JP2008299769A - Data transfer scheduling device, program, storage medium, and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To transfer finely divided partial files one-by-one via proximity transfer even if a node group in the transfer destination is changed for quickly transferring a file to the final node. <P>SOLUTION: A data transfer scheduler 10 generates and manages a schedule for finely dividing data and transferring them from a file sever 12 to nodes 16-1 to 16-5. When a transfer request receipt part 24 receives a transfer request designating a file name and a plurality of transfer destination node groups, a transfer order management part 26 disassembles the node transfer request into node transfer requests to the respective transfer destination nodes and combines them with the file name after sorting them to generate a file management list 46-1 showing the node transfer order, and asks a finely divided file delivery management part 32 to transfer the finely divided partial files via proximity node transfer according to the order of the file management list 46-1. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

The present invention relates to a data transfer schedule device, program, recording medium, and method for scheduling and managing data transfer from a file server to a plurality of nodes, and in particular, data that is subdivided and transferred in a bucket relay manner between adjacent nodes. The present invention relates to a transfer schedule device, a program, a recording medium, and a method.

  Conventionally, in an environment that uses a large number of servers such as large-scale server farms and data centers, data transfer that efficiently distributes input files when many applications that require large input files are executed Is required.

In conventional data transfer, a file transfer request is sent from a server that requires a file to a file server where the file is placed, and the file is transferred from the file server to the server in response to the request.
JP 2002-149473 A JP-A-9-160820

  However, in such conventional data transfer, when a file is required by a plurality of servers, for example, when a file is required by n servers, file transfer occurs n times from the file server. In such 1-to-n transfer, there is a problem that a file server becomes a bottleneck and it takes time until the files are distributed throughout.

  In order to solve this problem, there is a method in which a file transfer agent is arranged in each server, the file is held as a cache, and a file is transferred from a neighboring server. However, even in this data transfer, there is a problem that it takes time to transfer the file n times before the file is copied to the last one.

  On the other hand, in situations such as when a server is installed, a bucket relay that shortens the file transfer time by predetermining the transfer order of files to a plurality of servers and sequentially transferring the subdivided partial files between the servers in a predetermined order There is a data transfer method by a method.

  However, the data transfer by the bucket relay method can be used only in the case of distributing to the entire set of servers in which the server installation status can be grasped, and the data transfer in which the server that distributes the file changes for each transfer request. There is a problem that cannot be used.

The present invention relates to a data transfer schedule device, program, recording medium, and method capable of transferring a partial file subdivided by bucket relay transfer between adjacent nodes and transferring the file to the final node at a high speed even if the transfer destination node group is changed The purpose is to provide.

(apparatus)
The present invention provides a data transfer schedule device. The present invention relates to a data transfer schedule device that schedules and manages data transfer from a file server to a plurality of nodes.
A transfer request receiving unit that receives a transfer request specifying a file name and a plurality of transfer destination nodes;
After decomposing and sorting the transfer requests into node transfer requests for each transfer destination node, generating a file management list indicating the order of the transfer nodes by concatenating the file names, and subdividing the partial files according to the order of the file management list A file transfer order management unit that requests file transfer to transfer between adjacent nodes;
When a file transfer request is received, a subdivided file delivery management unit that instructs the file server and node that are the request destination to execute file transfer, and
It is provided with.

  Here, the file transfer order management unit limits the number of transfers to the next node of the same partial file at the node that requested the file transfer to one, and the partial files are arranged in the order of the node transfer request in the file management list. Transfer sequentially.

The file transfer order manager
After the transfer request is broken down into node transfer requests for each transfer destination node, the file transfer list is sorted by file name and transfer destination address, and the file transfer list indicating the order of transfer nodes is generated by concatenating the node transfer requests sorted into file names. A sort processing unit;
A cache location acquisition unit that collects a list of retained files cached in the file server and all nodes, and detects and sets the cache location of the transfer source file in the file management list;
For each node transfer request arranged in the file management list, progress management information indicating the transfer progress status of the partial file to the next node, and transfer count management that limits the transfer count of the same partial file to the next node to a predetermined maximum transferable number A transfer information management unit for generating and managing information,
File transfer that determines the transferable node based on the progress management information and the transfer count management information, requests the file transfer management unit to transfer the file, and updates the transfer count of the same partial file in the transfer count management information A request department;
When the file transfer completion notification requested from the subdivision file distribution management unit is received, the transfer progress status of the partial files described in the progress management information is updated, and the transfer completion of all the partial files is determined. The corresponding node transfer request is deleted from the file management list, and if the transfer count of the same partial file in the transfer count management information reaches the maximum transferable number, the transfer count is set to 1 for the next partial file transfer. A transfer completion processing unit to reduce
Is provided.

  The maximum transferable number of the same partial file for the next node managed by the transfer information management unit is set to 1.

The file transfer request department
Check the progress management information and transfer count management information of each node including the file server linked to the node transfer request arranged in the file management list in the order of arrangement,
If the number of partial file transfers in the transfer count management information is less than the maximum transferable number and the progress management information does not indicate completion of transfer of all partial files, it is determined as a transferable node.

(program)
The present invention provides a program for a data transfer schedule. The program of the present invention is a computer for a data transfer schedule device that schedules and manages data transfer from a file server to a plurality of nodes.
A transfer request reception unit that receives a transfer request specifying a file name and a plurality of transfer destination nodes;
After decomposing and sorting transfer requests into node transfer requests for each transfer destination node, generate a file management list indicating the order of the transfer nodes by concatenating the file names, and subdivide the partial files according to the order of the file management list. File transfer order management unit that requests file transfer to transfer between adjacent nodes,
When a file transfer request is received, a fragmented file distribution management unit for instructing and executing file transfer to the file server and node as a request destination,
It is made to function as.

(recoding media)
The present invention provides a computer-readable recording medium storing a program for a data transfer schedule. The recording medium of the present invention is
A data transfer scheduling device computer that schedules and manages data transfer from a file server to a plurality of nodes,
A transfer request receiving unit that receives a transfer request specifying a file name and a plurality of transfer destination nodes,
After decomposing and sorting the transfer requests into node transfer requests for each transfer destination node, generating a file management list indicating the order of the transfer nodes by concatenating the file names, and subdividing the partial files according to the order of the file management list File transfer order management unit that requests file transfer to transfer between adjacent nodes,
When a file transfer request is received, a fragmented file distribution management unit for instructing and executing file transfer to the file server and node as a request destination,
It stores the program which functions as

(Method)
The present invention provides a data transfer scheduling method. The present invention relates to a data transfer scheduling method for scheduling and managing data transfer from a file server to a plurality of nodes.
A transfer request reception process for receiving a transfer request specifying a file name and a plurality of transfer destination nodes;
After decomposing and sorting the transfer requests into node transfer requests for each transfer destination node, generating a file management list indicating the order of the transfer nodes by concatenating the file names, and subdividing the partial files according to the order of the file management list A file transfer order management process that requests file transfer to transfer between adjacent nodes;
When a file transfer request is received, a subdivided file distribution management process for instructing and executing file transfer to the file server and node as a request destination;
It is provided with.

According to the present invention, when a transfer request specifying a file name and a transfer destination node group is received, the received transfer request is decomposed into node transfer requests for each node so that transfer between adjacent nodes can be performed. For example, a file management list indicating the node transfer order arranged in units of files in the order sorted by the IP address of the node is generated, and a file transfer is determined by determining a transferable node based on the file management list, and By restricting the number of transfers of the same partial file between nodes to be allowed only once, the partial file subdivided according to the transfer order of the file management list is transferred in a bucket relay from the file server to each node, Files can be transferred at high speed to the final node.

  FIG. 1 is an explanatory diagram showing an embodiment of a data distribution system including a data transfer scheduler according to the present invention. In FIG. 1, a file server 12 serving as a file transfer source is connected to the data transfer scheduler 10 of the present embodiment, and nodes 16-1 to 16-5 serving as file transfer destinations are connected via a network 14. ing.

  Data transfer agents 18-1 to 18-5 are arranged in the nodes 16-1 to 16-5, respectively, and are subdivided into a predetermined fixed size between adjacent nodes in response to a transfer command from the data transfer scheduler 10. Transfer the partial file using the bucket relay method.

  Each of the nodes 16-1 to 16-5 is provided with transfer caches 20-1 to 20-5 for temporarily holding transfer files. The nodes 16-1 to 16-5 are connected to storages 22-1 to 22-5 for storing transferred files and using them in applications.

  The data distribution system including the data transfer scheduler 10 of the present embodiment shown in FIG. 1 is a grid environment such as a data farm or a data center, and the nodes 16-1 to 16-5 are relatively large from the file server 12. It functions as a server that is executed by an application upon receiving a file transfer.

  The network 14 connecting the transfer scheduler 10 and the nodes 16-1 to 16-5 is, for example, the Internet protocol (IP protocol network). The data transfer scheduler 10, the file server 12, and the nodes 16-1 to 16-5 are connected to each other. Each has a unique IP address, and each of the nodes 16-1 to 16-5 is preset with a unique node name for identification.

  FIG. 2 is a block diagram showing a functional configuration of the data transfer scheduler in the present embodiment. In FIG. 2, the data transfer scheduler 10 of this embodiment includes a transfer request receiving unit 24, a transfer order management unit 26, a control unit 28, a cache location acquisition unit 30, and a subdivided file distribution management unit 32.

  The transfer order management unit 26 includes a sort processing unit 34 and a transfer request storage unit 36. The control unit 28 includes a transfer information management unit 38, a file transfer request unit 40, and a transfer completion processing unit 42. Each processing unit provided in such a data transfer scheduler is a function realized by executing a program in a computer that implements the data transfer scheduler 10.

  In this example, host1 to host5 are set as the node names of the nodes 16-1 to 16-5 connected to the data transfer scheduler 10 via the network 14.

  The transfer request receiving unit 24 provided in the data transfer scheduler 10 receives and holds transfer requests 44-1 to 44-2 from clients connected to the network 14.

  The transfer requests 44-1 and 44-2 received by the transfer request receiving unit 24 have, for example, the contents shown in FIG. FIG. 3 shows an example of XML description of the transfer request 44-1 in FIG. In the transfer request 44-1, “file A” is described as the file name 48 to be transferred.

  The transfer request 44-1 describes the transfer destination node group 50. In this example, the five node names “host1 to host5” of the nodes 16-1 to 16-5 shown in FIG. is described.

  Referring to FIG. 2 again, the transfer requests 44-1 and 44-2 received by the transfer request receiving unit 24 are sent to the transfer order management unit 26 in the order of reception. The transfer order management unit 26 decomposes, for example, the transfer request 44-1 received from the transfer request reception unit 24 into node transfer requests for each transfer destination node, and then sorts the decomposed node transfer requests by, for example, IP addresses to obtain file names A file management list 46-1 shown in the transfer request storage unit 36 indicating the node transfer order arranged in combination with is generated.

  The transfer order management unit 26 subdivides the file A to be transferred into, for example, five partial files A1 to A5, and the subdivided file so that the transfer between adjacent nodes is performed according to the node transfer order of the file management list 46-1. Request the file transfer to the distribution management unit 32.

  When the subdivision file distribution management unit 32 receives a transfer request from the transfer order management unit 26, the subdivision file distribution management unit 32 instructs the file server 12 and nodes 16-1 to 16-4 as the request destination to perform file transfer, and performs transfer between adjacent nodes. Execute.

  The function of the transfer order management unit 26 will be described in detail together with the control unit 28 as follows. First, when the sort processing unit 34 receives, for example, the transfer requests 44-1 and 44-2 from the transfer request receiving unit 24, the sort request unit 34 decomposes the transfer requests 44-1 and 44-2 into node transfer requests for each transfer destination node. Thereafter, the file management lists 46-1 and 46 are sorted by the file name and the IP address as the transfer destination address, and the node transfer requests are arranged in the order of the node transfer in the file name unit, that is, the file name A unit and the file name B unit. -2 is generated.

  For each of the note transfer requests indicated by “host1 to host5” or “host3, host5” arranged in the file management lists 46-1 and 46-2, the progress management information to be clarified in the following explanation is initialized. Is linked. The progress management information is partial file transfer progress information for the next node, that is, information indicating how many sub-files have been sent to the next node.

  For example, since the progress management information linked to the file management list 46-1 of the file name A transfers the file A divided into, for example, six partial files A 1 to A 6, the progress management information is illustrated at a certain transfer timing. Thus, “1/6, 2/6, 3/6, 4/6, 5/6” is recorded in the transfer order.

  For example, taking the file management list 46-1 as an example, the transfer information management unit 38 provided in the control unit 28 applies to the next node based on the node transfer requests indicated by “host1 to host5” arranged in the file management list 46-1. It generates and manages progress management information indicating the transfer progress of the partial file and a transfer count management information section that limits the transfer count of the partial file to the next node to a predetermined maximum transfer count.

  4 shows the link destination file management list 46-1 generated from the transfer request 44-1 of FIG. 3 managed by the transfer information management unit 24 of FIG. 2, and progress management information 52 managed by linking to this. FIG. 5 is an explanatory diagram showing 54 transfer count management information.

  4, the file management list 46-1 is created by the processing of the sort processing unit 34 of FIG. 2 for the transfer request 44-1 of FIG. This file management list 46-1 is changed from the transfer destination node name included in the transfer request 44-1 to the node transfer requests 56-1 to 56-5 for each transfer destination node with respect to the file name 48 of the transfer request 44-1. After disassembling, the node transfer requests that have been disassembled are generated by sorting them into file units and IP address units and arranging the file names 48.

  Simultaneously with the generation of the file management list 46-1, the sort processing unit 34 initializes and links progress management information 60-1 to 60-5 indicating partial file transfer progress information for the next node for each transfer destination node. Yes.

  The progress management information 60-1 to 60-5 initialized and linked to the node transfer requests 56-1 to 56-5 records the transfer progress information of the fragmented partial files transferred between the nodes. For example, if the number of subdivisions is 6, every time a partial file is transferred to the next node, the transfer progress information records information that changes to 1/6, 2/6, 3/6, 4/6, 5/6. And manage.

  Here, the information of the node transfer requests 56-1 to 56-5 arranged in the file management list 46-1 in FIG. 4 includes the file name 68 and the transfer destination node name as shown in FIG. 70.

Also, the node progress management information 60-1 to 60-5 arranged in the progress management information 52 of FIG. 4 is extracted as shown in FIG. Is described. Here, the partial file progress ratio 72 is (partial file progress ratio) = (number of partial file transfers completed) / number of subdivisions)
Given in.

  Specifically, as shown in the node progress management information 60-1 to 60-5 of the progress management information 52 in FIG. 4, if the number of subdivisions = 6, the partial file progress ratio is determined according to the progress of the partial file transfer. It changes to 0/6, 1/6, 2/6, 3/6, 4/6, 5/6, 6/6. The partial file progress ratio 0/6 is an initial value at which partial files are not transferred, and the partial file progress ratio 6/6 is a value indicating completion of transfer when all the partial files have been transferred.

  4 describes the node name 62, the maximum transferable number 64, and the current transfer number 66, as shown in FIG. 5C. The transfer count management information 54 is subdivided and indicates the number of times that the same partial file can be transferred between the nodes when transferring the partial file from the file transfer source file server 12 to the nodes 16-1 to 16-5 by the bucket relay method. This is information for limiting.

  That is, the minimum transferable number 64 set in the transfer count management information 52 is the maximum number that can transfer the same partial file to the next node, and one current transfer number 66 is added to each transfer to the next node. When the current transfer number 66 reaches the maximum transferable number 64, the transfer of the same partial file to the next node is prohibited thereafter.

  In the present embodiment, by setting the maximum transferable number in the transfer count management information 54 to 1, the same part is transferred from the file server 12 to the divided partial files to the transfer destination nodes 16-1 to 16-5. Restrict files so that they can be transferred only once between nodes. As a result, bucket relay transfer can be realized in which partial files are sequentially transferred from the file server 12 to the nodes 16-1 to 16-5 one by one between adjacent servers.

  In the transfer count management information 54 shown in FIG. 4, as shown by the node names 62-1 to 62-5, the host names of the nodes 16-1 to 16-4 excluding the file server 12 and the last node 16-5. host1 to host4 are described.

  The next maximum transferable number 64-1 to 64-5 is “1” set in advance. The last current transfer number 66-1 to 66-5 is updated from 0 to 1 every time a partial file transfer is requested to the next node, and is changed from 1 to 0 when a notification of transfer completion of the requested partial file is received. Prepare for the next partial file transfer.

  Referring to FIG. 2 again, the file transfer request unit 40 provided in the control unit 28 can be transferred based on the progress information management information 52 and the transfer count management information 54 shown in FIG. A node is determined to request the subdivision file distribution management unit 32 to transfer the file. Further, the file transfer request unit 40 updates the current transfer numbers 66-1 to 66-5 indicating the transfer count of the same partial file at the time of transfer request from 0 to 1 at this timing corresponding to each node.

  The transfer completion processing unit 42 provided in the control unit 28 receives the node transfer management information 60-in the progress management information 52 of FIG. 4 when receiving the completion notification of the requested file transfer from the segmented file distribution management unit 32. Update to increase the numerator of the partial file progress ratio of 1-60-5 by one.

  As a result of this update, when the partial file progress ratio denominator number and the partial file transfer completion number of the numerator match, for example, 6/6, transfer completion of all partial files is determined, and the corresponding node transfer is performed. Delete the request from the file management list.

  If the current transfer numbers 66-1 to 66-5 in the transfer count management information 54 in FIG. 4 have reached the value of the maximum transferable numbers 64-1 to 64-5, specifically, 1, The current transfer number 66-1 to 66-5 is decreased by one to zero to transfer a partial file.

  Further, the cache location acquisition unit 30 provided in the data transfer scheduler 10 of FIG. 2 collects a list of retained files cached in the file server 12 and the nodes 16-1 to 16-5, and a file management list 46-1. Alternatively, the transfer source cache location is described and set in 46-2. Thereby, for example, transfer source file server information 58 indicating that the file server 12 has an original file to be transferred in the file management list 46-1 can be set in the file management list 46.

  In the file management list 46-1 of FIG. 4, the file server detected by the processing of the cache position acquisition unit 30 of FIG. Then, the file name 48 in the file management list 46 is linked.

  FIG. 6 shows that the file A is subdivided into six partial files A1 to A6 based on the file management list 46-1 arranged in the transfer request storage unit 36 of FIG. 2 and is transferred to the nodes 16-1 to 16-5 by the bucket relay method. 5 is a time chart showing how data is transferred.

  In FIG. 6, at the timing of step S1, the first partial file A1 subdivided from the file server 12 is first transferred to the first node 16-1. The number of transfers of the partial file A1 from the file server 12 to the node 16-1 is performed only once by setting the maximum transferable number to 1.

  At the timing of the next step S2, since the file server 12 can send the first partial file A1 only once, the next partial file A2 is transferred to the node 16-1. At the same timing, the node 16-1 transfers the partial file A1 received at the timing of step S2 stored in the cache to the next node 16-2.

  Similarly, the file server 12 sequentially transfers the partial files A3, A4, A5, A6 one by one to the next node 16-1 at each timing of steps S3 to S6. In the node 16-2, the partial files A1 to A6 are sequentially transferred to the next node 16-2 at the timings of steps S2 to S7.

  Similarly for the remaining nodes 16-3 to 16-4, the same partial file is sequentially transferred to the next node only once at each timing. As a result, the transfer of the last partial file A6 to the last node 16-5 is completed at the timing of step S10, whereby the nodes 16-1 to 16- that are the transfer destinations of the file A1 subdivided from the file server 12 are thereby obtained. The file transfer for all of 5 is completed.

  FIG. 7 is a time chart showing the state of data transfer based on the file management list 46-2 generated by the data transfer scheduler 10 from the transfer request 44-2 of FIG. As in the case, it is divided into six partial files B1 to B6 and transferred to two nodes 16-3 and 16-5 as transfer destination nodes.

  In FIG. 7, since the transfer destination nodes are two nodes 16-3 and 16-5, the file server 12 transfers the partial files B1 to B6 to the node 16-3 once at the timing of steps S1 to S6. To do.

  Further, the node 16-3 transfers the partial files B1 to B6 cached at the timings of steps S2 to S7 to the next node 16-5 one by one, thereby the node 16-3 of the file B based on the transfer request 44-2. , 16-5 is completed.

  As is apparent from the time charts of FIGS. 6 and 7, in this embodiment, the transfer destination node can be dynamically specified by designating an arbitrary number of nodes on the network by a transfer request. In addition, by restricting the number of transfers of the same partial file between the file server and the node to one, it is possible to execute the transfer between adjacent nodes by the bucket relay method between the nodes according to the sort order of the IP address, and a plurality of nodes In contrast, network congestion due to 1-to-n transfer that transfers partial files at the same time is avoided, and bucket relay transfer is executed in parallel if transfer is possible between nodes as shown in the time charts of FIGS. As a result, a file can be transferred from a file server to a plurality of nodes at a high speed.

  In the present embodiment, bucket relay transfer between nodes is realized by setting the maximum transferable number between nodes to 1, but if the maximum transferable number is set to an arbitrary integer larger than 1, a partial transfer is possible. 1 to n can be transferred. Such one-to-n transfer can be selected as necessary when the network load is low.

  FIG. 8 is a block diagram showing a hardware environment of a computer that executes the data transfer scheduler program of the present embodiment. 8, the computer constructing the data transfer scheduler 10 in FIG. 1 is connected to the device interface 86 for connecting the ROM 82, RAM 80, hard disk drive 84, keyboard 88, mouse 90, display 92 to the bus 78 of the CPU 76, and further to the network 14. A network adapter 94 to be connected is connected.

  The hard disk drive 84 is installed with a program for realizing the function of the data transfer scheduler 10 shown in FIG. 2 of the present embodiment.

  When the computer is started, the boot program in the ROM 82 is executed through self-diagnosis and initialization by the BIOS, and the OS is read from the hard disk drive 84 to the RAM 80 and executed by the CPU 76. Subsequently, a program for realizing the function of the data transfer scheduler of the present embodiment is read from the hard disk drive 84 by the execution of the OS into the RAM 80 and is executed by the CPU 76.

  The hardware environment of the computer shown in FIG. 8 is basically the same not only for the data transfer scheduler 10 of this embodiment but also for the server that constructs the file server 12 and the nodes 16-1 to 16-5. .

  FIG. 9 is a flowchart showing the data transfer schedule processing in the present embodiment, which will be described below with reference to FIG. In FIG. 9, in the data transfer schedule processing by the data transfer scheduler 10, the transfer request receiving unit 24 checks whether or not a file transfer request is received in step S1, and when the transfer request is received, the transfer request is sorted in step S2. A sort process is performed to generate a file management list that is passed to the unit 34 and manually sorted into node transfer requests.

  Subsequently, in step S3, transfer order management processing for performing transfer processing between adjacent nodes by the bucket relay method in which the number of partial file transfers between the nodes is limited to one time from the file server 12 to the transfer target node according to the file management list. Execute.

  Subsequently, whether or not the file transfer is completed is checked in step S4. When the file transfer is completed, the file management list is deleted in step S5, and the requesting source is notified of the normal transfer end in step S6. The processes in steps S1 to S6 are repeated until a stop instruction such as logoff is issued in step S7.

  FIG. 10 is a flowchart showing details of the sorting process in step S2 of FIG. In FIG. 10, the sort processing is decomposed into node transfer requests for each node, for example, the transfer request 44-1 shown in FIG. 3 received by the sort processing unit 34 from the transfer request receiving unit 24 in step S1.

  Subsequently, the node transfer requests decomposed in step S2 are sorted by file name and node IP address, and in step S3, for example, the node transfer decomposed into a file management list for each file as shown in the file management list 46-1 of FIG. Concatenate requests.

  Subsequently, in step S4, the progress management information 52 is initialized and linked to each of the node transfer requests 56-1 to 56-5 of the file management list as indicated by the node progress management information 60-1 to 60-5. In step S5, the generated file management list is transmitted as an additional event to the transfer order management unit 26 in FIG.

  FIG. 11 is a flowchart showing details of the transfer order management process in step S3 of FIG. In FIG. 11, the transfer order management process is the transfer information provided in the control unit 28 for the transmission request lists 46-1 and 46-2 received from the sort processing unit 34 and stored in the transfer request storage unit 36 of FIG. Processing of the management unit 38, the file transfer request unit 40, and the transfer completion processing unit 42 is executed.

  In FIG. 11, first, at step S1, it is checked whether there is a transfer request input event for a new node, or whether there is a partial file transfer completion event at step S2, and if any event occurs, the process proceeds to step S3. As shown, the progress management information 52 and the transfer count management information 54 generated and managed in correspondence with the file management list 46-1 are updated.

  Subsequently, the determination of the transferable node is checked based on the progress management information 52 and the transfer count management information 54 updated in step S4. If there is a transferable node in step S5 as a result of this discrimination processing, the partial file transfer management unit 32 is requested to transfer the partial file in step S6.

  Simultaneously with this request, in step S7, the current transfer count of the same partial file in the transfer count management information 54 is increased by one and updated. Subsequently, when it is determined that the requested node transfer request is completed in step S8, the completed node transfer request in step S9 is deleted from the file management list.

  FIG. 12 is a flowchart showing details of the transferable node determination process in step S4 of FIG. In FIG. 12, the transferable node determination process reads the partial file progress ratio 72 in the node progress management information 60 shown in FIG. 5B linked to the node in step S1, and the numerator transfer completion number is the denominator subdivision number. If it is determined in step S2 that it is less than that, the node is determined to be a transferable node because all of the partial files have not been transferred, and then the node is linked to the node in step S3 as shown in FIG. The current transfer number 66 of the transfer number management information 54 is read.

  Subsequently, in step S4, it is checked whether or not the current transfer number is less than the maximum transferable number set in the transfer count management information 54, and if it is less than the maximum transferable number, the process proceeds to step S5. Is determined.

  If the transfer completion number has reached the subdivision number in step S2, or the node whose current transfer number has reached the maximum transferable number in step S4, the partial file cannot be transferred, so the transferable node in step S5 The determination process is not performed.

  Subsequently, in step S6, it is checked whether or not the processing has been completed for all the nodes. If the processing has not been completed, the processing returns to step S1, and the same processing is performed for all the nodes arranged in the file management list in step S6. Repeat until the process is determined.

  FIG. 13 to FIG. 28 show the data transfer schedule processing according to the present embodiment in the case where the file is divided into two partial files and transferred to the two nodes 16-1 and 16-2 in 16 stages of steps S1 to S16. It is explanatory drawing divided and shown in detail.

  FIG. 13 shows the process in step S1, and the contents of the transfer request 44 received by the transfer request receiving unit 24 are divided into two partial files X1 and X2 obtained by subdividing the file X into the nodes 16-1 and 16-2 as file servers. The case of transferring from 12 is taken as an example. Here, the host name of the node 16-1 is Host1, and the host name of the node 16-2 is Host2.

  FIG. 14 shows the processing in step S2. The transfer request 44 received by the transfer request receiving unit 24 is sent to the sort processing unit 34, and is decomposed into node transfer requests 100 and 102 for the nodes 16-1 and 16-2. After that, the file management list 96 sorted by the IP addresses of the nodes 16-1 and 16-2 and linked to the entry of “file X” having the file name 98 is generated.

  FIG. 15 shows the processing of the next step S3, and the sort processing unit 34 further initializes and links the progress management information 104 and 106 to the node transfer requests 100 and 102 of the generated file management list 96, respectively.

  FIG. 16 shows the processing of the next step S4, in which the cache position acquisition unit 30 collects a list of held files from the file server 12 and all the nodes 16-1, 16-2, and the file management list of the transfer order management unit 26 As the position information of the file X in 96, file transfer source information 108 indicating that the transfer source is the file server 12 is recorded.

  FIG. 17 shows the processing of the next step S5. The node transfer requests 100 and 102 for the nodes 16-1 and 16-2 provided in the control unit 28 and linked to the file management list 96 are arranged in the order of arrangement (in terms of time). (In order of oldest), it is determined that the partial file X1 in the node transfer request 100 can be transferred from the file server 12 to the node 16-1.

  Based on the determination result, the control unit 28 updates the current transfer number 112 of the transfer number management information 110 corresponding to the node 16-1 from 0 to 1, and then sends the partial file X1 to the subdivision file distribution management unit 32. Requests transfer from the file server 12 to the node 16-1.

  FIG. 18 shows the process in step S6. The segmented file distribution management unit 32 that has received the transfer request in step S5 in FIG. 17 issues a transfer instruction command to the file server 12, and the file server 12 receives the request. The file X1 is transferred to the node 16-1.

  FIG. 19 shows the processing in step S7. When the transfer of the partial file X1 from the file server 12 to the node 16-1 is completed, the segmented file distribution management unit 32 notifies the control unit 28 of the transfer end.

  Upon receiving this transfer end, the control unit 28 reduces the current transfer number 112 in the transfer count management information 110 corresponding to the node 16-1 by 1 to prepare for the transfer of the next partial file. Further, the control unit 28 updates the progress management information 104 linked to the node transfer request 100 corresponding to the node 16-1 from the partial file progress ratio “0/2” before transfer in FIG. 18 to “1/2”. Then, the progress status of the transfer of the first partial file X1 is recorded.

  FIG. 20 shows the processing of the next step S8. The control unit 28 checks the progress management information 104 and 106 linked to the file management list 96 of the transfer order management unit 26 again, and the partial file X2 of the transfer target file X is found. It is determined that transfer from the file server 12 to the node 16-1 is possible, the file transfer management unit 32 is requested to transfer the determined file, and at the same time, the current transfer number 112 of the transfer count management information 110 is set to 0 so far. Update from 1 to 1.

  FIG. 21 shows the processing of the next step S9, in which the control unit 28 checks the progress management information 106 linked to the node transfer request 102 in the file management list 96, and the partial file from the node 16-1 to the node 16-2. It is determined that transfer of X1 is possible, and this determined transfer is requested to the subdivision file distribution management unit 32. At the same time, the control unit 28 updates the current transfer number in the transfer number management information 114 corresponding to the node 16-1 from 0 so far to 1.

  FIG. 22 shows the processing of the next step S10. The segmented file distribution management unit 32 that has received the file transfer request in the processing of step S8 of FIG. 20 and step S9 of FIG. 21 performs the file server 12 and the node 16-1. A transfer instruction command is sent to each of these, and the partial file X2 is transferred from the file server 12 to the node 16-1, and the partial file X1 is transferred from the node 16-1 to the node 16-2.

  FIG. 23 shows the processing in the next step S11. When the transfer of the partial files X1 and X2 in step S10 in FIG. 22 is completed, the segmented file distribution management unit 32 notifies the control unit 28 of the transfer end.

  Upon receipt of this transfer end, the control unit 28 decrements the current transfer count 112 of the transfer count management information 110 corresponding to the file server 12 by 1 and simultaneously sets the current count of the transfer count management information 114 corresponding to the node 16-1. The transfer number 116 is reduced by 1 to zero.

  Furthermore, the control unit 28 updates the partial file progress ratio of the progress management information 104 linked to the node transfer request 110 for the node 16-1 from “1/2” to “2/2”. Further, the partial file progress ratio of the progress management information 106 linked to the node transfer instruction 102 of the node 16-2 is updated from “0/2” to “1/2”.

  FIG. 24 shows the processing of the next step S12. The control unit 28 checks the progress management information 104 and 106 linked to the file management list 96 of the transfer order management unit 26 again, and the node 16-1 to the node 16-2. It is determined that the partial file X2 can be transferred to, and the determined transfer is requested to the subdivision file distribution management unit 32. Further, the current transfer number 116 in the transfer count management information 114 corresponding to the node 16-1 is set. Update from 0 to 1 until then.

  FIG. 25 shows the processing of the next step S13, and the control unit 28 transfers the file X to the node 16-1 from the partial file progress ratio “2/2” of the node progress management information 104 shown in step S12 of FIG. Is deleted, and the node transfer request 100 corresponding to the node 16-1 in FIG. 24 is deleted from the file management list 96.

  At this time, the transfer source information 108 for the file name 98 in the file management list 96 is rewritten to “Host 1” indicating the node 16-1. Then, the transfer request 44 of the transfer request receiving unit 24 records that the transfer of the file X to the node 16-1 is completed.

  FIG. 26 shows the processing of the next step S14. The segmented file distribution management unit 32 that has received the transfer request in step S12 of FIG. 24 issues a transfer instruction command to the node 16-1, and the nodes 16-1 to 16 The partial file X2 is transferred to 16-2.

  FIG. 27 shows the process of the next step S15. When the transfer of the partial file X2 from the node 16-1 to the node 16-2 is completed, the segmented file distribution management unit 32 notifies the control unit 28 of the transfer completion.

  The control unit 28 updates the partial file progress ratio of the progress management information 106 linked to the node transfer request 102 for the node 16-2 from “1/2” so far to “2/2”. Further, the current transfer number 116 of the transfer count management information 114 corresponding to the node 16-2 is decreased by one and returned to zero.

  FIG. 28 shows the processing of the next step S16, and the partial file progress ratio of the progress management information 106 linked to the node transfer request 102 corresponding to the node 16-2 in step S15 of FIG. 27 is “2/2”. Therefore, it is determined that the transfer of the file X to the node 16-2 is completed.

  When it is determined that the transfer of the file X is completed, the file management list 96 is deleted from the transfer order management unit 26, and the transfer request 44 records that the transfer of the file X to the node 16-2 is completed. Finally, the transfer request receiving unit 24 determines that the transfer of the requested file X to the nodes 16-1 and 16-2 has been completed, notifies the normal end of transfer, and ends a series of transfer processes.

  The present invention provides a program executed by the computer of the data transfer scheduler 10 of FIG. This program has the contents shown in the flowcharts of FIGS.

  The present invention also provides a recording medium storing a program executed by a computer having the hardware environment of FIG. 8 that implements the data transfer scheduler 10 of the present embodiment.

  Here, the recording medium is a portable storage medium such as a CD-ROM, a floppy disk (R), a DVD disk, a magneto-optical disk, an IC card, or a storage device such as a hard disk drive provided inside or outside the computer system. , Including a database or other computer system that holds the program via the line, the database, and a transmission medium on the line.

  In the above-described embodiment, a server is taken as an example of a transfer destination node. However, the present invention is not limited to this, and the data transfer from a file server having an appropriate information processing apparatus or device as a node is performed as it is. Applicable.

  In the above embodiment, the file is subdivided into a predetermined fixed-length partial file and transferred. However, if the file size is smaller than the partial file size, the number of partial files is One file is transferred from the file server to a plurality of transfer destination nodes by bucket relay transfer by transfer between adjacent nodes.

  Further, the present invention includes appropriate modifications that do not impair the objects and advantages thereof, and is not limited by the numerical values shown in the above embodiments.

Here, the features of the present invention are enumerated as follows.
(Appendix)

(Appendix 1) (Device)
In a data transfer schedule device that schedules and manages data transfer from a file server to a plurality of nodes,
A transfer request receiving unit that receives a transfer request specifying a file name and a plurality of transfer destination nodes;
After decomposing and sorting the transfer requests into node transfer requests for each transfer destination node, generating a file management list indicating the order of transfer nodes by concatenating the file names, and subdividing according to the order of the file management list A file transfer order management unit that requests file transfer to transfer a partial file between adjacent nodes;
When receiving the file transfer request, a subdivision file distribution management unit that instructs the file server and the node that are the request destination to execute file transfer, and
A data transfer schedule device comprising: (1)

(Appendix 2) (Details of File Transfer Order Management Unit)
In the data transfer schedule device according to attachment 1, the file transfer order management unit restricts the number of transfers to the next node of the same partial file at the node that requested the file transfer to one, and the file management list A data transfer scheduling apparatus for sequentially transferring partial files in the order of arrangement of node transfer requests in the system. (2)

(Appendix 3) (Details of File Transfer Order Management Unit)
In the data transfer schedule device according to attachment 1, the file transfer order management unit includes:
A file management list indicating the order of transfer nodes by concatenating the transfer requests sorted into file names and transfer destination addresses after separating the transfer requests into node transfer requests for each transfer destination node and concatenating the node transfer requests sorted into the file names. A sort processing unit to be generated;
A cache position acquisition unit that collects a list of cached files cached in the file server and all nodes, and detects and sets a cache position of a transfer source file in the file management list;
For each node transfer request arranged in the file management list, progress management information indicating the transfer progress status of the partial file to the next node, and the transfer count that limits the transfer count of the same partial file to the next node to a predetermined maximum transferable number A transfer information management unit that generates and manages management information;
Based on the progress management information and the transfer count management information, determine a transferable node, request the subdivision file distribution management unit to transfer the file, and update the transfer count of the same partial file in the transfer count management information A file transfer requesting unit,
When receiving the file transfer completion notification requested from the subdivided file distribution management unit, the transfer progress status of the partial files described in the progress management information is updated, and the transfer completion of all the partial files is determined. The corresponding node transfer request is deleted from the file management list, and if the transfer count of the same partial file in the transfer count management information reaches the maximum transferable number, the next partial file transfer is performed. A transfer completion processing unit that reduces the number of transfers by one;
A data transfer schedule device comprising: (3)

(Appendix 4) (Details of File Transfer Order Management Unit)
The data transfer schedule device according to appendix 3, wherein the maximum transferable number of the same partial file for the next node managed by the transfer information management unit is set to 1. (4)

(Appendix 5) (Determination of transferable nodes)
In the data transfer schedule device according to attachment 1, the file transfer request unit includes:
Check the progress management information and transfer count management information of each node including the file server linked to the node transfer request arranged in the file management list in the order of arrangement,
Data transfer characterized in that it is determined as a transferable node if the partial file transfer count of the transfer count management information is less than the maximum transferable number and the progress management information does not indicate completion of transfer of all partial files. Schedule device. (5)

(Appendix 6) (Program)
A data transfer scheduling device computer that schedules and manages data transfer from a file server to a plurality of nodes,
A transfer request receiving unit that receives a transfer request specifying a file name and a plurality of transfer destination nodes,
After decomposing and sorting the transfer requests into node transfer requests for each transfer destination node, generating a file management list indicating the order of transfer nodes by concatenating the file names, and subdividing according to the order of the file management list File transfer order management unit that requests file transfer to transfer partial files between adjacent nodes,
When receiving the file transfer request, a subdivided file distribution management unit for instructing and executing file transfer to the file server and node as a request destination,
A program characterized by functioning as (6)

(Appendix 7) (Details of File Transfer Order Management Unit)
In the program according to attachment 6, the file transfer order management unit restricts the number of transfers to the next node of the same partial file at the node that requested the file transfer to one node transfer in the file management list. A program that sequentially transfers partial files in the order of request arrangement.

(Appendix 8) (Details of File Transfer Order Management Unit)
In the program according to attachment 6, the file transfer order management unit includes:
A file management list indicating the order of transfer nodes by concatenating the transfer requests sorted into file names and transfer destination addresses after separating the transfer requests into node transfer requests for each transfer destination node and concatenating the node transfer requests sorted into the file names. A sort processing unit to be generated;
A cache position acquisition unit that collects a list of cached files cached in the file server and all nodes, and detects and sets a cache position of a transfer source file in the file management list;
For each node transfer request arranged in the file management list, progress management information indicating the transfer progress status of the partial file to the next node, and the transfer count that limits the transfer count of the same partial file to the next node to a predetermined maximum transferable number A transfer information management unit that generates and manages management information;
Based on the progress management information and the transfer count management information, determine a transferable node, request the subdivision file distribution management unit to transfer the file, and update the transfer count of the same partial file in the transfer count management information A file transfer requesting unit,
When receiving the file transfer completion notification requested from the subdivided file distribution management unit, the transfer progress status of the partial files described in the progress management information is updated, and the transfer completion of all the partial files is determined. The corresponding node transfer request is deleted from the file management list, and if the transfer count of the same partial file in the transfer count management information reaches the maximum transferable number, the next partial file transfer is performed. A transfer completion processing unit that reduces the number of transfers by one;
A program characterized by comprising:

(Appendix 9) (Details of File Transfer Order Management Unit)
The program according to appendix 8, wherein the maximum transferable number of the same partial file for the next node managed by the transfer information management unit is set to 1.

(Appendix 10) (Determination of transferable node)
In the program according to appendix 6, the file transfer request unit includes:
Check the progress management information and transfer count management information of each node including the file server linked to the node transfer request arranged in the file management list in the order of arrangement,
A program for determining a transferable node if the number of partial file transfers in the transfer count management information is less than the maximum transferable number and the progress management information does not indicate completion of transfer of all partial files.

(Appendix 11) (Recording medium)
A data transfer scheduling device computer that schedules and manages data transfer from a file server to a plurality of nodes,
A transfer request receiving unit that receives a transfer request specifying a file name and a plurality of transfer destination nodes,
After decomposing and sorting the transfer requests into node transfer requests for each transfer destination node, generating a file management list indicating the order of transfer nodes by concatenating the file names, and subdividing according to the order of the file management list File transfer order management unit that requests file transfer to transfer partial files between adjacent nodes,
When receiving the file transfer request, a subdivided file distribution management unit for instructing and executing file transfer to the file server and node as a request destination,
A computer-readable recording medium storing a program that functions as a computer. (7)

(Appendix 12) (Method)
In a data transfer scheduling method for scheduling and managing data transfer from a file server to a plurality of nodes,
A transfer request reception process for receiving a transfer request specifying a file name and a plurality of transfer destination nodes;
After decomposing and sorting the transfer requests into node transfer requests for each transfer destination node, generating a file management list indicating the order of transfer nodes by concatenating the file names, and subdividing according to the order of the file management list A file transfer order management process that requests a file transfer to transfer a partial file between adjacent nodes;
A subdivided file delivery management process for instructing and executing file transfer to the file server and node as a request destination when receiving the file transfer request;
A data transfer schedule method comprising: (8)

(Appendix 13) (Details of file transfer order management process)
The data transfer scheduling method according to appendix 12, wherein the file transfer order management process limits the number of times the same partial file is transferred to the next node at the node that requested the file transfer to one, and the file management list A data data transfer schedule method, wherein partial files are transferred sequentially in the order of arrangement of node transfer requests.

(Appendix 14) (Details of file transfer order management process)
In the data transfer schedule method according to attachment 12, the file transfer order management process includes:
A file management list indicating the order of transfer nodes by concatenating the transfer requests sorted into file names and transfer destination addresses after separating the transfer requests into node transfer requests for each transfer destination node and concatenating the node transfer requests sorted into the file names. Sort process to generate,
A cache location acquisition process that collects a list of retained files cached in the file server and all nodes, and detects and sets a cache location of a transfer source file in the file management list;
For each node transfer request arranged in the file management list, progress management information indicating the transfer progress status of the partial file to the next node, and the transfer count that limits the transfer count of the same partial file to the next node to a predetermined maximum transferable number A transfer information management process for generating and managing management information;
Based on the progress management information and the transfer count management information, determine a transferable node and request the subdivision file distribution management process to transfer the file, and update the transfer count of the same partial file in the transfer count management information File transfer request process to
When the file transfer completion notification requested from the subdivided file distribution management process is received, the transfer progress of the partial file described in the progress management information is updated, and the transfer completion of all the partial files is determined. The corresponding node transfer request is deleted from the file management list, and if the transfer count of the same partial file in the transfer count management information reaches the maximum transferable number, the next partial file transfer is performed. A transfer completion process for reducing the number of transfers by one;
A data transfer schedule method comprising:

(Appendix 15) (Details of file transfer order management process)
15. The data transfer schedule method according to claim 14, wherein the maximum transferable number of the same partial file for the next node managed by the transfer information management process is set to 1.

(Supplementary Note 16) (Determination of transferable node)
In the data transfer schedule method according to attachment 12, the file transfer request process includes:
Check the progress management information and transfer count management information of each node including the file server linked to the node transfer request arranged in the file management list in the order of arrangement,
Data transfer characterized in that it is determined as a transferable node if the partial file transfer count of the transfer count management information is less than the maximum transferable number and the progress management information does not indicate completion of transfer of all partial files. Schedule method.

Explanatory drawing which showed embodiment of the data delivery system provided with the data transfer scheduler of this invention The block diagram which showed the function structure of the data transfer scheduler in this embodiment Explanatory drawing showing transfer requests accepted in this embodiment Explanatory diagram showing a file management list, progress management information, and transfer count management information generated and managed by the data transfer scheduler of this embodiment Explanatory diagram showing the contents of the node transfer request, progress management information and transfer count management information of FIG. Time chart showing how file A in FIG. 2 is subdivided into 6 partial files in 5 nodes and transferred. Time chart showing how file B in FIG. 2 is subdivided into six partial files on two nodes and data transferred A block diagram showing a hardware environment of a computer that executes a program of the data transfer scheduler of the present embodiment The flowchart which showed the data transfer schedule process in this embodiment The flowchart which showed the detail of the sort process in step S2 of FIG. The flowchart which showed the detail of the transfer order management process in step S3 of FIG. The flowchart which showed the detail of the transferable node determination process in step S4 of FIG. Explanatory drawing which showed the process of step S1 which received the transfer request Explanatory drawing which showed the process of step S2 which decomposes | disassembles a transfer request and produces | generates a file management list Explanatory drawing which showed the process of step S3 which initializes and links progress management information to a file management list Explanatory drawing which showed the process of step S4 which acquires a file server and the possession file of all the nodes, and records a file position on a file management list Explanatory drawing of the process of step S5 which checks the transferable node and requests the first node to transfer the partial file X1 from the file server. Explanatory drawing of the process of step S6 which transfers the partial file X1 from the file server to the first node. Explanatory drawing which showed the process of step S7 which updates transfer order information and progress management information with the completion of transfer of the partial file X1. Explanatory drawing of the process of step S8 which checks the transferable node and requests the first node to transfer the next partial file X2 from the file server. Explanatory drawing of the process of step S9 which checks the transferable node and requests the transfer of the partial file X1 from the first node to the next node. Explanatory drawing of the process of step S10 which transfers partial file X1, X2 based on a transfer request Explanatory drawing which showed the process of step S11 which updates transfer order information and progress management information with the completion of transfer of partial files X1, X2 Explanatory drawing of the process of step S12 which checks the transferable node and requests transfer of the partial file X2 from the first node to the next node Explanatory drawing which showed the process of step S13 which discriminates the completion of transfer to the first node and deletes the corresponding node transfer request from the file management list Explanatory drawing of the process of step S14 which transfers the partial file X2 based on a transfer request Explanatory drawing which showed the process of step S15 which updates transfer order information and progress management information with the completion of transfer of partial file X2 Explanatory diagram showing the processing of step S16 for notifying completion of transfer to the request source after determining completion of transfer to the next node and deleting the corresponding node transfer request from the file management list

Explanation of symbols

10: Data transfer scheduler 12: File server 14: Networks 16-1 to 16-5: Nodes 18-1 to 18-5: Data transfer agents 20-1 to 20-5: Transfer caches 22-1 to 22-5: Storage 24: Transfer request receiving unit 26: Transfer order management unit 28: Control unit 30: Cache position acquisition unit 32: Subdivided file distribution management unit 34: Sort processing unit 36: Transfer request storage unit 38: Transfer information management unit 40: File transfer request unit 42: Transfer completion processing unit 44, 44-1, 44-2: Transfer request 46-1, 46-2, 96: File management list 48, 54, 78, 98: File name 50: Transfer destination node Groups 52, 104, 106: Progress management information 54, 110, 112: Transfer count management information 56-1 to 56-5, 100, 102: Node transfer request 58 Transfer source file servers 60, 60-1 to 60-5, 104, 106: Node progress management information 62, 62-1 to 62-5: Node names 64, 64-1 to 64-5: Maximum transferable number 66, 66-1 to 66-5, 112, 116: Current transfer number 68: File name 70: Transfer destination node name 72: Partial file progress ratio 74: Node name 76: CPU
78: Bus 80: RAM
82: ROM
84: Hard disk drive 86: Device interface 88: Keyboard 90: Mouse 92: Display 94: Network adapter

Claims (8)

In a data transfer schedule device that schedules and manages data transfer from a file server to a plurality of nodes,
A transfer request receiving unit that receives a transfer request specifying a file name and a plurality of transfer destination nodes;
After decomposing and sorting the transfer requests into node transfer requests for each transfer destination node, generating a file management list indicating the order of transfer nodes by concatenating the file names, and subdividing according to the order of the file management list A file transfer order management unit that requests file transfer to transfer a partial file between adjacent nodes;
When receiving the file transfer request, a subdivision file distribution management unit that instructs the file server and the node that are the request destination to execute file transfer, and
A data transfer schedule device comprising:
2. The data transfer schedule device according to claim 1, wherein the file transfer order management unit limits the number of transfers to the next node of the same partial file at the node that requested the file transfer to one time. A data transfer schedule device for transferring partial files sequentially in the order of arrangement of node transfer requests in a list.
The data transfer schedule device according to claim 1, wherein the file transfer order management unit includes:
A file management list indicating the order of transfer nodes by concatenating the transfer requests sorted into file names and transfer destination addresses after separating the transfer requests into node transfer requests for each transfer destination node and concatenating the node transfer requests sorted into the file names. A sort processing unit to be generated;
A cache position acquisition unit that collects a list of cached files cached in the file server and all nodes, and detects and sets a cache position of a transfer source file in the file management list;
For each node transfer request arranged in the file management list, progress management information indicating the transfer progress status of the partial file to the next node, and the transfer count that limits the transfer count of the same partial file to the next node to a predetermined maximum transferable number A transfer information management unit that generates and manages management information;
Based on the progress management information and the transfer count management information, determine a transferable node, request the subdivision file distribution management unit to transfer the file, and update the transfer count of the same partial file in the transfer count management information A file transfer requesting unit,
When receiving the file transfer completion notification requested from the subdivided file distribution management unit, the transfer progress status of the partial files described in the progress management information is updated, and the transfer completion of all the partial files is determined. The corresponding node transfer request is deleted from the file management list, and if the transfer count of the same partial file in the transfer count management information reaches the maximum transferable number, the next partial file transfer is performed. A transfer completion processing unit that reduces the number of transfers by one;
A data transfer schedule device comprising:
4. The data transfer schedule device according to claim 3, wherein the maximum transferable number of the same partial file for the next node managed by the transfer information management unit is set to 1.
The data transfer schedule device according to claim 1, wherein the file transfer request unit includes:
Check the progress management information and transfer count management information of each node including the file server linked to the node transfer request arranged in the file management list in the order of arrangement,
Data transfer characterized in that it is determined as a transferable node if the partial file transfer count of the transfer count management information is less than the maximum transferable number and the progress management information does not indicate completion of transfer of all partial files. Schedule device.
A data transfer scheduling device computer that schedules and manages data transfer from a file server to a plurality of nodes,
A transfer request receiving unit that receives a transfer request specifying a file name and a plurality of transfer destination nodes,
After decomposing and sorting the transfer requests into node transfer requests for each transfer destination node, generating a file management list indicating the order of transfer nodes by concatenating the file names, and subdividing according to the order of the file management list File transfer order management unit that requests file transfer to transfer partial files between adjacent nodes,
When receiving the file transfer request, a subdivided file distribution management unit for instructing and executing the file transfer to the file server and the node as the request destination,
A program characterized by functioning as
A data transfer scheduling device computer that schedules and manages data transfer from a file server to a plurality of nodes,
A transfer request receiving unit that receives a transfer request specifying a file name and a plurality of transfer destination nodes,
After decomposing and sorting the transfer requests into node transfer requests for each transfer destination node, generating a file management list indicating the order of transfer nodes by concatenating the file names, and subdividing according to the order of the file management list File transfer order management unit that requests file transfer to transfer partial files between adjacent nodes,
When receiving the file transfer request, a subdivided file distribution management unit for instructing and executing file transfer to the file server and node as a request destination,
A computer-readable recording medium storing a program that functions as a computer.
In a data transfer scheduling method for scheduling and managing data transfer from a file server to a plurality of nodes,
A transfer request reception process for receiving a transfer request specifying a file name and a plurality of transfer destination nodes;
After decomposing and sorting the transfer requests into node transfer requests for each transfer destination node, generating a file management list indicating the order of transfer nodes by concatenating the file names, and subdividing according to the order of the file management list A file transfer order management process that requests a file transfer to transfer a partial file between adjacent nodes;
A subdivided file delivery management process for instructing and executing file transfer to the file server and node as a request destination when receiving the file transfer request;
A data transfer schedule method comprising:

Images