data arrangement table (management for each arrangement ID) of the second comparative example. The distribution server of the second comparative example refers to the data allocation table for each allocation ID in Table 5 to identify the processing server that is the transaction distribution destination.
However, since the number of arrangement IDs is the number of transactions, the number of records in the data arrangement table increases, and the data arrangement table update process becomes a bottleneck of the transaction processing system.

Table 6 is a table illustrating a data arrangement table (management by remainder calculation) of the third comparative example. The function “mod (a, b) = c” is a remainder function indicating that the remainder when the dividend a is divided by the divisor b is c. Before the addition of the server P5, for example, the distribution server that has received the request for the arrangement ID = 10 is 10 ÷ 4 = 2 remainder 2. Therefore, the server P3 corresponding to the remainder 2 is selected as the distribution destination.
However, in the third comparative example, a large bottleneck occurs when the number of processing servers increases or decreases. For example, it is assumed that a server P5 is newly added to four servers P1 to P4. Since the number of servers corresponds to the divisor in the remainder calculation, even if only one server is added, the contents of the data arrangement table are greatly increased (strictly, if N is the number of processing servers, the total (N -1) / N) It is changed. For example, in the configuration of Table 6, with the change from N = 4 to N = 5, the entire 4/5 contents are changed in the data arrangement table.
It is necessary not only to change the data arrangement table but also to move the transaction data itself to the change destination in accordance with the contents of the data arrangement table after the change. This data movement process is the same as in the third comparative example. It is a big bottleneck.

  On the other hand, in the data arrangement table 42 (Tables 1 and 2) of the present embodiment, the arrangement ID is set as a value range, and when the number of processing servers 5 is changed, the arrangement ID value range is simply changed (divided or integrated). Therefore, the bottleneck of the transaction processing system generated in each of the first to third comparative examples does not occur, and the performance of the transaction processing system can be improved as the number of processing servers 5 increases.

DESCRIPTION OF SYMBOLS 1 Client 2 Load balancer 4 Distribution server 5 Processing server 6 Data management apparatus 8 Server apparatus 9 Distribution system 41 Transaction distribution part 42 Data arrangement table (2nd corresponding data)
43 Arrangement Search Unit 51 Transaction Processing Unit 52 Arrangement ID Assignment Unit 53 Set Change Unit 54 Other Management Data Acquisition Unit 55 Transaction Management Table (First Corresponding Data)

Claims (8)

There are a plurality of processing servers that process transaction request signals from clients, and the distributed processing device distributes the request signals to the processing servers,
The distributed processing device includes a plurality of the processing servers, a data management device that stores transaction data for each of the processing servers, and a distribution destination of the request signal from the plurality of processing servers. And a sorting server that selects the processing server of
In the storage means of the distributed processing device,
First corresponding data indicating a correspondence between a plurality of related transactions grouped and an arrangement ID indicating an arrangement destination of data of the transactions;
Second correspondence data indicating the correspondence between the device ID of the processing server connected to the data management device, which is the placement destination of the data corresponding to the placement ID, and the placement ID are stored, respectively.
The distribution server identifies the device ID of the corresponding processing server as the distribution destination of the request signal from the second correspondence data, using the arrangement ID as a search key, from the received request signal, and identifies it Transferring the request signal to the processing server;
The processing server receives the request signal, executes a transaction process, and performs a configuration change of a plurality of related transactions grouped by the execution, and is associated with the first corresponding data. Change the correspondence relationship between the placement ID and the transaction, and change the changed placement ID of the correspondence relationship to the data management device connected to the processing server associated with the second correspondence data. A distributed processing device, wherein transaction data is moved from the data management device connected to the previous processing server. The plurality of related transactions are call connections between the plurality of clients terminated by B2BUA (Back-to-Back User Agent) in SIP (Session Initiation Protocol),
The distributed processing apparatus according to claim 1, wherein the transaction configuration change is a change of a speaker of the client when a call process is performed between the clients by the call connection. When the processing server receives the request signal, executes a transaction process, and accesses the data in the data management device different from the data management device connected to the processing server by executing the processing. The corresponding processing server is identified from the storage means of the distributed processing device using the arrangement ID as a key, and the data management device connected to the identified processing server is accessed to the data. The distributed processing apparatus according to claim 1, wherein the distributed processing apparatus is performed. The storage unit of the distributed processing device stores, as the second correspondence data, value range correspondence data that associates one device ID with a value range indicating a range from the start point to the end point of the placement ID. ,
When a new processing server is added to the distributed processing device, one range of the range corresponding data is selected, the range of the selected range corresponding data is divided, and one of the divided ranges Associating a range with the device ID of the new processing server;
When the processing server is deleted from within the distributed processing device, the range corresponding to the processing server to be deleted is selected from the range corresponding data, and the range of the selected range corresponding data is selected as another range. The distributed processing apparatus according to claim 1, wherein the distributed processing apparatus is integrated into a value range of a processing server. There are a plurality of processing servers that process transaction request signals from clients, and a distributed processing method by a distributed processing device that distributes the request signals to the processing servers,
The distributed processing device includes a plurality of the processing servers, a data management device that stores transaction data for each of the processing servers, and a distribution destination of the request signal from the plurality of processing servers. And a sorting server that selects the processing server of
In the storage means of the distributed processing device,
First corresponding data indicating a correspondence between a plurality of related transactions grouped and an arrangement ID indicating an arrangement destination of data of the transactions;
Second correspondence data indicating the correspondence between the device ID of the processing server connected to the data management device, which is the placement destination of the data corresponding to the placement ID, and the placement ID are stored, respectively.
The distribution server identifies the device ID of the corresponding processing server as the distribution destination of the request signal from the second correspondence data, using the arrangement ID as a search key, from the received request signal, and identifies it Transferring the request signal to the processing server;
The processing server receives the request signal, executes a transaction process, and performs a configuration change of a plurality of related transactions grouped by the execution, and is associated with the first corresponding data. Change the correspondence relationship between the placement ID and the transaction, and change the changed placement ID of the correspondence relationship to the data management device connected to the processing server associated with the second correspondence data. A distributed processing method, comprising: moving transaction data from the data management apparatus connected to the previous processing server. The plurality of related transactions are call connections between the plurality of clients terminated by B2BUA (Back-to-Back User Agent) in SIP (Session Initiation Protocol),
6. The distributed processing method according to claim 5, wherein the transaction configuration change is a speaker change of the client when a call process is performed between the clients by the call connection. When the processing server receives the request signal, executes a transaction process, and accesses the data in the data management device different from the data management device connected to the processing server by executing the processing. The corresponding processing server is identified from the storage means of the distributed processing device using the arrangement ID as a key, and the data management device connected to the identified processing server is accessed to the data. The distributed processing method according to claim 5, wherein the distributed processing method is performed. In the storage unit of the distributed processing device, as the correspondence information between the arrangement ID and the apparatus ID, a value range correspondence that associates one device ID with a value range indicating a range from the start point to the end point of the arrangement ID. Data is stored,
When a new processing server is added to the distributed processing device, one range of the range corresponding data is selected, the range of the selected range corresponding data is divided, and one of the divided ranges Associating a range with the device ID of the new processing server;
When the processing server is deleted from within the distributed processing device, the range corresponding to the processing server to be deleted is selected from the range corresponding data, and the range of the selected range corresponding data is selected as another range. The distributed processing method according to any one of claims 5 to 7, wherein the processing server is integrated into a value range.

Images