JP5487891B2 - COMMUNICATION CONTROL DEVICE, COMMUNICATION CONTROL PROGRAM, AND DATA DISTRIBUTION SYSTEM - Google Patents

Description

This invention relates to a communication control apparatus, communication control program,及 Beauty, relates to a data distribution system, for example, it can be applied to data distribution server for streaming of moving image data into a plurality equipped VOD (Video On Demand) system.

  Conventionally, in a VOD system (data distribution system), when there are a plurality of data distribution servers (VOD servers) that distribute moving image data to a client, there is a communication control device called a dispatcher, and the client that has requested distribution However, the load was distributed by assigning and specifying the data distribution server of the distribution request destination.

  As a system including a conventional dispatcher, there is a multitask system described in Patent Document 1.

JP 2000-284980 A

  However, in the technique described in Patent Document 1, since communication control is concentrated by one dispatcher, the dispatcher becomes a bottleneck when the number of clients increases. For this reason, in a large-scale system, it is necessary to use multiple dispatchers in parallel. However, if multiple dispatchers communicate closely to check the load on each server, not only the server but also the network load Will increase.

Therefore, in the data distribution system comprising a dispatcher (communication control device), a communication control device capable of reducing the amount of communication between the data distribution server and dispatcher (communication control device), a communication control program,及 Beauty, data distribution A system is desired.

According to a first aspect of the present invention, there is provided a communication control device for controlling communication between a client device and a plurality of data distribution servers that perform distribution of distribution data to the client device. (1) For each of the data distribution servers, Server load information acquiring means for acquiring server load information relating to the load status of the server, (2) server allocation request receiving means for receiving an allocation request for allocating the data distribution server from the client device, and (3) acquiring the server load information. Data distribution allocated to the above allocation request by estimating the current load status of the data distribution server using the server load information for each data distribution server acquired by the means in the past and considering the estimation result From the server selection means for selecting a server, and (4) the data distribution server selected by the server selection means, The client device, to deliver the delivery data, the server selection means selects the data distribution server, and has a communication control means for controlling the client device, (5) server load information, the server load When the information is generated in the data distribution server, the stream distribution end schedule information relating to the scheduled end time of the stream currently being distributed in the data distribution server is included. (6) The server selection means includes the server The load information acquisition means estimates the current load status of the data distribution server by using stream distribution end schedule information included in the server load information acquired from the data distribution server in the past .

A communication control program according to a second aspect of the present invention includes: (1) a computer mounted on a communication control device that controls communication between a client device and a plurality of data distribution servers that perform stream distribution of distribution data to the client device. (2) Server load information acquisition means for acquiring server load information related to the current load status for each of the data distribution servers, and (3) a server allocation request for receiving an allocation request for allocating the data distribution server from the client device And (4) using the server load information for each data distribution server acquired in the past by the server load information acquisition unit, estimating the current load status of the data distribution server and considering the estimation result And server selection means for selecting a data distribution server to be allocated in response to the allocation request, and (5) the server From the data delivery server selection means selects, on the client device, to deliver the delivery data, the server selection means selects the data distribution server, and to function as a communication control means for controlling the client device, (6 ) The server load information includes stream distribution end scheduled information related to the scheduled end time of the stream currently being distributed by the data distribution server at the time when the server load information is generated in the data distribution server. 7) The server selection means estimates the current load status of the data distribution server using the stream distribution end schedule information included in the server load information acquired from the data distribution server in the past by the server load information acquisition means. vinegar Rukoto and features.

According to a fifth aspect of the present invention, there is provided a client device, a plurality of data distribution servers that perform stream distribution of distribution data to the client device, and one or a plurality of communication controls that control communication between the client device and the data distribution server. In the data distribution system having the apparatus, (1) each of the data distribution servers (1-1) server load information generation for generating server load information including information on the current load status of the data distribution server And (1-2) server load information transmitting means for giving the server load information generated by the server load information generating means to the communication control device. (2) First as each of the communication control devices. The communication control device of the present invention is applied.

  According to the present invention, in a data distribution system including a communication control device that distributes requests from clients to any of a plurality of data distribution servers, the amount of communication between the data distribution server and the communication control device can be reduced. .

It is a block diagram shown about the whole data distribution system composition concerning a 1st embodiment. It is explanatory drawing shown about the content of the server load information transmitted to the dispatcher from the data delivery server which concerns on 1st Embodiment. It is the sequence diagram shown about the operation example between each apparatus in the data delivery system which concerns on 1st Embodiment. It is the flowchart shown about the process of the server selection operation | movement in the dispatcher which concerns on 1st Embodiment. It is the flowchart shown about the process of the server selection operation | movement in the dispatcher which concerns on 2nd Embodiment. It is explanatory drawing shown about the example of the content of the stream delivery completion | finish schedule information which concerns on 3rd Embodiment. It is explanatory drawing shown about the example of the processing content of the deallocated function used in the process of the dispatcher which concerns on 3rd Embodiment. It is explanatory drawing shown about the specific example of the stream delivery completion | finish schedule information which concerns on 3rd Embodiment.

(A) First Embodiment Hereinafter, a first embodiment of a communication control device, a communication control program , and a data distribution system according to the present invention will be described in detail with reference to the drawings. In the first embodiment, the communication control device of the present invention is described as being a dispatcher.

(A-1) Configuration of the First Embodiment FIG. 1 is a block diagram showing the overall configuration of the data distribution system 10 of this embodiment. In FIG. 1, the reference numerals in parentheses are those used in the second and third embodiments described later.

  The data distribution system 10 includes four data distribution servers 20 (20-1 to 20-4), a communication control system 30, and N clients 40-1 to 40-N. In the communication control system 30, a Proxy 31 and three dispatchers 32 (32-1 to 32-3) are arranged.

  The number of data distribution servers 20, dispatchers 32, and clients 40 is not limited.

  The data distribution server 20 is a device that distributes moving image data to the client 40 according to the control of any one of the dispatchers 32. In this embodiment, the data distribution server 20 is described as distributing moving image data. However, the audio data may be added to the moving image data and distributed, or only the audio data may be distributed. Anyway.

  Here, for example, the data distribution server 20 distributes the data distribution of the embodiment to an apparatus having an execution configuration of a program such as a CPU, ROM, RAM, EEPROM, and hard disk and an interface for communicating with other communication apparatuses. You may make it build by installing a program etc.

  The data distribution server 20 provides each dispatcher 32 with information on the load status of the data distribution server 20 (hereinafter referred to as “server load information”) at predetermined intervals. Details of the server load information will be described in detail in an operation description to be described later.

  Further, in the data distribution server 20, the same configuration as that of the data distribution server of the existing data distribution system (VOD system) can be applied except that the server load information is generated and given to each dispatcher 32. .

  In FIG. 1, each of the data distribution servers 20-1 to 20-4 may store moving image data (may be the same data or different data), or may be a data distribution server. 20-1 to 20-4 may be provided with a common storage device (not shown), and moving image data may be read from the storage device and held. That is, the method by which the data distribution servers 20-1 to 20-4 hold moving image data is not limited. In FIG. 1, in order to simplify the description, the data distribution servers 20-1 to 20-4 are described as being capable of holding the same moving image data.

  The Proxy 31 has a function of relaying communication between the client 40 and the dispatcher 32. The Proxy 31 receives a request (hereinafter referred to as “server allocation request”) from the client 40 for allocating one of the data distribution servers 20 to data distribution, and distributes the server allocation request to one of the dispatchers 32 for transfer. Thereafter, the communication between the client 40 and the dispatcher 32 is relayed.

  In FIG. 1, as an example, the Proxy 31 exists between the client 40 and the dispatcher 32, but the Proxy 31 is not essential and can be omitted. For example, a specific client 40 is assigned to each dispatcher 32 in advance (for example, statically set on the client 40 side or assigned by a DNS server (not shown)), and the client 40 directly accesses the dispatcher 32. Also good.

  Further, the method by which the Proxy 31 allocates a server allocation request to any one of the clients 40 is not limited. For example, an existing load distribution method such as a random selection or a round robin method is applied. You may do it.

  As the clients 40-1 to 40-N, the same clients as those in the existing data distribution system (VOD system) can be applied.

  The dispatcher 32 performs control related to communication between the data distribution server 20 and the client 40.

  The dispatcher 32 installs the communication control program of the embodiment or the like in a device having an interface for communicating with other communication devices and an execution configuration of programs such as CPU, ROM, RAM, EEPROM, and hard disk, for example. You may make it build by.

  The dispatcher 32 acquires server load information of each data distribution server 20-1 to 20-4 at predetermined intervals.

  When the dispatcher 32 receives the server allocation request from the client 40, the dispatcher 32 uses the server load information acquired in the past from each of the data distribution servers 20-1 to 20-4 to determine the data distribution server 20 to be allocated to the client 40. select. Then, the dispatcher 32 transmits a response to the server allocation request (hereinafter referred to as “server allocation response”) having at least the identification information of the selected data distribution server 20 to the client 40 via the Proxy 31. Upon receiving the server allocation response, the client 40 extracts the identification information of the data distribution server 20 included in the server allocation response, accesses the data distribution server 20 of the identification information, and makes an actual distribution request. Thereafter, the data distribution server 20 performs stream distribution to the client 40 that has accessed based on the server allocation response.

  As a method in which each dispatcher 32-1 to 32-3 holds the server load information of the data distribution servers 20-1 to 20-4, for example, each dispatcher 32-1 to 32-3 is a data distribution server. The server load information may be requested and read from 20-1 to 20-4, and conversely, the server load information is distributed from the data distribution servers 20-1 to 20-4 to the dispatchers 32-1 to 32-3. You may do it. That is, if the dispatchers 32-1 to 32-3 can periodically hold the server load information from the data distribution servers 20-1 to 20-4, the communication procedure is not limited. Details of the server load information will be described in detail in the operation description to be described later.

(A-2) Operation of the First Embodiment Next, the operation of the data distribution system 10 of the first embodiment having the above configuration will be described.

(A-2-1) Regarding Server Load Information First, processing related to server load information in the dispatcher 32 will be described.

  FIG. 2 is an explanatory diagram showing an example of contents included in server load information given from the data distribution server 20 to the dispatcher 32.

  As illustrated in FIG. 2, here, the server load information is described as including information of “ServerID”, “Priority”, “stream”, and “updatetime”.

  “ServerID” is identification information of the data distribution server 20, and for example, address information, a domain name, or the like is applied.

  “Priority” is a parameter indicating the level of priority of the data distribution server 20. It is assumed that a different value is set in advance for each data distribution server 20 for Priority. Although details will be described later, the dispatcher 32 allocates stream distribution to the data distribution server 20 in consideration of the priority. As Priority, for example, a smaller numerical value may be set as the priority is higher.

  “Stream” is a parameter indicating the surplus processing capability of the data distribution server 20. In the first embodiment, “stream” indicates the number of remaining streams that can be distributed in the data distribution server 20. For example, when the number of streams that can be simultaneously distributed to the client 40 by the data distribution server 20 is 10, and currently 5 streams are being distributed, the stream indicating the number of remaining streams that can be distributed is “5”. Become.

  “Updatetime” indicates a time (time when the server load information is updated).

(A-2-2) Processing Between Each Device FIG. 3 is a sequence diagram showing processing between each device constituting the data distribution system 10.

  In FIG. 3, in order to simplify the description, a sequence diagram relating to processing among the client 40-1, the dispatcher 32-1, and the data distribution server 20-1 is shown. As shown in FIG. 1, a plurality of clients 40, dispatchers 32, and data distribution servers 20 are arranged.

  In FIG. 3, first, it is assumed that a server allocation request is given from the client 40-1 to the Proxy 31, and the server allocation request is given by the Proxy 31 to the dispatcher 32-1 (S101). In the following, the operation related to the dispatcher 32-1 will be described. However, the operations of the other dispatchers 32-2 and 32-3 when the server allocation request is given are the same, and the description thereof will be omitted.

  Next, when a server allocation request is given from the client 40-1, the dispatcher 32-1 selects the data distribution server 20 to be allocated to the client 40-1 (S102).

  Here, in the dispatcher 32-1, the data distribution server 20 is basically allocated in order of priority (priority order represented by priority in the server load information), but cannot be allocated because the distribution limit is reached. In order to avoid allocation to the data distribution server 20, a server having a server load information stream smaller than a predetermined value is not allocated. Since the reliability of the server load information is lowered depending on the elapsed time from the time of acquisition, it is determined which server is assigned based on the information.

  Here, in the process of step S102, it is assumed that the data distribution server 20-1 is selected by the dispatcher 32-1, but details of the process of selecting the data distribution server 20 of the dispatcher 32-1 will be described later.

  Next, the dispatcher 32-1 transmits a server allocation request to the data distribution server 20-1 selected in step S102 (S103).

  Next, the data distribution server 20-1 performs temporary allocation to the client 40-1, and transmits access information to the data distribution server 20-1 to the dispatcher 32-1 (S104).

  Next, the dispatcher 32-1 transmits access information to the client 40-1 that is the transmission source of the server allocation request (S105).

  Next, the client 40-1 transmits a distribution request to the data distribution server 20-1 based on the access information (S106).

  Next, data distribution based on the distribution request is performed from the data distribution server 20-1 to the client 40-1 (S107).

(A-2-3) Data Distribution Server Selection Process FIG. 4 is a flowchart showing the selection process of the data distribution server 20 of the dispatcher 32-1 in step S102 described above.

  In the flowchart of FIG. 4, the description will be made assuming that “I” is used as a variable used for the loop processing. In the flowchart of FIG. 4, M (representing the number of data distribution servers 20) is used as a constant used for loop processing.

  First, in the dispatcher 32-1, the threshold for the stream is obtained based on the elapsed time using the following equation (1) (S201).

Threshold = α (now-updatetime) + initial value (1)
In the equation (1), α is a predetermined positive value set in advance. Note that α is preferably a value proportional to the total number of dispatchers 32. This is set on the assumption that the Proxy 31 distributes the allocation request evenly to the dispatchers 32 (or the clients 40 are allocated equally to the dispatchers 32).

  In equation (1), “now” is the current time, and “updatetime” is the updatetime in the server load information acquired last. That is, “now-updatetime” indicates an elapsed time from the time when the server load information acquired last for the data distribution server 20 is generated to the present.

  Here, for simplicity of explanation, the threshold values calculated in step S201 described above are all the same in the data distribution servers 20-1 to 20-4 (that is, the servers in the data distribution servers 20-1 to 20-4). The load information is generated at the same timing). However, when the updatetime to be applied is different for each data distribution server 20, the processing in step S201 described above needs to be performed for each data distribution server 20. .

  In addition, the initial value in equation (1) is not limited, and an arbitrary value is set.

  Next, a variable I used for loop processing to be described later is initialized (I = 0) (S202). When I = 0, in the loop processing of steps S203 to S205 described later, processing is performed with the data distribution server 20 having the highest priority (order determined by Priority in the server load information) as the processing target server. In the following description, it is assumed that the processing of the data distribution server 20 having the second highest priority is performed when I = 1. For example, when the priority order is set in the order of the data distribution servers 20-1, 20-2, 20-3, and 20-4, the first (I = 0) is the data distribution server 20-1, 2. The first (I = 1) is the data distribution server 20-2, the third (I = 2) is the data distribution server 20-3, and the fourth (I = 3) is the data distribution server 20-4. It becomes.

  In the loop processing in the dispatcher 32-1, first, it is determined whether or not the latest stream in the server load information of the processing target server is larger than the threshold calculated in step S201 described above (S203).

  In step S203, if stream is greater than the threshold, the processing target server is selected as the data distribution server 20 to be allocated to the server allocation request, and the process ends.

  On the other hand, if stream is equal to or smaller than the threshold value in step S203, the processing target server is removed from the data distribution server 20 assigned to the server assignment request. Then, the variable I is incremented (I = I + 1) (S204), and it is further determined whether or not the variable I = M (S205). In step S205, when the variable I = M is not satisfied (when there is a data distribution server 20 that is not yet a server to be processed), the operation starts from the above-described step S203, and when the variable I = M (all If the data distribution server 20 is a server to be processed), the process ends assuming that there is no data distribution server 20 to be allocated to the server allocation request.

  As described above, the dispatcher 32 uses the stream value in the server load information acquired last for each data distribution server 20 and the elapsed time from the time when the server load information acquired last is generated to the present. Then, the current load situation of each data distribution server 20 is estimated, and the data distribution server 20 to be allocated to the server allocation request is selected in consideration of the estimation result.

  For example, in the above equation (1), α = 2 and initial value = 0 are applied, the stream of the data distribution server 20-1 is 5, the stream of the data distribution server 20-2 is 10, and the data distribution server 20- Assume that the priority of 1 is the highest and the data distribution server 20-2 is the next highest. In this case, since the threshold reaches the stream value (5) of the data distribution server 20-1 after the elapsed time (now-updatetime) is 2.5 seconds, the data distribution server 20-1 at this point After that, the data distribution server 20-2 is assigned. If the prediction of the threshold is correct, useless communication with the data distribution server 20-1 can be avoided.

(A-3) Effects of First Embodiment According to the first embodiment, the following effects can be achieved.

  The dispatcher 32 estimates the data distribution server 20 that can accept the distribution by using the content of the server load information and the elapsed time (now-updatetime) from the time when the server load information was last generated. Unnecessary communication with the impossible data distribution server 20 can be avoided. For example, in the first embodiment, if the estimation of the data distribution server 20 that cannot accept distribution is not performed, the server information allocated by another dispatcher 32 and the actual information are allocated as they are. There is a problem that communication continues to fail and communication fails.

  Further, since the data distribution server 20 that can accept distribution is estimated using not only the contents of the server load information but also the elapsed time (now-updatetime) from the time when the server load information was last generated, for example, Since the server load information only needs to be acquired less frequently than when estimated based only on the server load information, the amount of communication between the data distribution server 20 and the dispatcher 32 can be reduced.

  Further, in the data distribution system 10, since the dispatchers 32-1 to 32-3 are individually operated in parallel, the synchronization processing between the dispatchers 32 in a normal time or when a failure of some dispatchers 32 occurs. Switching process is not required, and the construction cost is reduced and the maintainability is improved.

(B) Second Embodiment Hereinafter, a second embodiment of a communication control device, a communication control program , and a data distribution system according to the present invention will be described in detail with reference to the drawings. In the second embodiment, the communication control device of the present invention is described as a dispatcher.

(B-1) Configuration of Second Embodiment A data distribution system 10A of the second embodiment can also be described with reference to FIG.

  In the data distribution system 10A of the second embodiment, the data distribution system 10 of the first embodiment is replaced with the data distribution system 10A, but the other configurations are the same as those of the first embodiment, and thus detailed. Description is omitted.

  In the data distribution system 10A, the dispatcher 32 (32-1 to 32-3) is replaced with the dispatcher 32A (32A-1 to 32A-3), but other configurations are the same as those in the first embodiment. Detailed explanation is omitted here.

  The dispatcher 32A is different from the dispatcher 32 of the first embodiment only in the process of selecting the data distribution server 20 to be allocated in response to the server allocation request (the process of step S102 in FIG. 3 described above). The details of the process of selecting the data distribution server 20 to be allocated to the server allocation request in the dispatcher 32A will be described in detail in the description of the operation described later.

(B-2) Operation of Second Embodiment Next, the operation of the data distribution system 10A of the second embodiment having the above configuration will be described.

  The processing between the devices constituting the data distribution system 10A can also be shown in FIG. However, in the data distribution system 10A of the second embodiment, only the processing of step S102 in FIG. 3 (processing for selecting the data distribution server 20 to be allocated to the server allocation request) is different from that of the first embodiment. Only the process according to step S102 will be described below.

  FIG. 5 is a flowchart showing the process corresponding to the above-described step S102 (selection process of the data distribution server 20) in the dispatcher 32A-1.

  In the flowchart of FIG. 5, as in the first embodiment (see FIG. 4 described above), “I” is used as a variable used for loop processing, and M (the number of data distribution servers 20) is used as a constant used for loop processing. Shall be used.

  Each dispatcher 32A accumulates server load information received from each data distribution server 20, and uses the history of the accumulated server load information in the operation according to the flowchart of FIG.

  First, in the dispatcher 32A, a variable I used for loop processing described later is initialized (I = 0) (S301). When I = 0, in the loop processing of steps S302 to S305 described later, the data distribution server 20 having the highest priority (the order determined by the priority in the server load information) is the same as in the first embodiment. The processing is performed as a server to be processed, and the processing of the data distribution server 20 having the second highest priority is performed when I = 1.

  In the loop processing in the dispatcher 32A-1, in the dispatcher 32A-1, each dispatcher 32A analyzes the change tendency of the stream value in the server load information of each data distribution server 20 based on the accumulated server load information history. . Then, the dispatcher 32A-1 calculates a value (hereinafter referred to as “new_stream”) indicating the surplus processing capability in each data distribution server 20 based on the change tendency (S302). A specific process for calculating new_stream will be described later.

  Next, the dispatcher 32A-1 determines whether or not the new_stream calculated in step S302 described above is larger than a predetermined threshold (S303). Although an arbitrary value can be set as the threshold used in step S303, for example, “0” may be applied.

  In step S303, when stream is larger than the threshold, the processing target server is selected as the data distribution server 20 to be allocated to the server allocation request, and the process ends.

  On the other hand, if stream is equal to or smaller than the threshold value in step S303, the processing target server is removed from the data distribution server 20 assigned to the server assignment request. Then, the variable I is incremented (I = I + 1) (S304), and it is further determined whether or not the variable I = M (S305). In step S305, if the variable I is not equal to M (when there is a data distribution server 20 that is not yet a server to be processed), the processing starts from the process in step S302 described above. If the data distribution server 20 is a server to be processed), the process ends assuming that there is no data distribution server 20 to be allocated to the server allocation request.

[About the calculation method of new_stream]
Next, the details of the new_stream calculation method in step S302 will be described.

  First, the dispatcher 32A-1 analyzes the change tendency of the stream value in the server load information of each data distribution server 20 based on the history of server load information received and accumulated from each data distribution server 20. Here, as an example, each dispatcher 32A will be described as determining whether the value of the stream associated with each data distribution server 20 is increasing or decreasing using the following equation (2).

Note that in equation _{(2), α 1 + α 2 + α} 3 + ... + α i = 1, and it is assumed that the value relationship holds for _{α 1> α 2> α 3} >...> α i are set in advance . In the expression (2), stream, stream ₁ , stream ₂ , stream ₃ ,..., Stream _i are stream information in the server load information given to the dispatcher 32 </ _b > A from the data distribution server 20 subject to change tendency determination. Shows a history of values. That is, stream indicates the latest value, and stream _i indicates the oldest value. An arbitrary value may be set as the value of i.

δ = α ₁ (stream-stream ₁ ) + α ₂ (stream ₁ -stream ₂ )
+ ... + α _i (stream _i-1 -stream _i ) (2)
Then, the dispatcher 32A estimates the number of streams that can be currently allocated to the data distribution server 20 by further applying δ obtained by the above equation (2) to the following equation (3).

  In equation (3), MAX (...) Is a function for calculating the value of new_stream. For example, when expressed as MAX (a, b), if a ≧ b, a is returned as the result of MAX (a, b), and if a <b, b is returned as the result of MAX (a, b). Shall. That is, in the expression (3), new_stream = a when a is greater than b (0), and new_stream = 0 when a is less than 0.

The server information update interval indicates an interval at which server load information is generated in the data distribution server 20. Although any value can be set as the server information update interval of each data distribution server 20, in the second embodiment, the server information update interval of each data distribution server 20 is all set to the same value. In the following description, it is assumed that each dispatcher 32 stores the value.

Next, a specific calculation example when (3) is applied will be described. Hereinafter, as an example, it is assumed that α ₁ = 0.8, α ₂ = 0.2, and server information update interval = 10 seconds. In the data distribution server 20-1, it is assumed that stream = 5, stream ₁ = 15, and stream ₂ = 25. Further, in the data distribution server 20-1, it is assumed that stream = 4, stream ₁ = 5, and stream ₂ = 6.

  Then, the following expression (4) is δ calculated 5 seconds after updatetime for the data distribution server 20-1 by the dispatcher 32A-1 under the above-mentioned conditions, and the following expression (5) is ( 4) new_stream calculated using δ calculated by equation (4).

  Further, the following expression (6) is δ calculated 5 seconds after updatetime for the data distribution server 20-2 by the dispatcher 32A-1 under the above-described conditions, and the following expression (7) is 6) new_stream calculated using δ calculated by equation (6).

  In the case of the above-described example, the data distribution server 20-1 becomes new_stream = 0 after 5 seconds as shown in the following expressions (4) and (5), and at this time, the data distribution server 20-1 Is not selected as a server to be allocated to the server allocation request.

On the other hand, the data distribution server 20-2 has a priority of data distribution server 20-1> data distribution server since new_stream = 3.5 after 5 seconds as shown in the following equations (6) and (7). The data distribution server 20-2 is also assigned to 20-2.

(B-3) Effects of Second Embodiment According to the second embodiment, the following effects can be achieved.

  The dispatcher 32A analyzes the trend of change of the stream value in the server load information of each data distribution server 20 based on the history of the server load information, and based on the analysis result, the current load status of each data distribution server 20 Is used for selection of the data distribution server 20. For example, in the first embodiment, if the estimation as in the second embodiment is not performed, even if the number of accesses of a certain data distribution server 20 has increased rapidly and has already reached the distribution limit, the data There is a problem that the distribution server 20 is selected.

(C) Third Embodiment Hereinafter, a third embodiment of a communication control device, a communication control program , and a data distribution system according to the present invention will be described in detail with reference to the drawings. In the third embodiment, the communication control device of the present invention is described as a dispatcher.

(C-1) Configuration of Third Embodiment A data distribution system 10B of the third embodiment can also be described with reference to FIG.

  In the data distribution system 10B of the third embodiment, the data distribution system 10 of the second embodiment is replaced with the data distribution system 10B. Further, the data distribution server 20 (20-1 to 20-4) of the second embodiment is replaced with the data distribution server 20B (20B-1 to 20B-4). Since it is the same as that of embodiment, detailed description is abbreviate | omitted.

  In the data distribution system 10B, the dispatcher 32A (32A-1 to 32A-3) is replaced with the dispatcher 32B (32B-1 to 32B-3), but other configurations are the same as those in the second embodiment. Detailed explanation is omitted here.

  In the data distribution server 20B, the content of the server load information given to the dispatcher 32B is scheduled to end the stream distribution for the stream currently being distributed by the data distribution server 20B in addition to the contents shown in FIG. Is different from the data distribution server 20 of the second embodiment in that information (hereinafter referred to as “stream distribution end scheduled information”) is included.

  FIG. 6 is an explanatory diagram showing an example of the content of the stream distribution end schedule information included in the server load information transmitted by the data distribution server 20B.

As shown in FIG. 6, the stream delivery end schedule information includes information on A _{1 to} A _j and K _{1 to} K _j . A _{1 to} A _j indicate how many seconds later from the time when the server load information is generated, and K _{1 to} K _j respectively indicate the data distribution server at the time related to A _{1 to} A _j. 20B shows the number of streams for which data distribution ends. For example, 1 is set in the K ₁ If the number of streams is one in which data delivery is terminated until after 0 seconds to A ₁ second, data distribution between ₁ second after A to ₂ seconds after the A termination If the number of streams to be performed is 2, K ₁ is set to 2.

_Although specific time set to A 1 to A _j is not intended to be limited, in this embodiment, for _example, it sets all in one second intervals A 1 to A _j (1 second, 2 seconds, 3 (Second,..., J seconds). Further, the value of j is not limited, but in this embodiment, the same value as the server information update interval is set. In the data distribution server 20B, the length (reproduction time) of the moving image data and the current transmission position (reproduction position) are grasped for each stream, and stream distribution end schedule information is based on the information. Is generated.

  The dispatcher 32B is different from the dispatcher 32 of the second embodiment only in the process of selecting the data distribution server 20 to be allocated in response to the server allocation request (the process of step S102 in FIG. 3 described above). Note that details of the process of selecting the data distribution server 20 to be allocated to the server allocation request in the dispatcher 32B will be described in detail in the description of the operation described later.

  Further, each dispatcher 32B accumulates server load information received from each data distribution server 20 as in the second embodiment.

(C-2) Operation of the Third Embodiment Next, the operation of the data distribution system 10B of the third embodiment having the above configuration will be described.

  The processing between the devices constituting the data distribution system 10B can also be shown in FIG. However, in the data distribution system 10B of the third embodiment, only the process of step S102 in FIG. 3 is different from that of the second embodiment, so only the process related to step S102 will be described below.

  The processing corresponding to step S102 described above in the third embodiment (selection processing of the data distribution server 20B) can also be shown using FIG. 5 described above. However, in the dispatcher 32B of the third embodiment, only the process of step S302 (new_stream calculation process) in FIG. 5 is different from that of the second embodiment, so only the process related to step S302 will be described below.

  First, in dispatcher 32B-1, the value of stream in the server load information of each data distribution server 20 is based on the history of server load information (including stream distribution end schedule information) received and accumulated from each data distribution server 20B. In the same way as in the second embodiment, the past change tendency is analyzed by obtaining δ by the above equation (2).

  Then, the dispatcher 32B estimates the number of streams that can be currently allocated to the data distribution server 20B by further applying δ obtained by the above equation (2) to the following equation (8).

In Expression (8), MAX is a function for calculating the value of new_stream, similar to Expression (3) above. For example, when expressed as MAX (a, b), if a ≧ b, a is returned as the result of MAX (a, b), and if a <b, b is returned as the result of MAX (a, b). Shall.

  In equation (8), deallocated is also a function for calculating the value of new_stream.

  FIG. 7 is an explanatory diagram showing the contents when the deallocated function process is written in a program language (for example, C language).

As shown in FIG. 7, when a time variable indicating time is input, the deallocated function returns K ₁ as a return value if time is greater than A ₁ in the stream distribution end scheduled information by an if statement. When K is larger than A _2, K ₁ + K ₂ is returned, and when time is larger than A _j , K ₁ + K ₂ +... + K _j is returned.

  That is, when the deallocated function inputs “now-updatetime”, stream distribution is performed in the data distribution server 20 from the time when the server load information was last generated in the data distribution server 20B to the current time (updatetime-now). It is possible to grasp the number of streams in which the process ends. In this embodiment, the deallocated function processing is shown by the program processing in FIG. 7, but the processing procedure is not limited as long as similar processing is possible.

  Thus, in the dispatcher 32B, as shown in the above equation (8), δ indicating a past change tendency of the stream value in the server load information of the data distribution server 20 and stream distribution in the data distribution server 20 are performed. New_stream is calculated using the number of streams to be terminated (deallocated (now-updatetime)).

Next, a specific calculation example when the above equation (8) is applied will be described. Hereinafter, as an example, it is assumed that α ₁ = 0.8, α ₂ = 0.2, and server information update interval = 10 seconds. In the data distribution server 20-1, it is assumed that stream = 5, stream ₁ = 15, and stream ₂ = 25. Further, in the data distribution server 20-1, it is assumed that stream = 4, stream ₁ = 5, and stream ₂ = 6.

  FIG. 8 is an explanatory diagram showing a specific example of the stream distribution end schedule information.

  It is assumed that the stream distribution end schedule information in the server load information of the data distribution server 20B-1 has the content as shown in FIG. 8A, and the stream distribution end schedule information in the server load information of the data distribution server 20B-1 is illustrated in FIG. It is assumed that the content is as shown in 8 (b).

  In this case, after 5 seconds from updatetime, new_stream = 0 of the data distribution server 20B-1 becomes 0 and allocation becomes impossible. However, since the data distribution server 20B-2 has several streams that end, new_stream = 0 does not become allocated. Is possible.

(C-3) Effects of Third Embodiment According to the third embodiment, the following effects can be obtained.

  The dispatcher 32B estimates the current load status of each data distribution server 20B using the stream distribution end schedule information. Thereby, compared with 2nd Embodiment, the estimation of a load condition with higher precision can be performed. For example, in the second embodiment, even if the load change trend of a certain data distribution server 20 is an upward trend in the past, if the stream distribution end schedule information includes a large number of streams scheduled for distribution end, the actual allocation is performed. Although possible, it may be determined that assignment is impossible.

(D) Other Embodiments The present invention is not limited to the above-described embodiments, and may include modified embodiments as exemplified below.

(D-1) In each of the above embodiments, in the stream distribution from the data distribution server to the client, the bit rate of each stream may be a different bit rate. In this case, the control signal such as a server allocation request may include information on the bit rate of data requested by each client, or the data distribution server may register the bit rate for each client in advance. You may make it leave.

  When the bit rate of each stream is variable, the unit of the stream parameter of the server load information is not the number of streams as shown in FIG. 2, but a unit indicating other surplus processing capacity is applied. Anyway.

  For example, 10 units of processing capacity may be required for one 1000 kbps stream, and 5 units of processing capacity may be applied to one 500 kbps stream. In this case, for example, assuming that an arbitrary data distribution server 20 has a processing capacity of 50 units as a whole, this data distribution server 20 can simultaneously distribute 10 streams for 500 kbps streams and 5 streams for 1000 kbps streams. It will be possible. When the data distribution server 20 is distributing three 1000 kbps streams, the stream parameter in the server load information is 50− (10 × 3) = 20 units.

(D-2) In each of the above embodiments, the dispatcher performs processing for selecting the data distribution server to be allocated to the server allocation request in consideration of the priority, but does not consider the priority. Also good. For example, in the first embodiment, the processing of step S203 to step S205 is performed in the order of priority assigned to the data distribution server. However, a random order, a fixed order that does not take priority into account, or identification information May be performed in an order determined by other criteria, such as an order in which the keys are arranged in ascending order.

(D-3) In each of the embodiments described above, a plurality of dispatchers are arranged in the data distribution system, but the number of dispatchers may be one.

(D-4) In each of the above embodiments, the server load information includes information on updatetime (the time when the server load information was generated), but this is omitted, and on the dispatcher side, The time when the server load information is received may be replaced with updatetime and applied.

  DESCRIPTION OF SYMBOLS 10 ... Data delivery system, 20, 20-1 to 20-4 ... Data delivery server, 30 ... Communication control system, 31 ... Proxy, 32, 32-1 to 32-3 ... Dispatcher, 40, 30-1 to 30- N ... Client.

Claims (7)

In a communication control device that controls communication between a client device and a plurality of data distribution servers that perform stream distribution of distribution data to the client device,
For each data distribution server, server load information acquisition means for acquiring server load information related to the current load status;
Server allocation request receiving means for receiving an allocation request for allocating a data distribution server from the client device;
Using the server load information for each data distribution server acquired in the past by the server load information acquisition means, the current load status of the data distribution server is estimated, and the estimation result is taken into consideration in the allocation request. Server selection means for selecting a data distribution server to be assigned to the server;
A data distribution server selected by the server selection unit and a communication control unit for controlling the client device so as to distribute distribution data from the data distribution server selected by the server selection unit to the client device; ,
The server load information includes stream distribution end schedule information related to the scheduled end time of the stream currently being distributed by the data distribution server when the server load information is generated in the data distribution server,
The server selection means also estimates the current load status of the data distribution server using the stream distribution end schedule information included in the server load information acquired from the data distribution server in the past by the server load information acquisition means. A communication control device.   2. The communication control apparatus according to claim 1, wherein the server load information includes information related to a time when the server load information is generated.   The server selection means analyzes the fluctuation tendency related to the load of the data distribution server based on the history of server load information acquired from the data distribution server in the past, and uses the analysis result to determine the data distribution server's load. The communication control apparatus according to claim 1, wherein a current load situation is estimated. A computer mounted on a communication control device that controls communication between a client device and a plurality of data distribution servers that perform stream distribution of distribution data to the client device,
For each data distribution server, server load information acquisition means for acquiring server load information related to the current load status;
Server allocation request receiving means for receiving an allocation request for allocating a data distribution server from the client device;
Using the server load information for each data distribution server acquired in the past by the server load information acquisition means, the current load status of the data distribution server is estimated, and the estimation result is taken into consideration in the allocation request. Server selection means for selecting a data distribution server to be assigned to the server;
Functioning as a data distribution server selected by the server selection unit and a communication control unit for controlling the client device so as to distribute distribution data from the data distribution server selected by the server selection unit to the client device;
The server load information includes stream distribution end schedule information related to the scheduled end time of the stream currently being distributed by the data distribution server when the server load information is generated in the data distribution server,
The server selection means also estimates the current load status of the data distribution server using the stream distribution end schedule information included in the server load information acquired from the data distribution server in the past by the server load information acquisition means. A communication control program characterized by the above. The communication control program according to claim 4, wherein the server load information includes information related to a time when the server load information is generated. The server selection means analyzes the fluctuation tendency related to the load of the data distribution server based on the history of server load information acquired from the data distribution server in the past, and uses the analysis result to determine the data distribution server's load. The communication control program according to claim 4 or 5, wherein a current load situation is estimated. A data distribution system having a client device, a plurality of data distribution servers that distribute distribution data to the client device, and one or a plurality of communication control devices that control communication between the client device and the data distribution server In
Each of the above data distribution servers
Server load information generating means for generating server load information having information relating to the current load status of the data distribution server;
Server load information transmitting means for giving the server load information generated by the server load information generating means to the communication control device;
A data distribution system, wherein the communication control device according to claim 1 is applied as each of the communication control devices.

Images