JP5928976B2 - Survival monitoring system - Google Patents

Description

  The present invention monitors the survival status of a server that transmits a request signal for requesting a predetermined process to an external system, such as an ordering server that transmits ordering information to a trading system for a financial product exchange. It relates to a survival monitoring system.

  In mission critical systems, servers are made redundant in case of unforeseen circumstances. As a result, when a failure occurs in the main server (operating server), it is possible to immediately shift the function to the secondary server (substitution server) so that the service for the user is not stopped. In order to realize such a mechanism, it is necessary to monitor the survival status of each server (see, for example, Patent Document 1).

  FIG. 9 is a schematic block diagram for explaining a conventional survival monitoring mechanism. FIG. 9 shows an example in which a survival monitoring mechanism is applied to an ordering system on the financial institution side that places an order with respect to a trading system for a financial product exchange. The ordering system on the financial institution side includes an ordering server A, an ordering server B, and an ordering server C, which are connected to a trading system at an exchange via a network. The ordering server A and the ordering server B are active main servers. Each of the ordering servers A and B performs ordering processing for transmitting ordering information to the trading system of the exchange via a switch. Upon receiving the ordering information, the exchange trading system transmits response information indicating that the ordering information has been received to the ordering server via the switch. The ordering server C is a standby secondary server (substitution server). When a failure occurs in one or both of the ordering server A and the ordering server B, the ordering server C is used as the ordering server instead. Conventionally, in order to monitor the survival status of each ordering server, each of the three ordering servers A, B, and C uses a signal for confirming the existence of other ordering servers as a survival monitoring process ( (Survival monitoring signal) is periodically transmitted. Then, when each of the ordering servers A, B, and C receives a response signal to the survival monitoring signal, it determines that the other ordering server is alive, while the response signal is transmitted. If not, it is determined that the other ordering server is not alive.

  Each ordering server A, B, and C not only performs survival monitoring processing for other ordering servers, but also performs processing for monitoring the survival status of the network used for ordering processing (network monitoring processing). ing. This network monitoring process is performed in the same manner as the server survival monitoring process using a network different from the network used for the ordering process.

JP 2008-299369 A

  By the way, if the server life monitoring process is not always performed, the effect cannot be obtained. For this reason, when each server performs this survival monitoring process, a load is generated on each server. In particular, in ordering systems that require high-speed transmission of ordering information from a financial institution to an exchange trading system, the processing load on the main server is reduced to a small extent, and the order processing that is the original business process is reduced. There is a desire to avoid delays.

  The present invention has been made based on the above circumstances, and an object thereof is to provide a survival monitoring system capable of monitoring the survival of the main server without causing a delay in processing performed by the active main server. Is.

  In order to achieve the above object, the present invention is a survival monitoring system that monitors the survival status of a main server connected to an external system via a network, and is transmitted between the main server and the external system. A relay switching means having a port mirroring function for copying a signal sent from the main server and sending it from a predetermined port when a signal is sent from the main server, and a relay switching means When it is determined that the signal from the main server sent from the predetermined port has not been sent for a predetermined time since the last reception, it is determined that the main server is not alive and And a management server that sends a signal indicating that the server is operating in place of the main server to a secondary server that is a server.

  Here, it is desirable that the management server performs the survival monitoring of the main server as follows, for example. That is, when the management server receives a signal from the main server sent from a predetermined port of the relay switching means, the management server determines that the main server is alive, and the determination result and the date and time when the signal is received Information indicating the survival status of the main server created based on the date / time information indicating the status is stored and managed in the storage means, and the signal from the main server is based on the information indicating the survival status of the managed main server. Is determined not to have been sent for a predetermined time since the last reception, it is determined that the main server is not alive.

  In the survival monitoring system according to the present invention, with the above configuration, the main server does not have a function of performing the survival monitoring process, and the management server performs the survival monitoring process on the main server. Thereby, since the main server does not need to perform the survival monitoring process, the processing load on the main server can be reduced. In addition, when the management server receives a signal from the main server sent from a predetermined port of the relay switching means and determines that the signal from the main server has not been sent for a predetermined time since the previous reception It is determined that the main server is not alive. In this way, the management server needs to send a signal for confirming the existence to the main server by determining the existence status of the main server using the signal sent from the main server to the external system. Therefore, the survival monitoring process by the management server does not affect the communication performed between the main server and the external system. Therefore, according to the survival monitoring system of the present invention, the processing load on the main server can be reduced, and the survival monitoring process by the management server does not affect the communication performed between the main server and the external system. Survival monitoring for the main server can be performed without causing a delay in processing performed by the main server.

  In the survival monitoring system of the present invention, the signal transmitted by the main server to the external system is a request signal for requesting a predetermined process and a communication path established between the external system on the network during operation. It is desirable to include a heartbeat signal for notifying that the self is alive so as not to be closed. Since the heartbeat signal is periodically transmitted from the main server to an external system, the survival status of the main server can be reliably monitored by using this heartbeat signal.

  In the survival monitoring system of the present invention, it is desirable to provide a network monitoring server that performs processing for monitoring whether the network in which the main server and the secondary server are connected to an external system is alive. This eliminates the need for the main server and the secondary server to perform network monitoring processing, thereby further reducing the processing load on these servers.

  In the survival monitoring system of the present invention, the management server periodically transmits a signal for confirming the existence to the secondary server, and a response signal to the transmitted signal for confirming the existence is the relevant secondary server. It is determined that the secondary server is alive, and a response signal to the transmitted confirmation signal is sent from the secondary server even if a predetermined time has passed since the transmission. It is desirable to determine that the secondary server is not alive when it does not. Thereby, the management server can monitor the survival status not only for the main server but also for the secondary server.

  Furthermore, in the survival monitoring system of the present invention, a log storage server that receives a signal from the main server sent from a predetermined port of the relay switching means and stores and manages a log of the received signal in the log storage means It is desirable to provide. As a result, since all logs of signals transmitted from the main server to the external system can be stored, it is possible to cope with an unexpected situation in the main server.

  According to the survival monitoring system according to the present invention, the processing load on the main server can be reduced and the survival monitoring process by the management server does not affect the communication performed between the main server and an external system. Survival monitoring of the main server can be performed without causing a delay in processing performed by the main server.

FIG. 1 is a schematic block diagram of an ordering system to which a survival monitoring system according to a first embodiment of the present invention is applied. FIG. 2 is a schematic block diagram of each ordering server of the ordering system. FIG. 3 is a schematic block diagram of the first management server in the survival monitoring system of the first embodiment. FIG. 4 is a flowchart for illustrating a procedure of processing performed by the first management server in the survival monitoring system. FIG. 5 is a schematic block diagram of an ordering system to which the survival monitoring system according to the second embodiment of the present invention is applied. FIG. 6 is a schematic block diagram of the third management server of the survival monitoring system. FIG. 7 is a flowchart showing a log collection processing procedure for ordering information in an ordering system to which the survival monitoring system is applied. FIG. 8 is a flowchart showing a log collection processing procedure for response information in the ordering system to which the survival monitoring system is applied. FIG. 9 is a schematic block diagram for explaining a conventional survival monitoring mechanism.

[First embodiment]
The best mode for carrying out the present invention will be described below with reference to the drawings.

  In the first embodiment, the survival monitoring system of the present invention is applied to an ordering system on the financial institution side that performs an ordering process on a trading system for a financial product exchange, which is one of high-speed information processing systems. explain. FIG. 1 is a schematic block diagram of an ordering system to which a survival monitoring system according to a first embodiment of the present invention is applied.

  The ordering system shown in FIG. 1 includes two main servers 100a and 100b and two secondary servers 100c and 100d. These servers 100a, 100b, 100c, and 100d are connected to a trading system 200 at an exchange via a network. The primary servers 100a and 100b are currently operating servers, and the secondary servers 100c and 100d have functions equivalent to those of the primary servers 100a and 100b. When a failure occurs in the primary servers 100a and 100b, the primary server It is an alternative server used in place of 100a and 100b. If a failure occurs in the primary server 100a, 100b, either or both of the secondary servers 100c, 100d will operate instead of the primary server. In particular, in the first embodiment, the secondary server 100c is used as an alternative server for the primary server 100a, and the secondary server 100d is used as an alternative server for the primary server 100b. In the following, the main server and the secondary server are collectively referred to as “ordering server”.

  Each of the ordering servers 100a, 100b, 100c, and 100d is a server that determines what algorithm is used for ordering in order to purchase a predetermined number of stocks of a certain company. The ordering servers 100a, 100b, 100c, and 100d are owned by financial institutions, but are installed in a data center of an exchange in order to perform ordering processing at high speed.

  An exchange trading system (external system) 200 is provided in an exchange of financial products (for example, Tokyo Stock Exchange). This exchange trading system 200 is a system that controls transactions relating to financial products handled by the exchange. In the following, a stock securities transaction will be described as an example, but the scope of the present invention is not limited to a stock securities transaction. Upon receiving the ordering information (request signal) from the ordering server, the trading system 200 at the exchange processes the order and sends response information (response signal) indicating that the ordering information has been received to the ordering server. . Since the trading system 200 of this exchange is publicly known, detailed description thereof is omitted here.

  Each main server 100a, 100b transmits various signals to the trading system 200 of the exchange. Specifically, each of the main servers 100a and 100b transmits order information (request signal) created by performing order processing, which is the original business process, to the trading system 200 of the exchange via the network. Here, each of the main servers 100a and 100b performs processing independently. For this reason, the two main servers 100a and 100b can simultaneously perform the ordering process. Each main server 100a, 100b is a heartbeat for notifying that it is alive so that the communication path established with the exchange trading system 200 on the network is not closed during operation. The signal is periodically transmitted to the trading system 200 of the exchange. Signal transmission is performed by packet communication between the main servers 100a and 100b and the trading system 200 at the exchange.

  The survival monitoring system of the first embodiment is for monitoring the survival status of each ordering server 100a, 100b, 100c, 100d of the ordering system. As shown in FIG. A server 20 and a second management server 30 are provided. The switch 10 relays a signal transmitted between the ordering servers 100a, 100b, 100c, and 100d and the trading system 200 at the exchange, and is a communication circuit that connects the ordering system and the trading system 200 at the exchange. It is provided on the line. Specifically, the four ordering servers 100 a, 100 b, 100 c, and 100 d are each connected to the switch 10, and the exchange trading system 200 is also connected to the switch 10. A first management server 20 is also connected to the switch 10. The second management server 30 is connected to each ordering server 100a, 100b, 100c, 100d.

  Next, the configuration of each ordering server 100a, 100b, 100c, 100d of the ordering system will be described. FIG. 2 is a schematic block diagram of each ordering server 100a, 100b, 100c, 100d of the ordering system. Each ordering server 100a, 100b, 100c, and 100d includes an order processing unit 111, a log control unit 112, and a storage unit (first log storage unit) 113, as shown in FIG.

  When the order processing unit 111 obtains market price information distributed from an exchange or information vendor, the order processing unit 111 analyzes and judges the market price information, and constructs a predetermined algorithm to purchase stocks at the exchange. Order information is generated and transmitted to the switch 10. Here, a serial number for identifying each order information is given to the order information by the order processing unit 111. Further, the order processing unit 111 has a function of transferring order information to the log control unit 112. When receiving the ordering information transferred from the order processing unit 111, the log control unit 112 stores the ordering information in the storage unit 113 as a log (history information) of the ordering information. The ordering information log stored in the storage unit 113 may store all of the ordering information including date information and serial number, or a part of the ordering information, for example, the serial number assigned to the ordering information You may make it memorize only. When a failure occurs in the ordering server, if this serial number is known, the corresponding ordering information can be specified, and whether or not the ordering information of the serial number has been accepted by the trading system 200 of the exchange will be described later. This is because it can be verified.

  Further, after the ordering server transmits the ordering information to the trading system 200 of the exchange, when response information for the ordering information is sent from the trading system 200 of the exchange to the ordering server, the order processing unit 111. Receives the response information and transmits it to the log control unit 112. Upon receiving the response information sent from the order processing unit 111, the log control unit 112 stores the response information in the storage unit 113 as a response information log (history information). A serial number is given to the response information by the trading system 200 of the exchange. Therefore, for this response information, for the same reason as the ordering information, all of the response information including date information and serial number may be stored as a response information log stored in the storage unit 113. For example, only the serial number given to the response information may be stored.

  Next, each component (switch 10, the 1st management server 20, the 2nd management server 30) of the survival monitoring system of 1st embodiment is demonstrated.

  The switch (relay switching means) 10 relays order information, response information, heartbeat signals and the like transmitted between the main servers 100a and 100b and the trading system 200 at the exchange. In the first embodiment, the switch 10 has a port mirroring function for copying a signal sent from the ordering server and sending it from a predetermined port when a signal is sent from each ordering server. Something is used. That is, the port mirroring function is a function of copying a packet flowing through a certain port of the switch 10 to another port arbitrarily designated by the user. The user can arbitrarily specify the copy source port and the copy destination port. Such port designation is performed when the switch 10 is installed.

  Specifically, in the example of FIG. 1, the primary server 100a is connected to the port “1” of the switch 10, the primary server 100b is connected to the port “2” of the switch 10, and the secondary server 100c is the port “3” of the switch 10. And the secondary server 100d is connected to the port “4” of the switch 10. The exchange trading system 200 is connected to the port “5” of the switch 10, and the first management server 20 is connected to the port “6” of the switch 10. In this case, in the switch 10, the port is designated so that the packet addressed to the exchange trading system received at each port “1” to “4” is copied and sent to the port “6”. . As a result, the switch 10 receives the signals (ordering information, heartbeat signal, etc.) addressed to the trading system of the exchange sent from the ordering servers 100a, 100b, 100c, and 100d, respectively, on the ports “1” and “2”. , “3” and “4”, the signal is sent to the trading system 200 of the exchange through the port “5”, and a copy of the signal is created through the port “6”. It is sent to the first management server 20.

  In the switch 10, the port is designated so that a packet addressed to each ordering server received at the port “5” is copied and sent to the port “6”. Thus, when the switch 10 receives a signal (response information, etc.) addressed to each ordering server from the trading system 200 of the exchange at the port “5”, the switch 10 sends the signal to the ordering server via the predetermined port. At the same time, a copy of the signal is created and sent to the first management server 20 via the port “6”. In this way, all copies of signals communicated between each ordering server and the trading system 200 of the exchange are transmitted to the first management server 20.

  The first management server (management server) 20 performs a survival monitoring process for determining whether each of the four ordering servers 100a, 100b, 100c, and 100d is alive. For example, a Sniffer server can be used as the first management server 20. FIG. 3 is a schematic block diagram of the first management server 20. As shown in FIG. 3, the first management server 20 includes a control unit 21 and a storage unit (storage unit) 22. The control unit 21 of the first management server 20 performs the following survival monitoring process on the main servers 100a and 100b. That is, when the control unit 21 receives a signal transmitted from the port “6” of the switch 10, the control unit 21 analyzes the content of the signal and is specified as the transmission source of the signal or the transmission destination of the signal. The main server is recognized and it is determined that the recognized main server is alive. This is because when the main server is specified as the transmission source of the signal, the signal is transmitted from the main server, and therefore the main server is considered to be actually operating at present. . Further, when the main server is specified as the transmission destination of the signal, the signal is response information transmitted from the trading system 200 of the exchange to the main server, but this response information is transmitted. This is because the order information is transmitted from the main server to the trading system 200 of the exchange, so that the main server is considered to be actually operating at present. When the control unit 21 determines that the main server is alive, the control unit 21 creates information indicating the survival status of the main server based on the determination result and date / time information indicating the date / time when the signal is received. 22 is stored and managed.

  Further, the control unit 21 determines whether or not a signal from the main server has been sent for a predetermined time from the previous reception based on information indicating the survival status of the main server stored in the storage unit 22. . Here, as the “predetermined time” as the determination criterion, a time longer than the transmission interval of the heartbeat signal from the main server is set. For example, the “predetermined time” can be set to about 5 seconds. In particular, when high-speed processing by the ordering server is required, it is desirable to set the “predetermined time” to a time as close as possible to the heartbeat signal, although longer than the heartbeat signal transmission interval. . Therefore, when the control unit 21 determines that the signal from the main server has not been sent for a predetermined time since the last reception, the heartbeat signal that should have been sent from the main server has not been transmitted. It is determined that the main server is not alive. Thus, when the control unit 21 determines that the main server is not alive, the control unit 21 transmits a signal indicating that the main server is operating instead of the main server to a predetermined sub server. Specifically, if it is determined that the primary server 100a is not alive, a signal indicating that the primary server 100a is operating as the primary server instead of the primary server 100a is transmitted to the secondary server 100c, and the primary server 100b is not alive. If it is determined, a signal indicating that it operates as a main server instead of the main server 100b is transmitted to the secondary server 100d.

  On the other hand, since the secondary servers 100c and 100d do not transmit / receive signals to / from the trading system 200 of the exchange, the survival monitoring process for the primary servers 100a and 100b is performed as the survival monitoring process for the secondary servers 100c and 100d. The same method cannot be applied. For this reason, the control part 21 of the 1st management server 20 transmits the signal (survival monitoring signal) for confirming survival with respect to each of the secondary servers 100c and 100d regularly, and with respect to the transmitted survival monitoring signal. When a response signal is sent from the secondary server, it is determined that the secondary server is alive. On the other hand, the control unit 21 determines that the secondary server is not alive when a response signal to the transmitted survival monitoring signal is not transmitted from the secondary server even after a predetermined time has elapsed since the transmission. This determination result is stored in the storage unit 22 as information indicating the survival status of the secondary server.

  The second management server (network monitoring server) 30 performs processing for monitoring whether or not the network in which the main servers 100a and 100b and the sub servers 100c and 100d are connected to the trading system 200 of the exchange is alive. is there. Conventionally, such network monitoring processing has been performed by each ordering server. However, in the survival monitoring system of the first embodiment, a second management server 30 is provided separately from the ordering server, and the second management server 30 has a network. The monitoring process is performed. Since the specific contents of the network monitoring process are the same as in the prior art, a detailed description thereof is omitted here.

  Next, an ordering process procedure in the ordering system of the first embodiment will be described. Here, consider a case where the main server 100a performs ordering processing.

  For example, a financial institution receives a request from a customer to purchase a predetermined number of shares when the stock price of company A reaches a specific value. When the order processing unit 111 of the main server 100a determines that the stock price of the company A has become a quote based on the market price information provided from an exchange or the like, it purchases a predetermined number of shares of the company. Determine and generate ordering information to purchase the stock. Then, the order processing unit 111 transmits the order information to the switch 10 and transfers the order information to the log control unit 112 in order to send the generated order information to the trading system 200 of the exchange.

  Upon receiving the order information from the order processing unit 111, the log control unit 112 stores a log of the order information in the storage unit 113. On the other hand, when the order information is sent from the order processing unit 111 to the switch 10 via the port “1”, the switch 10 sends the sent order information to the port “5” to the trading system 200 of the exchange. At the same time, a copy of the order information is generated, and the generated order information is sent to the port “6”. Here, the order information sent from the port “6” is sent to the first management server 20, and the first management server 20 performs the survival monitoring process for the main server based on the order information. As described above, the process of storing the order information log by the log control unit 112 and the process of transmitting the order information to the trading system 200 of the exchange via the switch 10 are performed asynchronously.

  Upon receiving the order information from the main server 100a, the exchange trading system 200 processes the order based on the order information, and transmits response information indicating that the order information has been received to the main server 100a. . Response information from the trading system 200 of this exchange is sent to the switch 10 via the port “5”. Then, the switch 10 sends the response information sent through the port “5” from the port “1” toward the main server 100a, generates a copy of the response information, and sends the generated response information to the main server 100a. , To port “6”. Here, the response information sent from the port “6” is sent to the first management server 20, and the first management server 20 performs the survival monitoring process for the main server based on this response information.

  When receiving the response information from the switch 10, the order processing unit 111 of the main server 100 a sends the received response information to the log control unit 112, and the log control unit 112 stores a log of the response information in the storage unit 113. To do.

  Next, a process procedure performed by the first management server 20 in the survival monitoring system according to the first embodiment will be described. FIG. 4 is a flowchart for illustrating a procedure of processing performed by the first management server.

  As shown in FIG. 4, the control unit 21 of the first management server 20 repeatedly executes the survival monitoring process for the primary server and the survival monitoring process for the secondary server. The survival monitoring process for the primary server includes a signal reception process (S11), a survival status information creation process (S12), a survival determination process (S13), and a switching process (S14). , Signal transmission processing (S15) and survival determination processing (S16). Here, in the processing flow of FIG. 4, in order to make the explanation easy to understand, it is described that each processing is executed in order, but in reality, a multi-core CPU is adopted as the control unit 21. The above processes are performed in parallel by multithreads.

  First, the control unit 21 performs signal reception processing (S11). In this signal reception process, when receiving a signal transmitted from the port “6” of the switch 10, the control unit 21 stores the content of the received signal in the storage unit 22 together with date / time information indicating the date / time of reception. Next, the control part 21 performs survival status information creation processing (S12). In this survival status information creation processing, the control unit 21 recognizes the main server specified as the transmission source of the signal or the transmission destination of the signal by analyzing the content of the signal stored in the storage unit 22. . Then, it is determined that the recognized main server is alive, and information indicating the survival status of the main server is created based on the determination result and the date and time information of the signal, and stored in the storage unit 22.

  Next, the control unit 21 performs a survival determination process (S13). In this survival determination process, the control unit 21 determines, for each of the main servers, a predetermined time from when the signal from the main server was previously received based on the information indicating the survival status of the main server stored in the storage unit 22. Determine if it has been sent. If the control unit 21 determines that the signal from the main server has not been sent for a predetermined time since the last reception, the control unit 21 determines that the main server is not alive and stores the determination result in the storage unit 22. .

  Next, the control unit 21 performs a switching process (S14). In this switching process, when the determination result that the main server is not alive is stored in the storage unit 22, the control unit 21 sends a signal indicating that the main server is operating instead of the main server to a predetermined secondary server. Send. If the main server 100a is not alive, a signal indicating that the main server 100a is operating instead of the main server 100a is transmitted to the secondary server 100c. A signal indicating that the server 100d is operating instead of the main server 100b is transmitted. When the secondary server receives from the first management server 20 a signal indicating that it operates in place of the primary server, the secondary server takes over the processing of the primary server and executes it. In this way, the secondary server becomes operational and operates as a primary server. The processes from step S11 to step S14 described above are the survival monitoring process for the main server.

  Next, the control unit 21 performs survival monitoring processing for the secondary server. First, the control unit 21 performs signal transmission processing for transmitting a survival monitoring signal for confirming survival to each secondary server (S15). Then, the control unit 21 determines that the secondary server is alive when a response signal to the transmitted survival monitoring signal is transmitted from the secondary server, while responding to the transmitted survival monitoring signal. If the signal is not sent from the secondary server even after a predetermined time has elapsed since the transmission, it is determined that the secondary server is not alive (S16). This determination result is stored in the storage unit 22 as information indicating the survival status of the secondary server. Here, when it is determined that the secondary server is not alive, the control unit 21 may send a signal to that effect to an external server or issue an alarm.

  In the survival monitoring system of the first embodiment, the primary server does not have a function of performing the survival monitoring process, and the first management server performs the survival monitoring process on the main server. Thereby, since the main server does not need to perform the survival monitoring process, the processing load on the main server can be reduced. When the first management server receives a signal from the main server sent from a predetermined port of the switch and determines that the signal from the main server has not been sent for a predetermined time since the last reception. It is determined that the main server is not alive. Thus, the first management server uses the signal transmitted from the main server to the trading system of the exchange to determine the survival status of the main server, thereby confirming the survival of the main server. Therefore, the survival monitoring process by the first management server does not affect the communication performed between the main server and the trading system of the exchange. Therefore, according to the survival monitoring system of the first embodiment, the processing load of the main server can be reduced, and the survival monitoring process by the first management server is performed between the main server and the trading system of the exchange. Therefore, it is possible to monitor the survival of the main server without causing a delay in processing performed by the main server.

  In addition, the heartbeat signal is included in the signal transmitted from the main server to the exchange trading system, and this heartbeat signal is periodically transmitted from the main server to the exchange trading system. By using, it is possible to reliably monitor the survival status of the main server.

  In the survival monitoring system of the first embodiment, the first management server periodically transmits a survival monitoring signal for confirming survival to the secondary server, and a response signal to the transmitted survival monitoring signal is When it is sent from the secondary server, it is determined that the secondary server is alive, and a response signal for the transmitted survival monitoring signal is not sent from the secondary server even if a predetermined time has elapsed since the transmission. Sometimes, it is determined that the secondary server is not alive. As a result, the first management server can monitor not only the main server but also the existence status of the secondary server, and the secondary server does not need to perform the survival monitoring process, reducing the processing load on the secondary server.

  Furthermore, in the survival monitoring system of the first embodiment, by including a second management server that performs processing to monitor whether the network in which the main server and the secondary server are connected to the trading system of the exchange is alive, Since the main server and the secondary server do not need to perform network monitoring processing, the processing load on these servers can be further reduced.

[Second Embodiment]
Next, a second embodiment of the present invention will be described with reference to the drawings. In the second embodiment as well, as in the first embodiment, the survival monitoring system of the present invention is one of the high-speed information processing systems. A case where the present invention is applied to the ordering system will be described. In the second embodiment, components having the same functions as those in the first embodiment are denoted by the same reference numerals, and detailed description thereof is omitted.

  FIG. 5 is a schematic block diagram of an ordering system to which the survival monitoring system according to the second embodiment of the present invention is applied. The ordering system shown in FIG. 5 includes two main servers 100a and 100b and two secondary servers 100c and 100d as in the first embodiment. The survival monitoring system of the second embodiment is for monitoring the survival status of each ordering server 100a, 100b, 100c, 100d of the ordering system. The switch 10, the first management server 20, Two management servers 30 and two third management servers (log storage servers) 40a and 40b are provided. That is, the survival monitoring system of the second embodiment is different from that of the first embodiment in that two third management servers 40a and 40b are provided. About another point, it is the same as that of the said 1st embodiment.

  The two third management servers 40a and 40b are connected to the switch 10, respectively. Specifically, in the example of FIG. 5, the third management server 40 a is connected to the port “7” of the switch 10, and the third management server 40 b is connected to the port “8” of the switch 10. In the second embodiment, the switch 10 copies packets addressed to the trading system of the exchange received at the ports “1”, “2”, “3”, and “4” to the ports “6”, “7”. ”And“ 8 ”, the port is designated. As a result, the switch 10 sends signals (ordering information, heartbeat signals, etc.) addressed to the trading system of the exchange from the ordering servers 100a, 100b, 100c, and 100d to the ports “1”, “2”, and “3”, respectively. , “4”, the signal is sent to the trading system 200 of the exchange via the port “5”, and a copy of the signal is created, and the first management server is sent via the port “6”. 20 to the third management server 40a via the port “7” and to the third management server 40b via the port “8”.

  Further, the switch 10 designates the port so that the packet addressed to each ordering server received at the port “5” is copied and sent to the ports “6”, “7”, “8”. Yes. Thus, when the switch 10 receives a signal (response information, etc.) addressed to each ordering server from the trading system 200 of the exchange at the port “5”, the switch 10 sends the signal to the ordering server via the predetermined port. At the same time, a copy of the signal is created and sent to the first management server 20 via the port “6”, to the third management server 40a via the port “7”, and via the port “8”. Each of them is sent to the three management servers 40b. In this way, all copies of signals communicated between each ordering server and the trading system 200 of the exchange are transmitted to the third management servers 40a and 40b, similarly to the first management server 20. become.

  Next, the third management servers 40a and 40b will be described. Each of the third management servers 40a and 40b receives a signal transmitted from a predetermined port of the switch 10 and collects logs (history information) about ordering information and response information among the received signals. . That is, in the second embodiment, the order information log and the response information log are not only stored in the storage unit 113 of each main server, but also stored and managed in the third management servers 40a and 40b. I have to. Here, the reason why the two third management servers 40a and 40b are provided is to increase redundancy and to improve the reliability of storing history information. As each 3rd management server 40a, 40b, a Sniffer server can be used, for example. FIG. 6 is a schematic block diagram of the third management servers 40a and 40b. As shown in FIG. 6, each third management server 40a, 40b includes a control unit 41, a raid controller unit 42, and two storage units (second log storage means) 43a, 43b.

  The control unit 41 stores a copy of order information sent from the main servers 100a and 100b to the exchange trading system 200, and a copy of response information sent from the exchange trading system 200 to the main servers 100a and 100b. To control the memory. Specifically, when receiving a copied signal sent from a predetermined port (“7” or “8”) of the switch 10, the control unit 41 analyzes the content of the signal. If it is determined as a result of the analysis that the signal received this time is ordering information, an instruction to store the ordering information together with the ordering information is sent to the raid controller unit 42, while the signal received this time is response information. If it is determined that the response information is stored, an instruction to store the response information is sent to the RAID controller unit 42 together with the response information. The RAID controller unit 42 is a RAID card, and is used in RAID 1 to simultaneously write information having the same contents into the two storage units 43a and 43b. When the control unit 41 sends an instruction to store the ordering information together with the ordering information, the raid controller unit 42 uses the ordering information as a log of ordering information (history information) and stores the two storage units 43a and 43b. On the other hand, when an instruction for storing the response information is sent together with the response information from the control unit 41, the two storage units 43a are used as the response information log (history information). , 43b is stored.

  In the second embodiment, the entire order information and the entire response information are stored as history information in the two storage units 43a and 43b. However, a part of the order information and a part of the response information are stored as history information. You may do it. In this case, for example, when the control unit 41 analyzes the content of the signal sent from the switch 10 and determines that the signal is ordering information, the control unit 41 acquires the serial number given to the information, An instruction to store the acquired serial number as a log of order information may be sent to the RAID controller unit 42. If it is determined that the signal is response information, the RAID controller 42 issues an instruction to acquire the serial number assigned to the information and store the acquired serial number as a log of response information. To send to.

  In the second embodiment, the RAID controller unit 42 is used in RAID1, but the present invention is not limited to this. For example, the RAID controller unit 42 is configured by combining RAID0 and RAID1. For example, high reliability may be realized and high speed may be achieved.

  The ordering process procedure in the ordering system of the second embodiment and the survival monitoring process procedure for the ordering server performed by the first management server 20 in the survival monitoring system of the second embodiment are the same as those in the first embodiment. It is the same. Therefore, detailed description thereof is omitted here.

  Next, a description will be given of log collection processing for order information transmitted from the main server to the trading system 200 of the exchange in the ordering system to which the survival monitoring system of the second embodiment is applied. FIG. 7 is a flowchart showing a log collection processing procedure for ordering information in the ordering system to which the survival monitoring system of the second embodiment is applied. Here, consider a case where the main server 100a performs ordering processing.

  When the order processing unit 111 of the main server 100a obtains the market price information provided from the exchange or the like, the order processing unit 111 analyzes and judges the market price information and determines to place an order, and generates the order information (S21). ). Then, the order processing unit 111 transmits the order information to the switch 10 in order to send the generated order information to the trading system 200 of the exchange (S22) and transfers the order information to the log control unit 112. (S23). Upon receiving the ordering information, the log control unit 112 stores the ordering information in the storage unit 113 as an ordering information log (S24). Thus, the process of step S22 and the processes of steps S23 and S24 are performed in parallel. Accordingly, in the main server 100a, the order information transmission process to the switch 10 and the order information log storage process are performed asynchronously, so that the main server 100a performs the order process compared to the case where these processes are performed synchronously. Can be performed at high speed.

  When the ordering information is sent from the order processing unit 111 to the switch 10 via the port “1” by the process of step S22 (S25), the switch 10 sends the sent ordering information to the port “5” for the transaction. At the same time, the information is sent to the local trading system 200 (S26), and three copies of the order information are generated, and the generated order information is sent to ports “6”, “7”, and “8”, respectively. Among these, the order information sent from the ports “7” and “8” is sent to the third management servers 40a and 40b, respectively (S27). Thus, the process of step S26 and the process of step S27 are performed in parallel. Here, the order information sent from the port “6” of the switch 20 is sent to the first management server 20, and the first management server 20 is based on the order information as described in detail in the first embodiment. Thus, the survival monitoring process for the main server is performed.

  The signal (order information) sent from the port “7” of the switch 10 is sent to the third management server 40a. When the control unit 41 of the third management server 40a analyzes the content of the transmitted signal and determines that the signal is ordering information, an instruction to store the ordering information together with the ordering information is sent to the RAID controller unit. 42. Then, the raid controller unit 42 of the third management server 40a simultaneously stores the order information in the two storage units 43a and 43b as an order information log in accordance with an instruction from the control unit 41 (S28). Similarly, a signal (order information) sent from the port “8” of the switch 10 is sent to the third management server 40b. When the control unit 41 of the third management server 40b analyzes the contents of the transmitted signal and determines that the signal is ordering information, the RAID controller unit issues an instruction to store the ordering information together with the ordering information. 42. Then, the raid controller unit 42 of the third management server 40b simultaneously stores the ordering information in the two storage units 43a and 43b as an ordering information log in accordance with an instruction from the control unit 41 (S29).

  Through the above processing, the order information generated by the main server 100a and transmitted to the trading system 200 of the exchange is stored in the storage unit 113 of the main server 100a and two third management servers 40a and 40b. Are also stored in the two storage units 43a and 43b.

  Next, log collection processing for response information transmitted from the trading system 200 of the exchange to the main server in the ordering system to which the survival monitoring system of the second embodiment is applied will be described. FIG. 8 is a flowchart showing a log collection processing procedure for response information in the ordering system to which the survival monitoring system according to the second embodiment is applied. Here, consider a case where response information is transmitted to the main server 100a.

  Upon receiving the order information from the main server 100a, the trading system 200 at the exchange processes the order based on the order information, and transmits response information indicating that the order information has been received to the main server 100a. Then, the response information from the trading system 200 of the exchange is sent to the switch 10 (S41). Upon receiving the response information, the switch 10 sends the response information to the main server 100a from the port “1” (S42), and copies the response information to generate three. Are sent to the ports “6”, “7”, “8” (S43). Thus, the process of step S42 and the process of step S43 are performed in parallel. Here, the response information sent from the port “6” of the switch 20 is sent to the first management server 20, and the first management server 20 is based on this response information as described in detail in the first embodiment. Thus, the survival monitoring process for the main server is performed.

  When the response information sent from the port “1” of the switch 10 is received by the order processing unit 111 of the main server 100a by the processing of step S42 (S44), the order processing unit 111 sends the received response information to the log control unit. 112 (S45). Upon receiving the response information from the order processing unit 111, the log control unit 112 stores the response information in the storage unit 113 as a response information log (S46).

  On the other hand, the signal (response information) sent from the port “7” of the switch 10 by the process of step S43 is sent to the third management server 40a. When the control unit 41 of the third management server 40a analyzes the content of the transmitted signal and determines that the signal is response information, the RAID controller unit issues an instruction to store the response information together with the response information. 42. Then, the RAID controller 42 of the third management server 40a simultaneously stores the response information in the two storage units 43a and 43b as a response information log in accordance with an instruction from the control unit 41 (S47). Similarly, the signal (response information) sent from the port “8” of the switch 10 by the process of step S43 is sent to the third management server 40b. When the control unit 41 of the third management server 40b analyzes the content of the transmitted signal and determines that the signal is response information, the RAID controller unit issues an instruction to store the response information together with the response information. 42. The RAID controller 42 of the third management server 40b simultaneously stores the response information in the two storage units 43a and 43b as a response information log in accordance with an instruction from the control unit 41 (S48).

  With the above processing, the response information transmitted from the trading system 200 of the exchange to the main server 100a is stored in the storage unit 113 of the main server 100a and two of the two third management servers 40a and 40b. It is also stored in the storage units 43a and 43b.

  The survival monitoring system of the second embodiment has the same operations and effects as the survival monitoring system of the first embodiment. In the second embodiment, when the third management server receives a copied signal transmitted from a predetermined port of the switch, the log of the signal is stored in the storage unit and managed. For this reason, all the order information logs sent from the main server to the trading system of the exchange can be stored in the third management server, so even if an unexpected situation occurs in the main server can do.

  By the way, according to the second embodiment, the whole of the survival monitoring system of the present invention and the ordering system on the financial institution side is regarded as one information processing system that transmits a signal to an external system via a network. When this information processing system is used, it can be seen that a specific effect can be obtained. That is, in the ordering system to which the survival monitoring system according to the second embodiment is applied, the ordering information is transmitted from the main server to the switch in parallel with storing the ordering information log in the main server, and the transaction is performed from the switch. In parallel with sending the order information to the local trading system, the order information copied from the switch is sent to the two third management servers to store the order information log. As a result, the main server sends the ordering information to the trading system of the exchange asynchronously with the process of storing the log of the ordering information. Therefore, the transmission process of the ordering information can be performed at a high speed while minimizing the delay caused by the log storage process. Can do. Moreover, since the order information log is stored not only in the main server but also in the third management server, the order information log transmitted to the trading system of the exchange can be reliably collected. For this reason, for example, even if a failure occurs in the main server before the main server stores the order information log in the storage unit, the order information sent to the trading system of the exchange is sent via the switch. Since the order information logs are collected by the two third management servers, the order information is processed by analyzing the logs collected by the third management server later. It is possible to verify whether or not the product has been shipped to a certain trading system. Therefore, according to the ordering system to which the survival monitoring system of the second embodiment is applied, even in algorithmic transactions that require high-speed processing, high reliability is ensured and ordering information is exchanged at a high-speed trading system. Can be sent to.

  In the ordering system to which the survival monitoring system of the second embodiment is applied, the switch relays response information transmitted from the trading system of the exchange and sends it to the main server that is the transmission destination. Is copied and sent to the third management server. The main server stores a log of response information sent from the switch, while the third management server also receives the copied response information from the switch and stores the received log of the response information. To do. Thereby, the response information can be stored together with the order information, and the history regarding the order can be completely collected.

[Other Embodiments]
The present invention is not limited to the above-described embodiments, and various modifications are possible within the scope of the gist.

  For example, in each of the above embodiments, the case has been described where the switch copies the response information transmitted from the trading system of the exchange and sends the response information to the first management server. However, the switch sends the response information to the first management server. It is not necessary to send it out. If the first management server can receive a signal (ordering information, heartbeat signal) transmitted from the main server to the trading system of the exchange, it can monitor the survival of the main server based on the signal. is there.

  In the second embodiment, the case has been described where the switch copies the response information transmitted from the trading system of the exchange and sends the response information to the third management server. However, the switch transmits the response information to the third management server. It is not necessary to send it to. That is, the third management server may collect only the order information log transmitted by the main server. This is because if only the order information log is collected and stored, the failure of the main server can be dealt with.

  In the second embodiment, the third management server may be provided with an analyzer function that can analyze, for example, a stored log, extract necessary information, and display the information on a display unit.

  Further, in each of the above embodiments, the case where the second management server performs the network monitoring process has been described. However, the second management server may not be provided, and each ordering server may perform the network monitoring process.

  In each of the above embodiments, the case where two main servers are provided has been described, but one or three or more main servers may be provided. In a system that requires high-speed ordering processing, it is desirable to provide a plurality of main servers. Thereby, since each main server can be processed simultaneously and in parallel, the processing of the main server can be performed efficiently.

  Moreover, although each said embodiment demonstrated the case where two secondary servers were provided, you may make it provide one or three or more secondary servers. In particular, by providing a plurality of secondary servers, the secondary servers can be made redundant. Therefore, even if an unexpected situation occurs, the system can be prevented from being stopped and the reliability of the system can be improved. Can be planned.

  In the second embodiment, the case where the third management server is provided separately from the first management server has been described. However, the first management server has an original function of performing the survival monitoring process for each ordering server. In addition to the execution, the third management server function of performing log storage processing may also be executed. Thereby, the number of third management servers can be reduced.

  Moreover, although said 2nd embodiment demonstrated the case where two 3rd management servers were provided, you may make it provide one or 3 or more 3rd management servers. In particular, by providing a plurality of third management servers, it is possible to store order information logs in the plurality of third management servers, thereby improving the reliability when sending order information at high speed. Can be planned. In addition, when a plurality of ordering systems are provided on one financial institution side, the third management server is not provided individually for each of the plurality of ordering systems, but the third management server is provided with a plurality of ordering systems. You may make it share.

  In each of the above embodiments, the case where the first management unit monitors the survival status of the primary server and the secondary server has been described. However, the present invention is not limited to this, and the primary management unit is not limited to the primary server. The survival status of the secondary server may be monitored, and the survival status of the secondary server may not be monitored.

  Further, in each of the above embodiments, the case where the survival monitoring system of the present invention is applied to an ordering system on the side of a financial institution that performs ordering processing on a trading system of a financial product exchange has been described. The present invention is not limited to this, and the present invention can be applied to any system as long as it is an information processing system that needs to transmit a request signal for a predetermined process to an external system.

  As described above, in the survival monitoring system of the present invention, the main server does not have a function of performing the survival monitoring process, and the management server performs the survival monitoring process on the main server. Thereby, since the main server does not need to perform the survival monitoring process, the processing load on the main server can be reduced. In addition, when the management server receives a signal from the main server sent from a predetermined port of the relay switching means and determines that the signal from the main server has not been sent for a predetermined time since the previous reception It is determined that the main server is not alive. In this way, the management server needs to send a signal for confirming the existence to the main server by determining the existence status of the main server using the signal sent from the main server to the external system. Therefore, the survival monitoring process by the management server does not affect the communication performed between the main server and the external system. Therefore, according to the survival monitoring system of the present invention, the processing load on the main server can be reduced, and the survival monitoring process by the management server does not affect the communication performed between the main server and the external system. Survival monitoring for the main server can be performed without causing a delay in processing performed by the main server. Therefore, the present invention can be applied to, for example, an ordering system on the financial institution that performs an ordering process on a trading system for a financial product exchange.

10 switch (relay switching means)
20 First management server (management server)
21 Control Unit 22 Storage Unit (Storage Unit)
30 Second management server (network monitoring server)
40a, 40b Third management server (log storage server)
41 control unit 42 raid controller unit 43a, 43b storage unit 100a, 100b primary server 100c, 100d secondary server 111 order processing unit 112 log control unit 113 storage unit 200 trading system for exchange (external system)

Claims (6)

A survival monitoring system for monitoring the survival status of an ordering main server connected to a trading system of an exchange via a network,
Relay the signal transmitted between the ordering main server and the trading system of the exchange, and when the signal is sent from the ordering main server, the signal sent from the ordering main server Relay switching means having a port mirroring function for copying and sending out from a predetermined port;
When it is determined that the ordering main server is not alive when it is determined that the signal from the ordering main server sent from the predetermined port of the relay switching means has not been sent for a predetermined time since the previous reception. determined, the management server sends a signal to the effect that running on behalf of the server the main for ordering server secondary server for ordering a alternative for for the orderer server,
A survival monitoring system comprising: The management server determines that the ordering main server is alive when receiving a signal from the ordering main server sent from a predetermined port of the relay switching means, and the determination result and with stores and manages information indicating the survival status of primary server for the order that is generated based on the date and time information indicating the date and time of receiving the signal in the storage means, indicating the survival status of the main server for ordering to its management according to which a signal from the main server for that order based on the information when it is determined that no sent a predetermined time from the time of last received, the main server for that orders and judging that no viable Item 1. The survival monitoring system according to Item 1. A signal transmitted from the ordering main server to the trading system of the exchange is a communication path established between a request signal for requesting a predetermined process and the trading system of the exchange on the network during operation. 3. A survival monitoring system according to claim 1, further comprising a heartbeat signal for notifying that the self is alive so as not to be closed. 2. The network monitoring server for performing processing for monitoring whether or not the network connected to the trading system of the exchange is alive, wherein the ordering main server and the ordering secondary server are provided. The survival monitoring system according to 2 or 3. The management server sends said signal for confirming the survival against sub for ordering server periodically, the response signal to the signal to confirm the survival and the transmission is sent from the secondary server for the order The ordering secondary server determines that the ordering secondary server is alive, and a response signal to the transmitted confirmation signal is sent from the ordering secondary server even if a predetermined time elapses from the time of transmission. 5. The survival monitoring system according to claim 1, wherein the ordering secondary server determines that the ordering secondary server is not alive when not picked up. And a log storage server that receives a signal from the ordering main server sent from a predetermined port of the relay switching unit and stores and manages a log of the received signal in a log storage unit. The survival monitoring system according to claim 1, 2, 3, 4, or 5.

Images