JP2016045802A - Information processing apparatus, load sharing system, load sharing method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an information processing apparatus which reduces the burden of a user and enables flexible and simple load sharing.SOLUTION: An information processing apparatus operates as follows: acquiring held distribution-destination identification information; transferring a request to a second information processing apparatus identified on the basis of the acquired distribution-destination identification information when it is determined that the request is to be distributed to other information processing apparatus connected over a network, on the basis of the amount of processing data included in the received request; controlling the processing data requested by the request and included in the request not to be executed when it is determined that the request is to be distributed to the other information processing apparatus; and controlling the processing data requested by the request and included in the request to be executed when it is determined that the request is not to be distributed to the other information processing apparatus.SELECTED DRAWING: Figure 1

Description

  The present invention relates to an information processing apparatus, a load distribution system, a load distribution method, and a program.

  2. Description of the Related Art In recent years, a system that causes an external server connected via a predetermined network to execute processing requested from a client is known. In such a system, when a request with a high processing load such as generation of image data is made from a client, a large amount of requested server-side resources are consumed.

  In addition, when a request with a high processing load as described above is made from multiple clients at the same time, the server resources are limited, so the server processing capacity cannot catch up and processing takes time. Will slow down the response.

  Therefore, in recent years, there is known a load distribution system in which a plurality of servers are arranged and a load is distributed by distributing requests from clients among the plurality of servers. Examples of general distributed configurations include a method in which front-end servers receive requests and distribute processing in a round-robin manner, and a method that monitors the load status of each server and distributes processing to servers with low load. Are known.

  For example, Patent Document 1 discloses a technique in which a load distribution device such as a so-called load balancer is provided between a plurality of clients and a plurality of servers. The load balancer estimates the amount of load that the service request gives to the server from the header information of the service request packet, and manages the estimated value for each allocated server, so that the load state of the server is real-time. Based load balancing.

  By the way, the load distribution system as disclosed in Patent Document 1 requires a dedicated server such as the above-described load distribution apparatus for load distribution in addition to a server that actually performs processing.

  However, when a separate server for load distribution is provided later, for example, it is necessary not only to set the load distribution server itself but also to perform settings for distributed processing on each server connected to it. Comes out. Therefore, in such a case, it takes time to introduce the system, for example, by stopping the system and requiring a major system change. Moreover, it is difficult to perform switching such as load distribution or no load distribution as necessary.

  The present invention has been made in view of such circumstances, and an object thereof is to reduce the burden on the user and to realize load distribution flexibly and easily.

  In order to solve the above-described problem, the first information processing apparatus of the present invention includes a request receiving unit that receives a request transmitted from a client, and a control that controls execution of a process requested by the request received by the request receiving unit. And distribution destination identification information for identifying a second information processing apparatus that is a distribution destination for distributing the request received by the request reception means, and acquiring the distribution destination identification information held by the distributed processing information holding means Whether to distribute the request to other information processing devices connected via the network based on the data amount of the processing data included in the request received by the distributed processing information acquisition unit and the request reception unit. A dispersion determination unit that determines the request to another information processing apparatus by the dispersion determination unit. And a transfer means for transferring the request to the second information processing apparatus identified based on the distribution destination identification information acquired by the distributed processing information acquisition means, When the distribution determination unit determines that the request is distributed to other information processing devices, control is performed so as not to execute the processing for the processing data included in the request requested by the request. When it is determined that the request is not distributed to other information processing apparatuses, control is performed so as to execute processing on the processing data included in the request requested by the request.

  According to the present invention, it is possible to reduce the burden on the user and realize load distribution flexibly and easily.

It is the schematic of the load distribution system containing the information processing apparatus in embodiment of this invention. It is a functional block diagram of an information processor in an embodiment of the present invention. It is a sequence diagram explaining the dispersion | distribution determination processing procedure in 1st Embodiment of this invention. It is a sequence diagram explaining the dispersion | distribution determination processing procedure in 2nd Embodiment of this invention. It is a sequence diagram explaining the dispersion | distribution determination processing procedure in 3rd Embodiment of this invention. It is a sequence diagram explaining the dispersion | distribution determination processing procedure in 3rd Embodiment of this invention. It is a sequence diagram explaining the dispersion | distribution determination processing procedure in 4th Embodiment of this invention. It is a functional block diagram of the information processing apparatus in 5th Embodiment of this invention. It is a sequence diagram explaining the dispersion | distribution determination processing procedure in 5th Embodiment of this invention. It is a sequence diagram explaining the dispersion | distribution determination processing procedure in 6th Embodiment of this invention. 1 is a schematic diagram of a system using a print server as a specific example to which an embodiment of the present invention is applied.

  An information processing apparatus according to an embodiment of the present invention will be described below with reference to the drawings. However, the present invention is not limited to the present embodiment as long as the gist of the present invention is not exceeded. In addition, in each figure, the same code | symbol is attached | subjected to the part which is the same or it corresponds, The duplication description is simplified thru | or abbreviate | omitted suitably.

  An outline of a load distribution system including an information processing apparatus according to an embodiment of the present invention will be described with reference to FIG. The load distribution system in the present embodiment includes, for example, a client 10, a server 20, a server 30, and a storage 31. In the present embodiment, each server may have both a role of receiving a request from a user terminal and a role of distributing a request to other servers as distribution destinations. That is, when a server distributes a request to another server that is a distribution destination, the server that is the distribution source may be called a client. Further, in the present embodiment, “distributing requests” refers to transferring an accepted request to another device and causing the other device to execute a process requested by the request.

  The client 10, the server 20, the server 30, and the storage 31 are connected by a network 51, a network 52, and a network 53. These networks may be constructed by, for example, a LAN (Local Area Network) or a WAN (Wide Area Network), and may be a wireless network as well as a wired network.

  The client 10 is, for example, an information terminal such as a personal computer (hereinafter referred to as “PC”) used by a user to make a processing request (hereinafter referred to as “request”) to the server 20. Therefore, the client 10 is not limited to a PC, and may be a smartphone, a mobile phone having an information processing function, or a portable information terminal such as a PDA (Personal Digital Assistant), or an information home appliance such as a television having an information processing function. It may be.

  In the present embodiment, the number of clients 10 is one. However, for convenience of illustration, it goes without saying that requests are sent from a plurality of clients on the system.

  The server 20 is an information processing apparatus such as a PC that accepts a request from the client 10 by means such as http communication via the network 51. For example, the server 20 provides its own functions, services, and data to a computer such as the client 10. Hardware.

  Since the server 30 is a PC or the like having substantially the same configuration as the server 20, description thereof will be omitted. In the present embodiment, the server 20 will be described below as a distribution source and the server 30 as a distribution destination. In the present invention, the server 20 is the first information processing apparatus and the server 30 is the second information processing apparatus, but these are defined formally for convenience of explanation. That is, since all apparatuses have substantially the same configuration, they are defined for distinction. That is, for example, the server 20 or the server 30 as the information processing apparatus in the present embodiment may have a necessary and sufficient configuration for determining the distribution.

  In the present embodiment, two servers 20 and 30 are provided. However, the load distribution system according to the present invention is not limited to two servers as information processing apparatuses. Needless to say, the above is also true. Further, when expressing the present invention as a load distribution system, for example, the distribution source server is referred to as a “first information processing apparatus”, the distribution destination server is referred to as a “second information processing apparatus”, and thereafter the distribution destination , And “nth information processing device” are formally defined according to the number of servers.

  For example, in the present embodiment, when the server 20 determines that processing is distributed to the server 30, the request sent from the client 10 to the server 20 includes access data, a clock, and the like for accessing the server 30. Except for this, it is transferred to the server 30 via the network 52 as it is.

  The storage 31 is a storage device such as a hard disk that stores processing data, bibliographic information, and the like included in a request sent from the client 10 via the server 20 in the server 30. Processing data, bibliographic information, and the like may be directly stored in a predetermined database.

  In the present embodiment, the storage 31 and the server 30 are connected via the network 53, but the storage 31 is not limited to being accessible only by the server 30 and can be accessed from the server 20. Needless to say, it may be. In the present embodiment, the storage 31 has a configuration independent of the server 20 and the server 30, but may be a configuration in which the storage 31 is built in the server 30, for example.

  Next, for example, functional blocks of the server 20 as the information processing apparatus in the present embodiment will be described with reference to FIG. Since the server 30 has substantially the same configuration as that of the server 20, this figure is also applied to the server 30.

  The server 20 in the present embodiment includes a management unit 21, a distribution determination unit 22, a transfer unit 23, a distributed processing information holding unit 24, a request analysis unit 25, a response analysis unit 26, a request processing unit 27, A control unit 28 is provided.

  The management unit 21 includes a request reception unit 211, a distributed processing information acquisition unit 212, a response reception unit 213, and a response unit 214. The request reception unit 211 is a request reception unit that receives a request transmitted from the client 10. The distributed processing information acquisition unit 212 is distribution destination identification information for identifying the server 30 that is a distribution destination for distributing the request received by the request reception unit 211, and the distribution destination identification held by the distributed processing information holding unit 24 It is a distributed processing information acquisition means for acquiring information. The response receiving unit 213 is a response receiving unit that receives a response returned from the second information processing apparatus in response to the request transferred to the server 30 by the transfer unit 23. The response unit 214 is a response unit that responds to the request to the client 10 that has transmitted the request based on the response received by the response reception unit 213.

  The distribution determination unit 22 determines whether to distribute the request to other information processing devices connected via the network based on the data amount of the processing data included in the request received by the request reception unit 211. It is a dispersion | distribution determination means to determine. Further, instead of the data amount, for example, the distribution determination may be performed using a CPU usage rate or a memory usage rate. In other words, equipment load can be measured quantitatively, such as CPU usage including the above CPU usage, memory usage including memory usage, and network usage including network usage. The determination of variance may be performed using. In the present embodiment, the network usage is indicated by the amount of data per unit time transmitted through the network between the server and the storage or database, for example.

  For example, the distribution determination unit 22 may determine that the request is distributed to other information processing apparatuses when the resource usage state is equal to or greater than a preset threshold. In the present embodiment, the variance determination unit 22 may internally include table data shown in the following table as a predetermined threshold value, and may refer to this table data when determining the variance.

  To explain with reference to Table 1, for example, when the “CPU usage rate” as the CPU usage status is used as a key for the distribution determination, the distribution determination unit 22 determines the CPU usage rate based on the threshold value in the table data of Table 1. When it is 90% or more, it is determined that distributed processing is performed. Similarly, when the “memory usage rate” as the memory usage status is used as a key for the distribution determination, when the memory usage rate is 90% or more based on the threshold value in the table data of Table 1, the distributed processing is performed. judge.

  Further, similarly, when “network usage” as the network usage status is used as a key for distribution determination, it is determined that distributed processing is performed when the network usage is 10 MBps or more based on the threshold value in the table data of Table 1. To do. That is, in the present embodiment, the distribution determination unit 22 determines that the server 20 has a high load and distributes the load to the server 30 when the resource usage state is equal to or greater than a preset threshold.

  Further, the distribution determination unit 22 may determine that the request is distributed to the servers 30 when the data amount of the processing data included in the request is equal to or greater than a predetermined data amount threshold. The “data amount of processing data included in the request” is, for example, data amount information in header information included in the request. In the present embodiment, the distribution determination unit 22 may refer to the table data shown in the following table as a data amount threshold value together with the resource usage status described above or instead of the resource usage status when determining the distribution.

  Describing with reference to Table 2, for example, when using the “data amount information in the header” as the key for the distribution determination, the distribution determination unit 22 uses the header based on the data amount threshold in the table data of Table 2. When the internal data amount information is 5 Mbytes or more, it is determined that the distributed processing is performed.

  If the same data amount threshold value is set in all information processing apparatuses at the distribution destination, a request having data amount information equal to or greater than the data amount threshold value is not processed permanently through all the distribution destinations. Occurs. In order to avoid this phenomenon, in the present embodiment, for example, 1 MBytes may be added at the data amount threshold held at the distribution destination for each distribution.

  When the distribution determination unit 22 determines that the request is distributed to other information processing apparatuses, the transfer unit 23 is identified based on the distribution destination identification information acquired by the distributed processing information acquisition unit 212. Transfer means for transferring to the server 30. Even in the server 30 that has received the request, the same distribution determination as that of the server 20 is performed. It is determined whether or not the distribution is performed based on the resource status and the like. Good. On the other hand, if the load is heavy, for example, the request may be transferred to a distribution destination instead of the server 30.

  The distributed processing information holding unit 24 is a distributed processing information holding unit that holds and stores a server IP address or the like as distributed destination identification information.

  The request analysis unit 25 analyzes the content of the received request, adds the IP address of the distribution destination server 30 to the header information, for example, and performs reconfiguration as a request from the server 20 to the server 30.

  The response analysis unit 26 analyzes the response content sent from the server 30, edits header information, for example, and reconfigures it as a response of the server 20 itself to the client 10. The request processing unit 27 performs a process requested by the received request and generates a response.

  When the distribution determination unit 22 determines that the request is distributed to other information processing apparatuses, the control unit 28 performs control so that processing for the processing data included in the request requested by the request is not executed. Control means. In addition, when the distribution determination unit 22 determines that the request is not distributed to other information processing apparatuses, the control unit 28 performs processing on the processing data included in the request requested by the request. To control.

  Based on the above, an outline of the operation of the server 20 in the present embodiment will be described. The management unit 21 receives a request request from the client or the server and makes a distribution determination request to the distribution determination unit 22. When performing distributed processing, the request contents and distribution destination information are transmitted to the request analysis unit 25 and a newly created request is received.

  The management unit 21 confirms the existence of the request with the distribution destination server 30 obtained from the distributed processing information holding unit 24, and then sends the request to the server 30 when the existence is confirmed. When the management unit 21 cannot confirm the survival, the management unit 21 requests the request processing unit 27 to perform processing. The management unit 21 sends the received response to the response analysis unit 26, receives the newly created response, and sends it to the client or server.

  Whether or not the server 30 is “alive” is confirmed based on the operating state of the server 30. For example, if the server 30 is powered on and is operating to a level where requests can be distributed, the management unit 21 determines that the server 30 is “alive”. Good. For example, a test request may be transmitted from the server 20 to the server 30, and the existence of the server 30 may be determined based on the response that has been returned. In addition, a Ping command may be used for confirming the existence of the server 30. In order to determine that the server is “alive”, the server 20 and the server 30 may be connected via a network, and the distributed processing program may be operating normally.

  On the other hand, when the server 30 is powered off, when the server 30 is out of order, or when the distributed processing program according to the present invention is not installed in the first place, the management unit 21 controls the server 30. Determines that it is not alive. For example, when the management unit 21 determines that the server 30 is not alive, the management unit 21 may not perform the distribution determination. In correspondence with the above-described example, for example, when there is no response response from the server 30, or when no Ping is returned, it may be determined that the server 30 is not alive. Furthermore, it may be determined that the server 20 and the server 30 are not connected via a network and that the distributed processing program is not operating normally is “not alive”.

[First Embodiment]
Next, the dispersion determination processing procedure in the first embodiment of the present invention will be described with reference to FIG. First, the client 10 sends a request to the management unit 21 using a protocol such as http, and receives a response from the management unit 21 [1].

  Next, the management unit 21 requests the distribution determination unit 22 to perform distribution determination [1.1]. Then, the distribution determination unit 22 determines whether or not to distribute the request to another information processing apparatus based on the data amount of the processing data included in the request [1.1.1].

  When the distribution determining unit 22 determines not to perform distributed processing, the management unit 21 requests the request processing unit 27 to perform request processing [1.2].

  On the other hand, when the distribution determining unit 22 determines that the distributed processing is to be performed, the management unit 21 requests the distributed processing information holding unit 24 to acquire information of the distribution destination server 30. The information of the distribution destination server 30 may use IP addresses as shown in the following table, for example.

  The management unit 21 accesses the server 30 based on the acquired IP address and checks whether the server 30 is alive [1.4]. When survival is confirmed, distributed processing is continued.

  Next, the management unit 21 sends the request content from the client 10 and the information of the distribution destination server 30 acquired by the distributed processing information holding unit 24 to the request analysis unit 25, and requests the request creation [1. .5]. The request analysis unit 25 analyzes the request [1.5.1], creates a request in the server 20 [1.5.2], and returns it to the management unit 21.

  The management unit 21 executes a request to the distribution destination server 30 based on the returned request content [1.6].

  Thereafter, when receiving a response from the server 30, the management unit 21 requests the response analysis unit 26 to create a response [1.7]. The response analysis unit 26 analyzes the response from the server 30 [1.7.1] and reconfigures it as the response of the server 20 [1.7.2].

  As described above, according to the first embodiment, when distributed processing is required, for example, a server 30 having substantially the same configuration as that of the server 20 that is the distribution destination need only be set as the distribution destination of the server 20. It is possible to realize load distribution flexibly and easily. Note that the substantially identical configuration is sufficient as long as the software that executes the distributed processing in this embodiment operates on each server, and does not prevent other software configurations and hardware configurations from being different. On the other hand, for convenience, the hardware configuration and software configuration of the apparatus may be the same.

[Second Embodiment]
Next, the dispersion determination processing procedure in the second embodiment of the present invention will be described with reference to FIG. Note that description of procedures similar to those of the first embodiment is omitted. In the second embodiment, the distribution determination is performed with reference to the in-header data amount information as the attribute information included in the request shown in Table 2 together with or separately from the resource status. The attribute information includes, for example, time information, processing designation information, and the like, but description thereof is omitted here.

  When receiving a distribution determination request from the management unit 21 [1.1], the distribution determination unit 22 starts the distribution determination process [1.1.1]. The distribution determination unit 22 acquires a data amount threshold such as 5 MBytes from the distributed processing information holding unit 24 in addition to the resource usage status [1.1.1.1]. Thereafter, the distribution determining unit 22 acquires the data amount information in the header included in the request from the request analyzing unit 25 [1.1.1.2]. Then, the distribution determination unit 22 compares the data amount in the header information with the data amount threshold, and determines that the distribution is performed when the data amount in the header information is equal to or greater than the data amount threshold [1.1.1. .3].

  The distribution determination may be performed by combining the resource status and the amount of data in the header information. For example, a correspondence relationship in which 10% of the CPU usage rate corresponds to 1 MB of the data amount is set in advance. When a conversion unit (not shown) determines the resource usage and the data amount in combination, first, the CPU usage rate is converted into the data amount based on the correspondence. Subsequently, the converted data amount that is the conversion result is added to the actual data amount, and the dispersion determination is performed based on whether or not the data amount after the addition is equal to or greater than the data amount threshold value. For example, when the resource status is 50% or more and the determination is performed only with this, the distribution determination is not performed. On the other hand, by combining the data amount in the header information, when the data amount is greater than or equal to the threshold value, it can be confirmed in advance that processing cannot be performed by the server that performs processing, so the load can be distributed more efficiently. It becomes possible. On the other hand, for example, the amount of data of 1 MB is converted into a CPU usage rate of 1 MB, etc., and added to the actual CPU usage rate, and a dispersion determination is performed based on whether the CPU usage rate after the addition is equal to or greater than the threshold shown in Table 1. Also good.

[Third Embodiment]
Next, the dispersion determination processing procedure in the third embodiment of the present invention will be described with reference to FIG. In addition, description is abbreviate | omitted about the procedure similar to 1st Embodiment and 2nd Embodiment.

  In the third embodiment, the distribution determination unit 22 acquires the data amount based on the attribute information included in the request. That is, the data amount is acquired from the data amount information in the header (hereinafter referred to as “content length”) as the attribute information described in the second embodiment. If it is determined that the acquired content length is distributed, the processing data related to the content length (hereinafter referred to as “content data”) is deleted from the request, and the transfer unit 23 transmits the processing data to, for example, the storage 31 illustrated in FIG. . Then, the transfer unit 23 transmits access information for reading processing data stored from the storage 31 to the server 30. The access information refers to, for example, a URL (hereinafter referred to as “content URL”) indicating a storage location of processing data in the storage 31.

  When the distribution determination unit 22 receives a distribution determination request from the management unit 21 [1.1], the distribution determination unit 22 performs determination processing [1.1.1]. At that time, the distribution determination unit 22 acquires the content length included in the request through the request analysis unit 25 [1.1.1.1]. If the distribution determination unit 22 determines to perform distribution determination based on the content length, the distribution determination unit 22 acquires content data from the request through the request analysis unit 25 and deletes the content data from the request [1.1.1.2]. In FIG. 5, for convenience in illustration, the content length is abbreviated as “c-length” and the content data is abbreviated as “c-data”.

  Since the content length is decreased by deleting the content data from the request, the distribution destination determines whether or not to further distribute based on the changed content length.

  When the distribution determination unit 22 determines to perform distribution determination, the management unit 21 transmits content data to the storage 31 via the transfer unit 23 and acquires a content URL from the storage 31 [2.2]. . In FIG. 5, the content URL is abbreviated as “c-URL” for convenience of illustration.

  The management unit 21 requests the request analysis unit 25 to create a request [2.3]. The request analysis unit 25 sets the content URL acquired by the management unit 21 as the content URL of the distribution request header [2.3.2]. The management unit 21 transmits a content URL when transferring a request to the server 30 via the transfer unit 23 as in the first embodiment [1.6]. On the other hand, when the distribution determination unit 22 determines not to perform distribution, the management unit 21 requests the request processing unit 27 to perform request processing [2.1].

  The above describes the process when the distributed server is the primary machine. The process when the distributed server is not the primary machine and the distribution source stores the content data in the storage or database is described with reference to FIG. I will explain.

  When the distribution server determines that the distribution determination unit 22 does not perform the distribution determination, the distribution server performs the processing for the content data by itself. Therefore, access information for acquiring content data is acquired from the content URL of the request header [1.1.1.2]. On the other hand, when it is determined by the distribution determining unit 22 that the content data is distributed, the content data is already stored in the storage or the like and is not included in the request, and thus does not need to be acquired. In addition, the content data is acquired by a server that performs processing without being distributed. Then, the management unit 21 of the distributed server can determine that the content data is stored in the storage 31 from the acquired content URL. For this reason, content data is acquired from the storage 31 using the acquired content URL [2.1] and processed in its own request processing unit 27 [2.2].

  As described above, according to the third embodiment, data with a large data capacity and heavy processing, that is, data with a large apparatus load is stored in an external storage device such as the storage 31, and the content URL is excluded from the distribution destination server except for the content data. Set the request header and execute the request. As a result, for example, because the load is high even at the server of the distribution destination, every time distributed processing is performed, such as further distribution to the destination, data with a large device load is distributed to the distribution destination. Can be avoided.

[Fourth Embodiment]
Next, the dispersion determination processing procedure in the fourth embodiment of the present invention will be described with reference to FIG. In addition, description is abbreviate | omitted about the procedure similar to each embodiment mentioned above. In the present embodiment, the distributed processing information holding unit 24 holds the distribution upper limit value, which is the upper limit of the number of times that the request can be transferred in total in the distributed system. As the dispersion upper limit value, for example, those shown in the following table may be used.

  When the distribution determination unit 22 determines to distribute the request to the server 30, for example, “1” is added to the number of distributions included in the request, and the added distribution number matches the distribution upper limit value. It is determined that the request is not distributed to the server 30.

  When the distribution determination unit 22 receives a distribution determination request from the management unit 21 [1.1], the distribution determination unit 22 acquires a distribution upper limit value, for example, “5” in Table 2 from the distributed processing information holding unit 24 [1.1.2]. .

  When making a request to the distribution destination server, the distribution determination unit 22 adds the number of distributions to the number of distributions included in the header and compares it with the obtained distribution upper limit [1.1.5]. At this time, the distribution determination unit 22 determines that the distribution is not performed if the number of distributions matches the distribution upper limit value, and the request processing unit 27 processes the request in its own device. On the other hand, if the number of times of dispersion is equal to or less than the upper limit of dispersion, the dispersion determination unit 22 determines to perform dispersion processing.

  According to the fourth embodiment, it is possible to determine whether the device itself accepts a request from a client or a request as a distributed processing destination by checking the header. For example, when the number of distributions is 0, it may be determined that a request is received from a client. When functioning as a primary machine that receives a request from a client, all servers may have the function to determine whether or not the machine is a primary machine. For example, if a request requires processing to store large data included in the request in the storage, distributing requests containing large data may increase the load on the network. Such processing may be performed by the primary machine. More specifically, for example, when a request requests a process for storing data whose size is greater than or equal to the data amount threshold in the storage, the request is executed on the primary machine without distributing the request. Good.

[Fifth Embodiment]
Next, a dispersion determination processing procedure according to the fifth embodiment of the present invention will be described with reference to FIGS. In addition, description is abbreviate | omitted about the structure similar to each embodiment mentioned above, and the same procedure. In this embodiment, a timeout check unit 32 is provided in addition to the block diagram shown in FIG. The timeout check unit 32 is a timeout check unit that determines whether or not a measurement time calculated after the management unit 21 receives a request has passed a predetermined time. The predetermined time is a threshold time for performing timeout determination, and may be a time-out value such as 10 seconds, for example.

  When the time-out check unit 32 determines that the measurement time has passed the predetermined time, the management unit 21 stops the distributed processing. That is, the state waiting for the response of the request transferred to the distribution destination is canceled.

  First, the management unit 21 extracts the request input time from the request content from the client 10, and measures the measurement time from the request input by a timer (not shown), for example. Then, the management unit 21 requests the timeout check unit 32 to perform a timeout check [1.8]. The timeout check unit 32 checks the timeout value held in the apparatus or the like, and determines whether or not the measurement time exceeds the timeout value [1.8.1].

  When the timeout check unit 32 receives a determination that the timeout value has been exceeded, the management unit 21 stops the distributed processing and stops waiting for a response [1.9]. Then, the management unit 21 requests the request processing unit 27 for request processing [1.10].

  According to the fifth embodiment, when processing is performed at the distribution destination, for example, when an error occurs and the processing does not end or cannot respond to the distribution source, the distribution source returns a response after a certain period of time. Responsiveness can be improved by canceling the wait state and performing processing. That is, according to the fifth embodiment, even in such a case, referring to the timeout value, if the measurement time in the distributed processing exceeds the timeout value, the distribution is not performed at that time, and the processing is performed in any server. It may be executed.

[Modification]
Next, a modification of the fifth embodiment of the present invention will be described with reference to FIG. In addition, description is abbreviate | omitted about the procedure similar to each embodiment mentioned above. In the present modification, unlike the fifth embodiment, the management unit 21 may notify the client 10 of an error when making a timeout determination.

  That is, when it is determined that the measurement time exceeds the timeout value, the management unit 21 returns an error notification to the client 10 [1.11]. As a result, the user can recognize that the server cannot perform distributed processing.

  Each of the above-described embodiments is a preferred embodiment of the present invention, and various modifications can be made without departing from the scope of the present invention. For example, each process in the information processing apparatus of the present embodiment described above can be executed using hardware, software, or a combined configuration of both.

  In the case of executing processing using software, it is possible to install and execute a program in which a processing sequence is recorded in a memory in a computer incorporated in dedicated hardware. Alternatively, the program can be installed and executed on a general-purpose computer capable of executing various processes.

  In each of the embodiments described above, the distributed processing information holding unit 24 holds the IP address of the distribution destination server and refers to the distributed processing information holding unit 24 when performing distributed processing. The distribution destination information such as the IP address may be separately stored in an external server different from the distribution destination server. As a result, for example, information on the distribution destination such as an IP address can be centrally managed.

  Also, when a plurality of distribution destination servers are set, for example, when managing the IP addresses of all of the servers collectively with an external server and performing distributed processing, when referring to the external server, for example, An IP address may be acquired as server information of a distribution destination using a round robin method.

  Alternatively, the resource status of each server may be centrally managed by an external server, and the distribution destination server may be determined by referring to the resource status stored in the external server when determining the distribution.

  Further, the present embodiment described above is preferably applied to a pull print system using a print server as shown in FIG. In this system, a client 100, a print server 200, a print server 300, a storage 310, and a storage 320 are connected by a network 510, a network 520, a network 530, and a network 540, respectively.

  In this example, as a request from the client 100, for example, “save print job”, “get print job list”, “change print settings”, “generate preview image”, “get print job” and so on.

  The internal processing executed by the print server includes, for example, “print job, print job attribute storage”, “print job parsing or editing”, “print job parsing and image data generation”, “print job “List acquisition”, “Print job acquisition”, and the like. For example, bibliographic information such as print job attribute information is stored in the storage 310, and a print job and preview image data are stored in the storage 320. The print server acquires these pieces of information when necessary and executes processing.

  In such a case, when deciding the upper limit number of clients that each print server can process smoothly, how much parallel print job parsing, editing, and image data generation in the server is a CPU-intensive process It is important to be able to do it.

  For example, when processing such as the above is performed in parallel for 100 processes, if the CPU usage rate of the print server is always 100%, the response may be delayed. Therefore, it is necessary to take measures to increase the number of clients by avoiding, for example, the suspension of processing, which is an incentive for this.

  For example, there is a measure such that processing is performed in a queue so that parallel processing cannot be performed for a certain number or more, but in this case, the waiting time of the user becomes long. On the other hand, if this embodiment is applied, the load on the CPU in the print server can be distributed, and the upper limit number of clients can be easily increased.

  For example, when a print job as a request from the client 100 is input to the print server 200 in the raw data format, the print server 200 performs distribution determination to which the present embodiment is applied.

  If the print server 200 determines that the process for the print job is to be distributed according to the distribution determination, the print server 200 confirms the existence of the print server 300 next. When the print server 200 confirms the existence of the print server 300, the print server 200 inputs a print job to the print server 300.

  When the print server 300 also performs the distribution determination process to which the present embodiment is applied and determines that the distribution is not performed, the print server 300 generates a preview image from the print job and executes a process of storing the preview image and the print job in the storage 320. Further, the print server 300 executes processing for extracting bibliographic information of a job from API (Application Programming Interface) parameters and print jobs that the client calls when the print job is input and storing them in the storage 310. Then, the processing result is returned to the client 100 via the print server 200.

  On the other hand, the print server 200 may determine not to distribute the processing for the print job by the distribution determination based on the internal processing to be executed. Here, the job data is generally a job control language command portion such as a printer job language (hereinafter referred to as “PJL”) command for controlling a print job, and a page description language format such as a postscript or a printer control language. Print data main body.

  For example, when internal processing for editing attributes such as the number of copies, color, and both sides of a print job is executed, the PJL portion of job data is edited while the print data body is not edited. Good.

  The PJL portion of job data generally has a significantly smaller data amount than the job data body. The processing for the PJL portion of the job data consumes less resources than the processing for the job data main body and can be processed in a short time, so that a large load is not applied to the print server.

  Furthermore, when the amount of data in the job data body is large, the job data body is sent to the distribution destination along with the request, so the load on the distribution source due to the request transmission, the load on the distribution destination due to the reception of the request, or the network due to data transfer There is a possibility that the load on Therefore, when internal processing for editing the attributes of a print job is executed, it is desirable to execute the processing by the own apparatus without performing the distributed processing by the distribution determination.

10, 100 Client 20, 30 Server 21 Management unit 22 Distribution determination unit 23 Communication unit 24 Distributed processing information holding unit 25 Request analysis unit 26 Response analysis unit 27 Request processing unit 31, 310, 320 Storage 32 Timeout check unit 51, 52, 53, 510, 520, 530, 540 Network 200, 300 Print server

JP 2002-91936 A

Claims (13)

A request receiving means for receiving a request sent from a client;
Control means for controlling execution of processing requested by the request received by the request receiving means;
The distribution destination identification information for identifying the second information processing apparatus that is the distribution destination that distributes the request received by the request reception means, and acquires the distribution destination identification information held by the distributed processing information holding means Distributed processing information acquisition means for
A distribution determination unit that determines whether to distribute the request to other information processing devices connected via a network based on the data amount of the processing data included in the request received by the request reception unit. When,
When it is determined by the distribution determination unit that the request is distributed to the other information processing apparatus, the request is identified based on the distribution destination identification information acquired by the distributed processing information acquisition unit. Transfer means for transferring to the second information processing apparatus,
The control means includes
When it is determined by the distribution determination unit that the request is distributed to the other information processing apparatus, control is performed so as not to execute the processing for the processing data included in the request requested by the request, When it is determined by the distribution determination unit that the request is not distributed to the other information processing apparatus, control is performed so as to execute processing for the processing data included in the request requested by the request. A first information processing apparatus. The dispersion determining means includes
2. The first information processing apparatus according to claim 1, wherein when the data amount is equal to or greater than a preset data amount threshold, it is determined that the request is distributed to the other information processing apparatus. The distribution determination unit acquires the data amount based on attribute information included in the request,
Based on the acquired amount of data, determine whether to distribute the request to the other information processing apparatus,
The transfer means includes
When it is determined by the distribution determination unit that the request is distributed to the other information processing device, the processing data included in the request is transmitted to the first information processing device and the distributed processing information acquisition unit. Transmitted to the storage means accessible by the second information processing apparatus identified based on the distribution destination identification information acquired by
The first information processing apparatus according to claim 1, wherein access information used to acquire the transmitted processing data is transmitted to the second information processing apparatus. Conversion means for converting resource usage into the data amount;
The dispersion determining means includes
The other information processing apparatus based on the data amount of the processing data included in the request received by the request receiving unit and the converted data amount that is the data amount converted from the resource usage state by the converting unit 4. The first information processing apparatus according to claim 1, wherein the first information processing apparatus determines whether to distribute the request to the first information processing apparatus. The distributed processing information acquisition means includes
Obtaining from the distributed processing information holding means a distribution upper limit value that is an upper limit of the number of times the request is distributed to the other information processing apparatus held by the distributed processing information holding means;
The dispersion determining means includes
When determining whether to distribute the request to the other information processing apparatus, the distribution count information indicating the number of times the request included in the request is distributed, and the distributed processing information acquisition unit 5. The first information processing apparatus according to claim 1, wherein the determination is made based on the dispersion upper limit value. The dispersion determining means includes
6. The first information processing apparatus according to claim 5, wherein when the distribution count information reaches the distribution upper limit value, it is determined that the request is not distributed to the other information processing apparatus. Response accepting means for accepting a response returned from the second information processing apparatus in response to the request transferred to the second information processing apparatus by the transfer means;
Based on the response received by the response receiving means, response means for responding the request to the client that sent the request;
Timeout check means for determining whether or not the measurement time calculated after the request reception means has received the request exceeds a preset time,
The control means includes
When the distribution determination unit determines that the measurement time has exceeded a preset time by the timeout check unit, the distribution request is controlled to stop the distribution of the request, and the request requested by the request Control to execute processing on the processing data included in
The response means includes
7. The first information according to claim 1, wherein the request is returned to the client that has transmitted the request based on an execution result of the process controlled to be executed. Processing equipment. The request receiving means
8. The first information processing apparatus according to claim 7, wherein an error is notified to the client when the time-out check means determines that the measurement time has exceeded a preset time. The process requested by the request received by the request receiving unit is a process for print job data included in the request,
The dispersion determining means includes
9. The first information according to claim 1, wherein when the process is not a process for a page description language of the print job data, it is determined that the request is not distributed to the other information processing apparatus. Processing equipment. The dispersion determining means includes
The first information processing apparatus according to claim 9, wherein when the process is a process for a printer job language of the print job data, it is determined that the request is not distributed to the other information processing apparatus. A load balancing system in which two or more information processing devices are connected by a network,
The first information processing apparatus
First request accepting means for accepting a request transmitted from a client;
First control means for controlling execution of processing requested by the request received by the first request receiving means;
Distributing destination identification information for identifying a second information processing apparatus that is a distribution destination for distributing the request received by the first request receiving means, and held by the first distributed processing information holding means First distributed processing information acquisition means for acquiring distribution destination identification information;
Based on the amount of processing data included in the request received by the first request receiving means, it is determined whether to distribute the request to other information processing apparatuses connected via a network. First dispersion determination means;
The distribution destination identification information acquired by the first distributed processing information acquisition unit when the first distribution determination unit determines to distribute the request to the other information processing apparatus. And a first transfer means for transferring to the second information processing apparatus identified based on
The first control means includes
When the first distribution determination unit determines that the request is distributed to the other information processing apparatus, the processing for the processing data included in the request requested by the request is not performed. And when the first distribution determination unit determines that the request is not distributed to the other information processing apparatus, the request for the processing data included in the request requested by the request is performed. Control to run,
The second information processing apparatus
Second request receiving means for receiving a request transferred from the first information processing apparatus;
Second control means for controlling execution of processing requested by the request accepted by the second request acceptance means;
Distributing destination identification information for identifying a third information processing apparatus that is a distribution destination for distributing the request received by the second request receiving means, and held by the second distributed processing information holding means Second distributed processing information acquisition means for acquiring distribution destination identification information;
Based on the amount of processing data included in the request received by the second request receiving means, it is determined whether to distribute the request to other information processing apparatuses connected via the network. A second dispersion determination unit;
The distribution destination identification information acquired by the second distributed processing information acquisition unit when the second distribution determination unit determines to distribute the request to the other information processing apparatus. And a second transfer means for transferring to the third information processing apparatus identified based on
The second control means includes
When the second distribution determining unit determines that the request is distributed to the other information processing apparatus, the processing for the processing data included in the request requested by the request is not performed. And when the second distribution determination unit determines that the request is not distributed to the other information processing apparatus, the processing for the processing data included in the request requested by the request is performed. A load balancing system that is controlled to execute. A load balancing method for a system in which two or more information processing devices are connected by a network,
The first information processing apparatus
Receiving a request sent from the client and storing it in the first storage unit;
Controlling execution of processing requested by the request stored in the first storage unit;
The distribution destination identification information for identifying a second information processing apparatus that is a distribution destination for distributing the request stored in the first storage unit, the distribution being held in the first distributed processing information holding unit Obtaining pre-identification information and storing it in the first storage unit;
Determining whether to distribute the request to other information processing devices connected via a network based on the amount of processing data included in the request stored in the first storage unit When,
When it is determined that the request is distributed to the other information processing apparatus, the request is identified based on the distribution destination identification information stored in the first storage unit. Transferring to the processing device;
When it is determined that the request is distributed to the other information processing apparatus, control is performed so as not to execute processing for the processing data included in the request requested by the request;
When it is determined that the request is not distributed to the other information processing apparatus, the request is made to control the processing data included in the request to be executed.
The second information processing apparatus
Receiving a request transferred from the first information processing apparatus and storing it in a second storage unit;
Controlling execution of processing requested by the request stored in the second storage unit;
The distribution destination identification information for identifying a third information processing apparatus that is a distribution destination for distributing the request stored in the second storage unit, the distribution being held in the second distributed processing information holding unit Obtaining pre-identification information and storing it in the second storage unit;
Determining whether to distribute the request to other information processing apparatuses connected via a network based on the amount of processing data included in the request stored in the second storage unit When,
When it is determined that the request is distributed to the other information processing apparatus, the request is identified based on the distribution destination identification information stored in the second storage unit. Transferring to the processing device;
When it is determined that the request is distributed to the other information processing apparatus, control is performed so as not to execute processing for the processing data included in the request requested by the request;
When it is determined that the request is not distributed to the other information processing apparatus, the control requesting the request to perform processing on the processing data included in the request;
A load balancing method comprising: A program for a first information processing apparatus,
A process of receiving a request transmitted from a client and storing it in a storage unit;
A process for controlling execution of the process requested by the request stored in the storage unit;
The distribution destination identification information for identifying a second information processing apparatus that is a distribution destination that distributes the request stored in the storage unit, and acquires the distribution destination identification information held by the distributed processing information holding unit Processing to be stored in the storage unit,
A process of determining whether to distribute the request to other information processing apparatuses connected via a network based on the amount of processing data included in the request stored in the storage unit;
When it is determined that the request is distributed to the other information processing apparatus, the request is transferred to the second information processing apparatus identified based on the distribution destination identification information stored in the storage unit Processing to
When it is determined that the request is distributed to the other information processing apparatus, a process for controlling the processing data included in the request that is requested by the request not to be executed;
When it is determined that the request is not distributed to the other information processing apparatus, a process for controlling the process data included in the request that is requested by the request to be executed;
A program that causes a computer to execute.

Images