JP2008152792A - Data transfer processor and data transfer system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To obtain a data transfer processor and a data transfer system improving data transfer efficiency to the whole processing device and achieving highly reliable data transfer. <P>SOLUTION: This system comprises a division processing means 2 which determines a processor at a data transfer destination based on processor group list data associating the processor connected via a communication line with index data concerning the throughput of each processor and which divides the processor group list data and allocates them to the determined processor at each transfer destination, and a transfer means 4 which transfers the data to be transferred to the processor at each transfer destination determined by the division processing means 2 and the processor group list data allocated to the processor at each transfer destination by the division processing means 2. When the data transfer completed normally, the transfer means 4 notifies the processor at higher hierarchy that the data transfer completed normally. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a data transfer processing device and a data transfer system for transferring data at high speed from one processing device to another processing device in an environment where a large number of processing devices such as a distributed processing system can use communication means such as a network. .

  A conventional data transfer system is a data transfer system that expands data transfer as shown by a tree structure to which a mouse calculation is applied. In this data transfer method, the processing data at the apex of the tree structure, together with the same data to be transferred to other processing devices, the recognition code group of the processing device to be transferred (individual processing devices can be identified) The data is transferred by adding the data of the code), and the transfer is developed while capturing the processing device that has not completed the transfer with the data of the recognition code group of the processing device (for example, patent document) 1).

  A specific data transfer method is realized by each processing device operating with the same algorithm as described below. That is, each processing device divides the list of recognition code groups out of the received data and the list of recognition code groups of the processing device by the number n or n + 1 of communication lines that the processing device can open simultaneously. .

  Furthermore, one processing device to be transferred is selected from each of the divided n or n + 1 recognition code group lists. Then, n communication lines connected to the n selected processing devices are opened, and data is transferred together with the recognition code group of the processing device including the selected processing device to each selected processing device. Thereby, when transferring the same data to many processing apparatuses, the data transfer to all the processing apparatuses can be completed in a short time.

Japanese Patent No. 3302769 (page 5, FIG. 2)

  However, the prior art has the following problems. In the prior art, when dividing the list of recognition code groups indicating the processing device, no conditions are obtained, and the criteria and method for division are not disclosed. Further, it does not describe a method for determining a processing device to be transferred from a list of recognition code groups divided into n or n + 1.

  That is, in the prior art, since the capacity and connection environment of the processing apparatus and the current load status are not considered, the processing apparatus with a low capacity, the processing apparatus connected to the dial line, or the processing apparatus with a high workload. May be selected as the next transfer destination processing apparatus. In such a case, due to the low processing capacity, for example, it is possible to easily imagine a specific situation where only one line can be opened or data cannot be transferred easily (data transfer speed is slow). As a result, the efficiency of shortening the transfer time by distributed processing decreases.

  The present invention has been made to solve the above-described problems, and provides a data transfer processing device and a data transfer system capable of improving data transfer efficiency to the entire processing device and realizing highly reliable data transfer. Objective.

The data transfer processing device according to the present invention determines a processing device as a data transfer destination based on processing device group list data in which processing devices connected by a communication line and index data relating to the processing capability of each processing device are associated with each other. The division processing means to be assigned to each transfer destination processing device determined by dividing the processing device group list data, and the data to be transferred to each transfer destination processing device determined by this division processing means and the division processing means A transfer means for transferring the processing device group list data assigned to each transfer destination processing device, and the transfer means indicates that the data transfer has been normally completed when the data transfer has been normally completed. To the processing device.
Further, the data transfer system according to the present invention is a data transfer system for transferring data to a plurality of processing devices. Each of the plurality of processing devices is index data relating to the processing capability of the processing device for each processing device to be retransferred For each processing device to be retransmitted from the index data included in the processing device group list data received by the receiving means, and receiving means for receiving the processing device group list data associated with Division processing means for obtaining division index data and determining n (n is an integer of 2 or more) processing devices and n post-division processing device group list data from the processing device group list data based on the division index data; A connection means for connecting to n processing devices determined by the division processing means using n communication lines, and a group of n post-division processing devices determined by the division processing means. A transfer unit that adds data to be retransferred to each of the data and transfers the data to each of the n processing devices connected by the connection unit, and the receiving unit receives data from the processing device of the transmission source After the transfer is normally completed, a data transfer completion list including an identifier indicating the transmission source processing device is received from the transmission source processing device, and the transfer unit transfers data to each of the n processing devices. And monitoring whether or not the data communication is normally completed based on the responses from the n processing devices. When the data transfer to the n processing devices is normally completed, an identifier indicating the own processing device Is added to update the data transfer completion list, and the updated data transfer completion list is transferred to n processing devices.

  According to the present invention, a processing device to which data transfer is to be performed is determined based on index data related to processing capability from among unreceived processing devices, and a list of processing device groups is divided based on the index data. It is possible to obtain a data transfer processing device and a data transfer system capable of improving data transfer efficiency to the entire device and realizing highly reliable data transfer.

  Hereinafter, preferred embodiments of a data transfer processing device and a data transfer system of the present invention will be described with reference to the drawings.

Embodiment 1 FIG.
FIG. 1 is a configuration diagram of a processing apparatus used in the data transfer system according to the first embodiment of the present invention. The data transfer system according to the present invention efficiently distributes one common transfer data to a plurality of processing devices, and each processing device has the basic configuration shown in FIG.

  The processing device takes in the processing device group list data as input data together with one common transfer data. The processing device group list data is obtained by associating an identification code for identifying a processing device to be transferred with index data relating to the processing capability of each processing device. Specifically, the index data includes index data related to the processing capability of each processing device, such as CPU capability, memory load, network bandwidth allocated to communication, operation status, and the number of lines that can be communicated simultaneously.

  The processing device divides the processing device group list data included in the input data, thereby determining the processing device group list data corresponding to the processing device of the transfer destination and the processing device. Then, the processing device generates data in which common transfer data is added to the processing device group list data corresponding to each processing device group as output data for the processing device group divided into n pieces, and n processing is performed. The generated output data is transferred to n processing devices selected from each of the device groups based on the index data.

  As described above, when the processing apparatus according to the present invention performs the division, the processing apparatus performs processing based not only on the identification code described in the processing apparatus group list data but also on the index data attached to the identification code. The overall transfer efficiency is improved.

  Next, the internal configuration of the processing apparatus of the present invention will be described with reference to FIG. Each processing apparatus includes a reception unit 1, a division processing unit 2, a connection unit 3, and a transfer unit 4. The receiving means 1 is means for receiving transfer data and processing device group list data as input data from another device connected by a network or the like or a processing device in an upper layer.

  The division processing unit 2 obtains division index data serving as an index for performing division from the index data included in the processing apparatus group list data, and based on the division index data, the transfer destination processing apparatus and the corresponding divided processing apparatus Group list data is generated, and output data to which transfer data is added is generated.

  Here, the division index data can be obtained from the items described in the index data, and adopts one item of the index data or data obtained by combining a plurality of items included in the index data. You can also

  The connection unit 3 connects to the transfer destination processing device determined by the division processing unit 2 in accordance with the number of lines through which the processing device can simultaneously communicate. Further, the transfer unit 4 transmits the output data generated by the division processing unit 2 corresponding to each transfer destination to the processing device at each transfer destination.

  In this way, the processing device that has received the output data from the higher-layer processing device similarly performs the division process, transfers it to the lower-layer processing device, and repeats this processing, whereby the original processing device group Data transfer to all the processing devices included in the list data is performed.

  Next, the flow of processing as the entire data transfer system will be described. FIG. 2 is a diagram showing a tree structure of the divided processing devices in the data transfer system according to the first embodiment of the present invention. In FIG. 2, the processing device 10 in the first layer is a processing device that first receives a list of data and processing device groups. The second tier processing devices 20, 30, and 40 are processing devices that the first tier processing device 10 has determined as the next processing device to transfer data. Similarly, the data is divided and transferred to the processing devices in the third and fourth layers.

  Each processing apparatus is indicated by a square divided into an upper stage and a lower stage, and the upper stage shows an identification code including division index data of the processing apparatus. On the other hand, the lower part shows a list of processing device groups to be retransmitted, and is similarly shown as a list of identification codes including division index data. Further, a broken-line square drawn at the end of the broken-line arrow indicates the same processing apparatus that repeatedly performs the transfer process after the data transfer through the communication line is completed.

  For example, in the tree structure of FIG. 2, the broken line processing device 10 at the right end of the second hierarchy moves to the processing device group corresponding to the remaining identification codes after the data transfer to the processing devices 20, 30, and 40 is completed. The processing apparatus 10 in the first hierarchy that performs the data transfer is shown. Therefore, the broken line processing apparatus is physically the same as the solid line processing apparatus of the same number. In addition, processing devices in the same hierarchy indicate processing devices that are simultaneously working in time.

  FIG. 3 is a diagram showing an example of a list of processing apparatus groups according to the first embodiment of the present invention. The identification code given to each processing device is a code that can uniquely identify the processing device, such as a host name, an IP address, a MAC address, and a telephone number. In addition, as index data corresponding to each identification code, the CPU performance of the processing device, the amount of installed memory, the network bandwidth that can be used for data transfer, the network distance from a reference point such as the processing device that receives data first, A summary of the load status of each processing apparatus that dynamically changes is included.

  The basis of the present invention is to grasp the concept indicating the performance, function, positional relationship, and the like of such a processing apparatus as index data, determine division index data according to the conditions, and determine the processing apparatus to be transferred next. There is a concept. Therefore, in the present invention, in addition to the index data corresponding to each identification code illustrated in FIG. 3, for example, position information by GPS, relative position in a building, or priority order determined by arbitrary conditions (each process An arbitrary weight on the apparatus) can be included as index data.

  FIG. 4 is an example of an identification code of each processing apparatus according to Embodiment 1 of the present invention, showing an example in which the list of processing apparatuses is simplified based on the above-described conditions and the data amount is reduced. Yes.

  Just because the processing capacity of the processing device is high does not necessarily mean that many communication lines can be opened simultaneously. For example, even if the processing performance indicated by the CPU performance, the amount of installed memory, etc. is the highest level, if there is only one thin telephone line, the lowest level in this data transfer according to one thin telephone line It may be positioned as a processing device.

  For simplification of explanation, in FIG. 4, as a result of ranking the division index data for 26 processing devices a to z, each processing device is classified into three ranks A, B, and C. A list of identification codes when arranged so as to belong to any of the above is shown.

  Further, as the division index data in this example, class A indicates a processing device that can open only one communication line at a time, class B indicates a processing device that can open only two lines, and class C opens three lines. A processing device shall be indicated. In FIG. 4, the classes A to C, which are the division index data, are attached after each of the 26 processing devices a to z to form a list of identification code groups including the division index data. Each processing apparatus in the tree structure shown in FIG. 2 is configured by the group of 26 processing apparatuses shown in FIG. 4, and the following description will be made with this configuration.

  Next, a specific process for transferring data to each layer will be described. FIG. 5 is a flowchart showing a series of data transfer processes in the processing devices of each hierarchy according to the first embodiment of the present invention. This operation is common to the processing devices that transmit and receive data, but will be described below using the processing device 10 in the first hierarchy as an example.

  First, the receiving means 1 of the processing device 10 receives transfer data and processing device group list data as input data. Next, the division processing means 2 of the processing device 10 compares the number n of communication lines that the processing device 10 can open simultaneously with the number of processing devices included in the processing device group list data (step S501).

  Here, the identification code of the processing device 10 is aC, which is a processing device that opens three lines. In addition, the number of processing devices included in the processing device group list data is 25 other than bC to zA. Therefore, the number of processing devices is larger than the number n of communication lines.

  Therefore, the division processing means 2 preferentially determines the processing devices having a large division index data as n processing devices to be transferred from the processing device group list data. Specifically, the division processing means 2 determines the processing device 20 (corresponding to the identification code bC) and the processing device 30 (corresponding to the identification code cC) of the C rank, and further, the processing device 40 (identification) from the B rank. 3 corresponding to the number of communication lines that can be opened simultaneously by the processing device 10 (step S502).

  Next, the division processing unit 2 calculates the remaining number of units obtained by excluding the three processing devices 20, 30, and 40 from the 25 processing devices included in the processing device group list data as 22 units. Further, the total number of communication lines opened simultaneously by the processing apparatus 20, the processing apparatus 30, and the processing apparatus 40 is calculated as 3 + 3 + 2 = 8 lines. Then, the division processing means 2 compares the remaining number of processing devices 22 with the total number of lines 8. (Step S503)

  As a result, since the number of processing devices is larger, the division processing means 2 adds m to 3 + 1 = 4 by adding the number of divisions m to the number of lines 3 that can be communicated at the same time. That is, the division of the processing device group is performed for four processing devices 10, the processing device 20, the processing device 30, and the processing device 40 (step S504).

  Here, the division method (detailed processing in step S504) will be described with reference to FIG. FIG. 6 is a flowchart showing a series of division processing in the processing devices of each hierarchy according to Embodiment 1 of the present invention. More specifically, when the processing device 10 divides the processing device included in the processing device group list data into the division number m, the processing device 10 includes the processing devices included in the divided processing device group list data. The total of the division index data and the total number of units are set to the same ratio as the ratio of the division index data of the determined m processing apparatuses.

  First, the division processing means 2 of the processing device 10 performs division processing based on the number of lines opened simultaneously corresponding to the division index data of each processing device. That is, since the division index data A is the number of lines 1 that can be opened simultaneously, B is 2, and C is 3, the ratio of the division index data of the processing device 10, the processing device 20, the processing device 30, and the processing device 40 is obtained. It can be seen that C: C: C: B = 3: 3: 3: 2 (step S601).

  Then, in order to divide the remaining 22 units included in the processing device group at this ratio, the division processing means 2 can be divided into 6: 6: 6: 4 units when applied to the ratio ( Step S602). Here, when too many units are generated, the division processing means 2 performs distribution in order from the processing device having the highest division index data.

  Next, the division processing means 2 divides the remaining 22 processing device groups according to the division index data, so that B is 10 units and A is 12 units (step S603). Therefore, the division processing means 2 divides the 10 B, which has high processing capability as the division index data, so that the ratio becomes the ratio of the above processing devices 10 to 40. As a result, the processing devices 10 to 30 have 3 Two or two units and the processing apparatus 40 are divided into two units (step S604). In the example of FIG. 2, two units are distributed to the processing device 30 and three units are distributed to the processing devices 10 and 20.

  Next, the division processing unit 2 determines whether or not there is still a processing device group to be divided for each division index data after the division of the ten units having high processing capability B (step S605). Specifically, the division processing means 2 can determine that 12 A rank ranks still remain, and the division for 12 units will be performed next.

  Furthermore, the division processing means 2 determines whether or not the current division ratio matches the ratio of the division index data (step S606). Specifically, the processing device 10 divides the ratio B of the processing devices of rank B divided in step S604 and the ratio 3: 3: 3: 2 of the division index data of m processing devices obtained in the previous step S602. It is determined whether or not the ratio matches.

  At present, three B-rank processing devices are distributed to the processing devices 10 and 20, whereas only two B-rank processing devices are distributed to the processing device 30. Since one unit is insufficient, the division processing unit 2 applies one from the rank A to the processing device 30 (step S607).

  Then, the division processing unit 2 repeats the process of step S604 for the 11 A ranks remaining without being distributed so that the ratio of the division index data becomes 3: 3: 3: 2. Distribute m processing units. When the distribution of the rank A processing devices is completed, there is no processing device group to be divided, and the division processing unit 2 ends the series of distribution processing (step S605).

  The result of the division by the distribution method shown in FIG. 6 is a processing device group described in the lower part of each square of the processing devices 10, 20, 30, and 40 in the second hierarchy in FIG. Looking at the total and total number of division index data of each processing device group, regarding the processing device 20, the total of the division index data is 9 and the total number is 6, and for the processing device 30, the total of the division index data is 8. The total number is 6, the total number of division index data is 6 for the processing device 40, the total number is 4, and the total number of division index data is 9 for the processing device 10 indicated by the broken line.

  In other words, the total number is the same as the capacity ratio of 3: 3: 3: 2, but the total of the division index data is 9: 9: 8: 6, and the ratio is 3: 3: 3: 2. Are not exactly the same. However, since the total and total number of division index data are not necessarily divisible, there is a case where the ratio of the division index data is 3: 3: 3: 2, but the distribution is performed according to the flowchart of FIG. Thus, an even distribution close to the ratio of the number of lines of processing devices can be performed.

  When the division of the processing device group in step S504 is completed in this way, the connection means 3 of the processing device 10 opens three communication lines and connects to the processing device 20, the processing device 30, and the processing device 40. Then, the transfer unit 4 transmits the processing device group list data divided together with the data to be transferred to the processing device 20, the processing device 30, and the processing device 40 (step S506).

  When the data transmission is completed, the division processing unit 2 checks whether there is a list of untransmitted processing device groups (step S507). Specifically, in this example, a list of processing device groups (hB, 1B, nB, pA, tA, and xA) divided for the processing device 10 indicated by the broken line in the second hierarchy in FIG. List).

  Therefore, the division processing means 2 returns to the first processing in the flowchart of FIG. 5 and repeats a series of processing in order to perform the division processing of the processing device group assigned to its own processing device 10. This processing corresponds to the division processing of the processing device group assigned to itself corresponding to the (n + 1) th processing device after the division into n + 1.

  Specifically, the division processing unit 2 compares the number of processing devices 6 included in the processing device group with the number of simultaneously communicable lines 3 (step S501), and the number of processing devices is larger. From the list of devices, the B rank processing devices 11, 12 and 13 are determined as the next transfer destination processing devices (step S502). Next, the division processing means 2 compares three units obtained by subtracting the already determined three processing devices from the six processing device groups with the total number of simultaneous communication lines 6 of the three units (step S503).

  As a result, since the number of simultaneous total communication lines is larger, the division processing unit 2 determines that the division number may be m = n. Then, the division processing unit 2 divides the remaining processing device group, but since all of the three transfer destination processing devices have the same division index data, the list of the remaining processing device groups is divided into the division index data. Are distributed so that the total number is equal and the total number is also equal. As a result, the division processing unit 2 divides the three units remaining in the list of processing units into one unit for each of the transfer destination processing units 11, 12, and 13 to be transferred this time (step S505).

  Furthermore, the connection means 3 and the transfer means 4 transfer the data to be transferred to the processing devices 11, 12, 13 and the output data with a list of the processing device groups assigned to each of them (step S506). . Finally, the division processing unit 2 confirms whether or not the processing device group for the processing device 10 remains, and completes the processing of the processing device 10 because it does not remain (step S507).

  As described above, according to the first embodiment, when there is a difference in performance such as CPU performance, memory load, and network type in the processing device, according to the division index data based on the index data viewed from these various viewpoints, By deciding the processing device to be transferred and dividing the list of processing device groups, it is possible to sequentially transmit the processing devices from the processing devices having high ability determined as the division index data from various viewpoints.

  As a result, for example, if the division index data is a communication line, data transfer can be performed deterministically from a processing device that opens more communication lines. It becomes possible to increase. Accordingly, the number of distribution layers is reduced as compared with the prior art, and the time required for completion of transfer to all the processing devices is shortened, thereby realizing highly reliable data transfer.

Embodiment 2. FIG.
In the first embodiment, first, n processing devices to be transferred are determined based on the division index data, and then the processing device group is distributed according to the ratio of the division index data of the n processing devices. Explained when to do. In the second embodiment, first, after the processing device group is divided into n so that the sum of the division index data is equal, the processing device to be transferred is determined from each of the n processing device groups. Will be described.

  FIG. 7 is a diagram showing a tree structure of the divided processing devices in the data transfer system according to the second embodiment of the present invention. The processing apparatus numbers quoted in FIG. 7 are the same as those in FIG. That is, in both figures, the same processing apparatus number is the same, and both figures show the differences in the same system according to the embodiment.

  Next, a specific process for transferring data to each layer will be described. FIG. 8 is a flow chart showing a series of data transfer processes in the processing devices of each hierarchy according to the second embodiment of the present invention. This operation is common to the processing devices that transmit and receive data, but will be described below using the processing device 10 in the first hierarchy as an example.

  First, the receiving means 1 of the processing device 10 receives transfer data and processing device group list data as input data. Next, the division processing means 2 of the processing device 10 compares the number n of communication lines that the processing device 10 can open simultaneously with the number of processing devices included in the processing device group list data (step S801).

  When the number of processing devices is larger than the number n of communication lines, the division processing unit 2 divides the processing device group into division numbers m (m is n or n + 1). At this time, the division is performed so that the total and total number of the division index data of each of the divided processing apparatus groups are equal (step S802).

  The division method may be obtained so that the ratio of the division index data of the processing device is 1: 1: 1: 1 in the division method described above with reference to FIG. However, unlike Embodiment 1, in this Embodiment 2, since the processing apparatus which should be transferred at present is not determined, it should divide equally with respect to 25 units of all the processing apparatus groups bC-zA. In addition, since it is divided into n + 1 in this example, it should be noted that the processing device group including 26 units including aC must be divided.

  The result of the division is as shown in the processing devices 20, 30, 40 and the dashed processing device 10 in FIG. 7. That is, the processing device of the second hierarchy includes a processing device group including the processing device 10 in which the total of the division index data is 12 and the total number is 7, and a processing device in which the total of the division index data is 11 and the total number is 7. A processing device group including 20, a processing device group including a processing device 30 in which the total of the division index data is 10 and the total number is 6, and a processing device 40 in which the total of the division index data is 10 and the total number is 6. It is divided into four processing device groups.

  The division processing means 2 determines the processing device having the highest ability determined as the division index data from among these four processing device groups as the processing device to be transferred next (step S803). In this example, the division processing means 2 determines that the processing devices 20, 30, and 40 are the next processing devices corresponding to the three communication lines.

  Further, the connection means 3 opens three communication lines to connect to the processing devices 20, 30, 40, and the transfer means 4 together with the processing device group list data after dividing the processing devices 20, 30, 40. The transfer data is transferred (step S804).

  In the first embodiment, first, n processing devices to be transferred are determined based on the division index data, and then the processing device group is distributed according to the ratio of the division index data of the n processing devices. In order to do so, the division processing shown in FIGS. 5 and 6 is required. On the other hand, in the second embodiment, after the processing device group is divided into n so that the total of the division index data is equal, the processing device to be transferred is determined from each of the n processing device groups. Therefore, only the simple division process shown in FIG. 8 is required.

  As described above, according to the second embodiment, although the disturbance of the branches and leaves of the tree structure occurs as compared with the first embodiment, the processing device group can be simply divided by a dramatically simplified algorithm. it can. Thereby, the data transfer efficiency to the whole processing apparatus can be improved, and highly reliable data transfer can be realized. Such division processing can be expected to be applied to a resource-limited environment such as a data receiver, an STB (Set-top Box), and a digital home appliance.

Embodiment 3 FIG.
In the first and second embodiments, a series of processing based on the assumption that data transfer is normally performed has been described. In the third embodiment, a data transfer system that can cope with a case where a failure occurs in a transfer destination processing apparatus by monitoring the data transfer status by the transfer source processing apparatus will be described.

  FIG. 9 is an explanatory diagram of an operation when a failure occurs during data transfer according to the third embodiment of the present invention. Specifically, a coping operation is shown when a failure occurs in the data transfer destination processing device b while the data transfer source processing device a is transferring data. The transfer means 4 of the processing device a monitors whether or not the data transfer has been normally completed based on a response from the transfer destination processing device b.

  Specifically, the transfer means 4 of the processing device a monitors whether or not a failure occurs during data transfer by monitoring whether or not a response from the data transfer destination processing device b is returned within a predetermined time. It can be detected. Further, the processing device group list data transferred to the processing device b is originally generated by the processing device a, and the processing device a may hold the processing device group list data transferred to the processing device b. Is possible.

  When a failure occurs in the processing apparatus b under such circumstances, the division processing means 2 of the processing apparatus a uses the processing apparatus group list data sent to the processing apparatus b as division index data other than the processing apparatus b. The processing device c having the highest capability is selected. Then, the connection means 3 of the processing device a switches and connects the communication line from the processing device b to the processing device c.

  Further, the transfer means 4 of the processing device a can transfer data to the processing device c together with the processing device group list data including the processing device b. At this time, it is still effective to inform the subsequent processing devices by describing in the processing device group list data that a failure has occurred in the processing device b.

  Furthermore, if the content of the processing device group is only the processing device b in the terminal processing device, and the failure of the processing device b continues, the processing device b gives up data transfer as being incapable of data transfer. .

  As described above, according to the third embodiment, when a problem occurs in a data transfer destination processing device during data transfer, the processing device can be skipped and data transfer can be continued. Data transfer never stops. In addition, the data transfer system has an advantage that a processing apparatus to which data cannot be transferred can be definitely found. Regarding these, after recovering from a failure, data transfer may be performed to the failed processing device group in the same manner.

  Although a failure has occurred in the third embodiment, this may be a case where the power is simply cut off in an STB, a receiver, a digital home appliance, or the like.

Embodiment 4 FIG.
In the third embodiment, the data transfer system has been described in which a failure occurs in a data transfer destination processing apparatus during data transfer. In the fourth embodiment, a data transfer system corresponding to a case where a failure occurs in a data transmission source processing apparatus during data transfer will be described.

  FIG. 10 is an explanatory diagram of an operation when a failure occurs during data transfer according to the fourth embodiment of the present invention. Specifically, the processing device that is the transmission source of the data while the processing device a completes the data transfer to the processing device b and the processing device b is transferring the data to the processing device c. b shows a case where a failure occurs.

  The transfer means 4 of the processing device a monitors the situation where the processing device b is transferring data to the remaining processing device group even after the data transmission to the processing device b is completed. Specifically, the processing apparatus b that is the data transfer destination appropriately notifies the processing apparatus a of how much data has been transferred normally. On the other hand, the transfer means 4 of the processing device a can detect the presence or absence of a failure during data transfer by monitoring the transfer status of the processing device b.

  When a failure occurs in the processing device b, the transfer means 4 of the processing device a only has a shortage of remaining data with respect to the processing device c that is being transmitted by the processing device b based on the transfer status monitored. Can be sent. As a result, the normal processing device c can continue to distribute to the remaining processing device group.

  As described above, according to the fourth embodiment, a failure has occurred not only in the data receiving destination processing device during communication as in the third embodiment, but also in the data transmission source processing device in the central channel. Even in this case, it is possible to continue data transfer as the entire data transfer system. In this case, since the processing device b has already received the data, the data transfer system needs to handle the data reception of the processing device b as complete.

Embodiment 5. FIG.
In the first to fourth embodiments, description has been given of the case where the number of communication lines is ranked as the division index data in consideration of the CPU clock, the memory mounting amount, the connection network, and the like. However, other than these index data, weighting may be performed so that a processing device that is desired to complete transfer earlier may be placed at the top of the tree structure, and a processing device that may be delayed in reception may be placed behind the tree structure.

  Therefore, in such a case, a numerical value such as a priority and a code indicating the priority can be easily set as index data instead of the communication line. The division processing means 2 can prioritize data transfer to a processing device having a high priority by performing the division processing based on the division index data with priority added.

  As described above, according to the fifth embodiment, the transfer according to the priority can be realized by giving each processing apparatus the priority in the transfer order. For example, when considering the distribution of digital video, some people want it as soon as possible, and some people can do it any time within one week. In view of such circumstances, it is possible to realize a data transfer system having practical convenience in accordance with a sales form such that a person who wants within one hour responds with an additional charge.

Embodiment 6 FIG.
In the first to fifth embodiments, the mode in which data is transferred in one direction from the upper layer processing device to the lower layer processing device has been described. In the sixth embodiment, a method will be described in which each processing device in the second and subsequent layers notifies the data processing result to the processing device 10 in the first layer that is first input.

  FIG. 11 is a diagram showing a tree structure of divided processing devices and a data transfer completion list from each processing device in the data transfer system according to the sixth embodiment of the present invention. After all data transfer is completed, each processing device in the second layer and later returns a data transfer completion list to the upper layer processing device as a notification of transfer completion.

  The tree structure of each processing apparatus having four layers in FIG. 11 is the same as the tree structure of FIG. 2, and a method for returning the result of data transfer will be described based on this tree structure. In FIG. 11, the data transfer completion list is shown as a1 to x1.

  First, the fourth-layer processing devices 24, 25, 26, 34, 35, 36, 43, 44, 14, 15 corresponding to the terminal processing devices that have completed the transfer and do not have the processing device group to be transferred next. , 16 has been successfully transferred to the third-tier processing devices 21, 22, 23, 31, 32, 33, 41, 42, 11, 12, 13 as the data transfer source. Is transmitted, a data transfer completion list in which each recognition code is written is sent.

  The third-tier processing devices 21, 22, 23, 31, 32, 33, 41, 42, 11, 12, and 13 to which the data transfer completion list has been sent are sent identification codes for identifying the respective processing devices. In addition, the updated data transfer completion list is sent to the second-layer processing devices 20, 30, 40, and 10 that were the data transfer sources.

  The second tier processing devices 20, 30, and 40 that have received the data transfer report each time the data transfer completion list is sent from the third tier processing device, the updated data by the identification code of its own processing device. The transfer completion list is transmitted to the processing device 10 in the first layer.

  Alternatively, the second-layer processing devices 20, 30, and 40 that have received the data transfer report wait for the data transfer completion lists sent from the three, three, and two processing devices, respectively. All are added and updated as one data transfer completion list, the recognition codes of the respective processing devices 20, 30, and 40 are added to the updated data transfer completion list, and the updated data transfer completion list is added to the updated data transfer completion list. It can also be transmitted to the processing device 10 in one layer.

  The processing device 10 of the transmission source that has started the first transfer additionally updates the data transfer completion list sent from the processing devices 20, 30, 40, 11, 12, and 13 as one confirmation, and completes the processing.

  As described above, according to the sixth embodiment, each processing device sends a data transfer completion list to a higher-level processing device by going back the route in which the data is transferred, and moves toward the top of the tree structure. The processing apparatus of the transmission source that first started data transfer can finally grasp the data transfer result of the entire processing system.

  At this time, for example, even when the failure described with reference to FIGS. 9 and 10 occurs, data cannot be transferred to any of the processing devices described in the processing device group. By providing means for recognizing that it is a terminal processing device, the data transfer completion list is displayed on the processing device of the transmission source that first started data transfer, including processing devices that have not yet completed data transfer. Will be sent. Thereby, the processing apparatus 10 that wishes to confirm completion of data transfer can obtain a data transfer completion list with low load without increasing the load unnecessarily.

  Obtaining the transfer result determined according to the present embodiment is, for example, as a new data transfer after totaling the processing devices that could not be transferred and generating a list of new processing device groups and putting time. Data transfer can be continued. Accordingly, in consideration of the business of selling digital data (video, music, etc.), it is possible to provide a practical data transfer system capable of deterministically transmitting data to the ordered processing device. Also, there will be no scandals such as charging even if the data transfer is not completed by mistake.

  In the above description, the case where the data transfer completion list is sent to the processing device one layer above has been described. However, the data transfer completion list is directly sent to the processing device which is the transmission source of each processing device. By doing so, the same effect can be obtained.

Embodiment 7 FIG.
In the sixth embodiment, the case where the data transfer completion list is additionally updated by tracing the route through which the data is transmitted in the reverse order has been described. In the seventh embodiment, a case will be described in which a failure occurs in the processing device along the way.

  FIG. 12 is a diagram showing a tree structure of divided processing devices and a data transfer completion list from each processing device in the data transfer system according to the seventh embodiment of the present invention. This shows a case where a failure has occurred in the processing device 20 along the way of returning the data transfer completion list. Similarly to the sixth embodiment, the fourth layer processing devices 24, 25, and 26 transfer the data transfer completion list to the upper third layer processing devices 21, 22, and 23.

  Next, the processing devices 21, 22, and 23 in the third layer try to update the data transfer completion list and send it to the processing device 20 in the upper second layer, but a failure has occurred and communication is not possible. It is in a state.

  In this case, each of the processing devices 21, 22, and 23 of the third hierarchy additionally adds that the data transfer completion of the processing device 20 of the second hierarchy is completed when the data transfer completion list is additionally updated. The data transfer completion list that has been additionally updated is sent to the upper level processing apparatus 10.

  Since the processing device 20 of the second layer has transferred the data to the processing devices 21, 22, and 23 of the third layer, the data reception from the processing device 10 of the first layer, which is the upper layer, is naturally completed. Can be regarded as being.

  As described above, according to the seventh embodiment, when a failure occurs at the timing when the data transfer completion list is returned, or when a processing device that has been turned off is present, the processing device directly in the upper layer Thus, it is possible to obtain a data transfer system capable of sending the data transfer completion list to a higher level by jumping over the above.

  As a result, the function of notifying the completion of data transfer to the processing apparatus of the transmission source that first started data transfer can be realized with high reliability. Also, each processing device on the path gradually summarizes the data transfer completion list compared to the case where all the processing devices transmit the data transfer completion list to the processing device of the transmission source that first started data transfer. Therefore, the data transfer completion list finally received by the first processing device is small, and the load can be drastically reduced.

  Further, the data transfer completion list can be transmitted by skipping the processing apparatus in which the failure has occurred, including the completion of data transfer of the processing apparatus in which the failure has occurred. Thereby, even when a failure occurs during sending of the data transfer completion list, the processing device of the transmission source that first started data transfer can accurately grasp the transfer completion status of the entire data transfer system.

Embodiment 8 FIG.
In the above-described first to seventh embodiments, the network is not particularly described, but the protocol supported by the so-called IP network (v4, v6), the telephone line, the wireless line, the private dedicated line, etc. It may be a form. Broadcast and satellite lines are also included in such communication lines.

Embodiment 9 FIG.
In the above-described first to eighth embodiments, since attention is paid to transferring data to each processing device, a misunderstanding as if the processing device needs data to be transferred tends to occur. As in Embodiment 8, a processing device may be used as a simple transfer path when it is desired to transfer certain identical data via a heterogeneous network. Hereinafter, some specific examples will be described. In any case, the processing apparatus that has received the data does not necessarily need the data.

  For example, referring to FIG. 2, the processing device 10 is connected by a broadband network (so-called Internet), but the processing device 20 can be connected by broadband at a1 and communicates at b1, b2, and b3. The data transfer of the present invention can also be applied to a shared server connected through a possible in-house network. In the case of such a utilization method, the shared server itself does not need data, and a lower-layer processing device that receives data via the shared server actually uses the data.

  In another example, a configuration in which the processing device 10 in the first layer is a broadcasting station, the processing devices 20, 30, and 40 in the second layer are receivers, and the processing devices in lower layers are also considered as a home network. . In this case, it is considered that the receiver receives data from the broadcast line and is connected to the home via IEEE1394, a power line network, a wireless LAN, or the like.

  Even in such a case, the receiver does not necessarily need data, and the data is sent to a specific device in the home (for example, a white home appliance not having a broadcast receiving function = a future digital home appliance). It is also possible to go through a receiver.

  In yet another example, in a mobile phone network, a processing device positioned higher in the tree structure, such as the processing devices 10, 20, 30, 40, is a mobile phone base station, and a processing device positioned lower than that is a mobile phone. You can also think about it.

  In yet another example, an in-vehicle device (such as a so-called car navigation system) of an automobile can send information on road construction information or accidents from another in-vehicle device, so that location information can also be transmitted to nearby in-vehicle devices. It can be distributed in order. Usually, the information transmission source requires a large amount of processing power. In this case, it is only necessary to send information data to at most several in-vehicle devices, and to distribute the data transfer load to each in-vehicle device. And is efficient.

  As a method for creating a processing device group in this case, a command for requesting a response may be thrown so that a nearby in-vehicle device can receive the command, and the command may be created based on the command. After that, each in-vehicle device can dynamically advance the processing while generating the processing device group one after another in the same manner, and can transfer data to a wide range of in-vehicle devices. Recently, a camera is often mounted on a car, but it can also be used as a method of transferring the camera video to another on-vehicle device one after another.

  The data may be a large volume of data such as a file, or a small amount of unit data such as an IP packet. When it is realized as a small amount of data, a route to the processing device at the end is established with the first unit data, and then transmitted according to the established route, so that video on demand (VOD) or live stream ( (Broadcast) etc. is also conceivable.

  As described above, according to the ninth embodiment, the data transfer system of the present invention can be applied even when a processing device is used as a simple transfer path, such as when it is desired to transfer certain identical data via a heterogeneous network. is there.

  In the above-described embodiment, a case has been described in which each processing apparatus employs common division index data, but the data transfer system of the present invention is not limited to this. Each processing apparatus can perform division processing based on individual division index data. For example, the processing device of the first layer can also perform the division processing using the number of lines that can be simultaneously communicated as the division index data, and the processing device of the second layer can perform the division processing using the network bandwidth as the division index data, Division processing can be performed using division index data suitable for each processing device in the data transfer system.

  Furthermore, in the above description, the division index data is determined based on the index data given from the upper layer, but the data transfer system of the present invention is not limited to this. For example, the division processing unit 2 can measure the response time of each processing device in the current network environment by sending a command for requesting a response to the processing device to be transferred.

  In this case, the division processing means 2 updates the index data with the measured response time, and adopts the response time or data consisting of a combination of this response time and other index data items as the division index data. , Division processing can be performed.

  Alternatively, the division processing unit 2 acquires the current data such as the CPU usage rate and the memory usage rate of the processing device as the latest system load information from the processing device to be transferred, thereby updating the index data and performing a new division. Division processing based on the adoption of index data can be realized.

  In the above description, the division processing is described in which the total of the division index data of each processing device group and the total number of processing devices are both equal, but the data transfer system of the present invention is limited to this. Is not to be done. It is also possible to improve the efficiency of data transfer to the entire processing apparatus by performing a dividing process that equalizes only the sum of the division index data.

It is a block diagram of the processing apparatus used for the data transfer system in Embodiment 1 of this invention. It is a figure which shows the tree structure of the processing apparatus divided | segmented in the data transfer system of Embodiment 1 of this invention. It is a figure which shows an example of the list of the processing apparatus groups in Embodiment 1 of this invention. It is an example of the identification code | symbol of each processing apparatus in Embodiment 1 of this invention. It is a flowchart which shows a series of data transfer processes in the processing apparatus of each hierarchy of Embodiment 1 of this invention. It is a flowchart which shows a series of division | segmentation processes in the processing apparatus of each hierarchy of Embodiment 1 of this invention. It is a figure which shows the tree structure of the processing apparatus divided | segmented in the data transfer system of Embodiment 2 of this invention. It is a flowchart which shows a series of data transfer processes in the processing apparatus of each hierarchy of Embodiment 2 of this invention. It is explanatory drawing of operation | movement at the time of the failure generation in the data transfer in Embodiment 3 of this invention. It is explanatory drawing of operation | movement at the time of the failure generation in the data transfer in Embodiment 4 of this invention. It is a figure which shows the tree structure of the divided | segmented processing apparatus in the data transfer system of Embodiment 6 of this invention, and the data transfer completion list from each processing apparatus. It is a figure which shows the tree structure of the processing apparatus divided | segmented in the data transfer system of Embodiment 7 of this invention, and the data transfer completion list from each processing apparatus.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 Reception means, 2 Division | segmentation process means, 3 Connection means, 4 Transfer means 10, 11, 12, 13, 20, 21, 22, 23, 30, 31, 32, 33, 40, 41, 42 Processing apparatus.

Claims (13)

Based on the processing device group list data in which the processing devices connected by the communication line and the index data relating to the processing capability of each processing device are associated with each other, the data transfer destination processing device is determined and the processing device group list data is divided And dividing processing means to be assigned to each of the determined transfer destination processing devices,
Transfer means for transferring data to be transferred to each transfer destination processing device determined by the division processing means and processing device group list data assigned to each transfer destination processing device by the division processing means, and
When the data transfer is normally completed, the transfer means notifies the upper layer processing device that the data transfer has been normally completed. The data transfer processing device according to claim 1,
The data transfer processing device, wherein when the data transfer is normally completed, the transfer means notifies the processing device of one layer higher level that the data transfer is normally completed. The data transfer processing device according to claim 1,
The data transfer processing device, wherein when the data transfer is normally completed, the transfer means notifies the highest level processing device that the data transfer is normally completed. The data transfer processing device according to any one of claims 1 to 3,
The transfer means determines that the data transfer has been normally completed when it is notified that the data transfer has been normally completed from each processing terminal to which the processing device has transferred the data, and the data transfer has been normally completed. A data transfer processing device characterized by notifying a higher-level processing device. The data transfer processing device according to claim 4,
The transfer means determines that the data transfer has been normally completed when it is notified from all the processing terminals to which the processing device has transferred the data that the data transfer has been normally completed, and the data transfer has been normally completed. A data transfer processing device characterized by notifying a higher-level processing device. In a data transfer system for transferring data to a plurality of processing devices,
Each of the plurality of processing devices is
Receiving means for receiving processing device group list data in which index data relating to processing capability of the processing device is associated with each processing device to be retransmitted, and data to be retransmitted;
Division index data for each processing device to be retransmitted is obtained from the index data included in the processing device group list data received by the receiving unit, and n is calculated from the processing device group list data based on the division index data. Division processing means for determining the number of processing devices (n is an integer of 2 or more) and n post-division processing device group list data;
connection means for connecting to the n processing devices determined by the division processing means using n communication lines;
Each of the n processing apparatuses connected by the connecting means by adding the data to be retransmitted to each of the n divided post-processing apparatus group list data determined by the division processing means. And a transfer means for transferring to
The receiving unit receives a data transfer completion list including an identifier indicating the transmission source processing device from the transmission source processing device after data transfer from the transmission source processing device is normally completed.
The transfer means monitors whether or not data communication has been normally completed based on a response from the n processing devices when transferring data to each of the n processing devices, When the data transfer to one processing apparatus is normally completed, transfer completion data including an identifier indicating its own processing apparatus is added to update the data transfer completion list, and the updated data transfer completion list is A data transfer system characterized by transferring data to a single processing device. The data transfer system according to claim 6, wherein
The division processing unit obtains the total number of lines that the determined number of the n processing devices can simultaneously communicate from the index data, and determines the n units from all the processing devices included in the processing device group list data. If the total number of processing devices to be retransmitted is determined by the n processing devices excluding the processing devices, and if the total number of processing devices to be retransmitted is greater than the total number of lines, the self processing is performed. The apparatus is assigned as the (n + 1) th processing apparatus, n + 1 post-partition processing apparatus group list data is determined, and n pieces other than the (n + 1) th post-split processing apparatus group list data corresponding to the (n + 1) th processing apparatus The post-split processing device group list data is determined as the n post-split processing device group list data corresponding to the determined n processing devices. The data transfer system according to claim 7, wherein
The division processing means includes the n processing devices and the n divided post-processing devices until the number of processing devices included in the (n + 1) -th post-processing device group list data is n or less. A data transfer system characterized by repeatedly executing a process for determining group list data. The data transfer system according to any one of claims 6 to 8,
When the transfer means is a terminal processing device that does not have a processing device to be transferred, the transfer means displays the updated data transfer completion list as the first transmission source described in the data transfer completion list. A data transfer system, characterized by being transmitted to a processing device. The data transfer system according to any one of claims 6 to 8,
If the processing unit is a terminal processing device that does not have a processing device to be transferred, the transfer unit displays the updated data transfer completion list as a source immediately before the data transfer completion list. If the processing device is not the terminal processing device, the data transfer completion list received from any one of the processing devices at the transmission destination is displayed immediately before the data transfer completion list. A data transfer system, wherein the data is transferred to a processing device at a transmission source. The data transfer system according to claim 10, wherein
The transfer unit generates a single data transfer completion list by combining the data transfer completion lists received from all of the transmission destination processing devices, and the generated single data transfer completion list is used as the immediately previous transmission source processing device. A data transfer system characterized by being transferred to The data transfer system according to claim 10 or 11,
When the transfer means transfers the data transfer completion list to the immediately previous transmission source processing device, the transfer means normally transfers the data transfer completion list based on a response from the previous transmission source processing device. Monitoring whether or not the transfer of the data transfer completion list is not completed normally, and if the previous transmission of the processing device of the previous transmission source described in the data transfer completion list is detected. A data transfer system, wherein the data transfer completion list is transferred to an original processing device. The data transfer system according to claim 12,
When the transfer means detects that the transfer of the data transfer completion list is not normally completed, the transfer means adds to the data transfer completion list that the previous transmission source processing device has been transferred, and A data transfer system, wherein the added data transfer completion list is transferred to a transmission source processing device further preceding the previous transmission source processing device described in the data transfer completion list.

Images