WO2014013674A1

WO2014013674A1 - Communication keeping system, terminal device, communication keeping method, and connection keeping program

Info

Publication number: WO2014013674A1
Application number: PCT/JP2013/003763
Authority: WO
Inventors: 貴弘城島
Original assignee: 日本電気株式会社
Priority date: 2012-07-17
Filing date: 2013-06-17
Publication date: 2014-01-23
Also published as: JP6176247B2; JPWO2014013674A1

Abstract

A connection management unit (91) sets a sustained connection with a server device (80) by designating an extension interval. A packet transmission unit (81) transmits a keep-alive packet to a terminal device (90) upon the elapse of the extension interval designated by the setting of the sustained connection. The connection management unit (91), upon the elapse of a high-reliability interval since the setting of the sustained connection, sets a new sustained connection with the server device (80) designating an extension interval determined in accordance with the reception state of the keep-alive packet, and, upon reception of the keep-alive packet based on the new sustained connection from the server device (80), sets the extension interval designated by the setting of the sustained connection as the high-reliability interval.

Description

Communication maintenance system, terminal device, communication maintenance method, and connection maintenance program

The present invention relates to a communication maintenance system that maintains communication between devices, a terminal device, a communication maintenance method, a connection maintenance method with a server device, and a connection maintenance program.

With the development of mobile terminals, cooperative systems between server devices and mobile terminals such as smartphones have come to be used. For example, communication connection is performed between an arbitrary application that operates on a mobile terminal such as a smartphone and a server device arranged on the network. Using the connection, the server device causes the mobile terminal to execute arbitrary processing.

For such communication between the server device and the portable terminal, a connection type connection protocol called TCP / IP (Transmission Control Protocol / Internet Protocol) is generally used, and transmission / reception is performed using this connection protocol. There are many cases.

In this TCP / IP, one device that wishes to connect transmits a packet indicating a connection start request to the other device, and the device that receives the connection start request packet sends a connection acceptance response to the transmission source terminal. Return the meaning packet. By this procedure, the connection associated with the connection start request is set in each device. Thereafter, communication is executed by transmitting and receiving packets using this connection.

Also, mainly due to security problems, there are many cases where a device collectively called a middle box for relaying or blocking communication is installed on a network connecting a server and a mobile terminal. Examples of the middle box include a NAT (Network Address Translation), a firewall, and a proxy server.

The middle box is mainly used to send and receive packets between a terminal having a private IP address assigned by a network operator or user and a server device having a global IP address that can also be used in a global network such as the Internet. Convert IP addresses to each other. Such conversion enables a mobile terminal connected to a private network to communicate with a server device connected to a global network such as the Internet.

When a connection start request packet in TCP / IP is sent from a server on the global network to a terminal on the private network, a normal middle box discards the packet unless special conditions are set in advance. Do not allow connection between devices.

On the other hand, when the middle box receives a connection start request packet for a server device on the global network from a terminal on the private network, the middle box relays the connection start request packet to the server device. Further, when the middle box receives the connection acceptance reply packet from the server device, the middle box relays the reply packet to the terminal on the private network. In this way, the setting of the connection associated with the connection start request is permitted between the terminal on the private network and the server device on the global network.

Thereafter, the middle box continuously relays the packet transmitted from the server device or the terminal using the permitted connection.

Also, the middle box usually monitors the transmission and reception of packets on the connection. If transmission / reception of a packet is not executed between devices for a certain period of time (hereinafter also referred to as a connection invalidation deadline), the middle box invalidates the connection permission setting, and thereafter transmits packets transmitted / received using that connection. Discard.

Patent Document 1 describes a communication system in which such a middle box is interposed. In the communication system described in Patent Document 1, since a packet is transmitted from a server device on the global network to a terminal on the private network at an arbitrary timing, a connection is established from the terminal on the private network to the server on the global network. Is set. Thereafter, the packet is transmitted and received between the devices every predetermined period (hereinafter referred to as keep alive interval) so that the persistent connection is not invalidated by the middle box. Connection is maintained.

JP 2003-18235 A

In the communication system described in Patent Document 1, it is determined whether or not the persistent connection has been invalidated by transmitting and receiving packets after the keep alive interval has elapsed after the persistent connection is set. When the persistent connection is valid, the keep alive interval is increased. When the persistent connection is invalidated, the persistent connection is set again after decreasing the keep alive interval.

However, when the method described in Patent Document 1 is used, if the keep alive interval is longer than the connection invalidation deadline, a period calculated by “keep alive interval−connection invalidation deadline” (hereinafter referred to as a no-response period). Will be in a state where the persistent connection is disabled. Therefore, during this time, the server cannot transmit a packet to the terminal.

Also, when the n-th (n ≧ 1) keep-alive interval when the persistent connection is valid is Kn, the increment of the keep-alive interval to be increased next is d, and the connection invalidation deadline is T, The response period is calculated by (Kn + d) −T. In this case, the maximum no-response period is a period when Kn = T, that is, the period is d. Therefore, in order to shorten the non-response period, the increment d of the keep alive interval may be reduced.

However, when K1 <T is satisfied and K1 is a value much smaller than T, Kn does not reach T easily, and packets are continuously transmitted and received at a short interval, which places a wasteful load on the network.

Therefore, the present invention provides a communication maintenance system, a terminal device, and a communication that can suppress a communication load for maintaining a connection while maintaining the connection between devices even in a communication network in which the connection between the devices becomes invalid for a predetermined period. It is an object to provide a maintenance method, a connection maintenance method with a server device, and a connection maintenance program.

A communication maintenance system according to the present invention includes a server device and a terminal device that communicates with the server device using a persistent connection set between the server device, and the server device and the terminal device use the persistent connection. When communication is not performed during a predetermined period of keep-alive interval, the terminals are connected to each other via at least one middle box device that invalidates the persistent connection, and the persistent connection is valid. The high-reliability interval, which is the maximum period of the keep-alive interval when it can be confirmed, and the extended interval with a longer period than the high-reliability interval used to check whether the persistent connection is valid or not are extended. A connection management unit is provided for setting a persistent connection with the server device by specifying an interval. Including a packet transmission unit that transmits keep-alive packets to the terminal device when the extended interval specified in the setting of the persistent connection has elapsed since the session was set, and the connection management unit of the terminal device sets the persistent connection. When the high-reliability interval elapses, a new persistent connection that specifies an extended interval determined according to the keep-alive packet reception status is set with the server device, and the keep based on the new persistent connection is set. When the alive packet is received from the server device, the extended interval designated by the setting of the persistent connection is set as a high-reliability interval.

The terminal device according to the present invention is a terminal device that communicates with the server device using a persistent connection set up with the server device, and communication with the server device using the persistent connection is predetermined. If the keep alive interval is not performed, it is connected via at least one middlebox device that invalidates the persistent connection, and the keep alive interval when it is confirmed that the persistent connection is valid. The maximum reliable period and the extended interval longer than the reliable interval used to check whether the persistent connection is valid are retained, and the extended interval is specified to establish a persistent connection with the server device. A connection management unit is provided, and when the high reliability interval elapses after the connection management unit sets a persistent connection, A new extension interval that is determined according to the reception status of the keep-alive packet sent from the server device when the extension interval specified in the setting of the persistent connection elapses after a persistent connection is set between When a persistent connection is set with the server device and a keep-alive packet based on the new persistent connection is received from the server device, the extended interval specified by the setting of the persistent connection is set as a reliable interval. It is characterized by that.

The communication maintenance method according to the present invention communicates with a server device using a persistent connection established between the server device and the server device for a predetermined period of communication using the persistent connection. When the keep alive interval is not performed, the keep alive interval when the terminal device connected via at least one middle box device that invalidates the persistent connection can confirm that the persistent connection is valid. Among these, the high-reliability interval that is the maximum period, and the extended interval that is longer than the high-reliability interval that is used to check whether the persistent connection is valid or not are retained. The persistent connection is established after the server device establishes a persistent connection with the terminal device. When the extended interval specified in the setting elapses, a keep-alive packet is transmitted to the terminal device. When the high-reliability interval elapses after the terminal device establishes a persistent connection, it depends on the reception status of the keep-alive packet. When the terminal device receives a keep-alive packet based on the new persistent connection from the server device, a new persistent connection that specifies the extended interval determined in this way is set with the server device. The extended interval specified by the step is set to a highly reliable interval.

The method for maintaining a connection with a server device according to the present invention communicates with the server device using a persistent connection set up with the server device, and communication using the persistent connection is predetermined with the server device. If the terminal device connected via at least one middlebox device that invalidates the persistent connection can be confirmed that the persistent connection is valid when the keep alive interval is not performed. The high-reliability interval, which is the maximum period of the keep-alive interval, and the extended interval longer than the high-reliability interval used to check whether the persistent connection is valid are retained, and the terminal device specifies the extended interval. To establish a persistent connection with the server device, and when the high reliability interval has elapsed since the terminal device established the persistent connection. Specify the extension interval that is determined according to the reception status of the keep-alive packet sent from the server device when the extension interval specified in the setting of the persistent connection has elapsed since the persistent connection was set up with the server device. When the terminal device receives a keep-alive packet based on the new persistent connection from the server device, the extended interval specified by the setting of the persistent connection is increased. The confidence interval is set.

The connection maintenance program according to the present invention communicates with a server device using a persistent connection set up with the server device, and the server device has a predetermined period of communication using the persistent connection. A connection maintenance program applied to a computer connected via at least one middlebox device that invalidates the persistent connection when the keep alive interval is not performed, and the persistent connection is valid for the computer. The high-reliability interval, which is the maximum period of the keep-alive interval when it can be confirmed, and the extended interval with a longer period than the high-reliability interval used to check whether the persistent connection is valid or not are extended. Connection management is executed by specifying the interval and executing connection management processing to set up a persistent connection with the server device. Therefore, when a high-reliability interval elapses after setting a persistent connection, when the extended interval specified in the setting of the persistent connection elapses after the persistent connection is set with the server device, A new persistent connection specifying an extension interval determined according to the reception state of the keep-alive packet to be transmitted is set between the server apparatus and the keep-alive packet based on the new persistent connection is received from the server apparatus. In some cases, the extended interval designated by the setting of the persistent connection is set to a highly reliable interval.

According to the present invention, it is possible to suppress a communication load for maintaining a connection while maintaining a connection between devices even in a communication network in which a connection between devices is invalidated during a predetermined period.

It is a block diagram which shows one Embodiment of the communication maintenance system by this invention. It is a flowchart which shows the operation example in which a terminal device sets a persistent connection. It is a flowchart which shows the operation example in which a server apparatus sets a persistent connection. It is a flowchart which shows the example of the process which a server apparatus transmits a message. It is a flowchart which shows the example of the process which a terminal device receives a message. It is a flowchart which shows the operation example when the high reliability interval passage event generate | occur | produces in the terminal device. It is a flowchart which shows the example of a process when a keep alive event generate | occur | produces in the server apparatus. It is a flowchart which shows the example of a process when a terminal device receives a keep alive packet. It is a flowchart which shows the example of a process when an extended interval elapsed event generate | occur | produces in the terminal device. It is explanatory drawing which shows the operation example of the communication maintenance system of this embodiment. It is a block diagram which shows the outline | summary of the communication maintenance system by this invention. It is a block diagram which shows the outline | summary of the terminal device by this invention.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of a communication maintenance system according to the present invention. The communication maintenance system of this embodiment includes a terminal device 1 and a server device 2. The terminal device 1 is operated by a user, for example. The server device 2 is connected to the other server device 3 and mediates a message transmitted from the other server device 3 to the terminal device 1. The other server device 3 provides a service to a user who operates the terminal device 1 by transmitting a message to the terminal device 1.

The middle box device 4 is connected between the terminal device 1 and the server device 2 via a communication network. The middle box device 4 monitors and relays communication between the terminal device 1 and the server device 2 by a firewall function or a NAT function. 1 illustrates a case where one middle box is connected between the terminal device 1 and the server device 2. However, the number of middle box devices 4 is not limited to one. It may be more than one.

Further, the middle box device 4 monitors whether a packet is transmitted / received over a TCP / IP connection between the terminal device 1 and the server device 2. If no packet transmission / reception is executed during the connection invalidation period, the middle box device 4 invalidates the connection permission setting, and thereafter discards the packet transmitted / received using the connection. That is, the middle box device 4 invalidates the connection when communication using the connection set between the terminal device 1 and the server device 2 is not performed for a predetermined period.

The terminal device 1 includes a persistent connection management unit 11, a duration calculation unit 12, a timeout processing unit 13, a terminal identifier management unit 14, a connection management unit 15, a message reception unit 16, and a timer 17.

For example, the persistent connection management unit 11 is configured such that when the terminal device 1 is activated, when a network device (not shown) included in the terminal device 1 is activated, or when a user operates the terminal device 1, the server device 2. Setting up a persistent connection with 2. The persistent connection management unit 11 holds the maximum period among the periods in which it is confirmed that the connection (sustained connection) maintained with the server device 2 is valid as a highly reliable interval. Further, the persistent connection management unit 11 holds an interval for confirming whether or not the persistent connection set next time with the server device 2 is valid as an extended interval.

Here, since the period that is not invalidated by the middle box device 4 is the keep-alive interval, the high-reliability interval is the maximum period of the keep-alive interval when it can be confirmed that the persistent connection is valid. it can. The expansion interval can be said to be a period obtained by adding a difference in keep alive interval to be increased and a highly reliable interval. In other words, the expansion interval means a period increased by a predetermined period to the high reliability interval.

This extension interval is a value set to determine whether or not the determined high-reliability interval can be further extended. For example, when a high-reliability interval is determined for a certain persistent connection, an extended interval obtained by extending the high-reliability interval is determined, and a continuous connection with the specified extended interval is newly set between devices. A packet is transmitted as keep-alive from one device to the other device at the timing when this expansion interval elapses. When this packet arrives, the highly reliable keep-alive interval can be expanded.

Note that the high-reliability interval and the extension interval may be held by the persistent connection management unit 11 or may be held by the duration calculation unit 12 described later.

In addition, the persistent connection management unit 11 sets a new persistent connection with the server device 2 after the high reliability interval has elapsed when the maximum value of the high reliability interval has not been determined. On the other hand, when the maximum value of the high-reliability interval is determined, the persistent connection management unit 11 receives a packet from the server device 2 that confirms whether the persistent connection is valid (hereinafter referred to as a keep alive packet). A new persistent connection is set up with the device 2. Note that whether or not the maximum value of the high confidence interval has been determined is determined based on a change mode held by the duration calculation unit 12 described later. In this way, the persistent connection management unit 11 continues to maintain a persistent connection with the server device 2.

The duration calculation unit 12 determines the extension interval. The duration calculation unit 12 holds one of the three states (enlargement mode, reduction mode, and steady mode) as the change mode. The expansion mode is a mode that maintains the high reliability interval and the expansion interval and gradually increases the expansion interval of the newly set persistent connection. The reduction mode is a mode indicating that the extension interval of a newly established persistent connection is gradually reduced. The steady mode is a mode in which the expansion interval is not changed. The state in which the duration calculation unit 12 holds the steady mode as the change mode indicates a case where the maximum value of the high reliability interval is determined. The duration calculation unit 12 determines the expansion interval based on these modes and the high reliability interval. A method for determining the expansion interval will be described later.

The timeout processing unit 13 sets a setting for generating two types of events in the timer 17 attached to the terminal device 1. The first event is a high-reliability interval elapsed event that occurs after a high-reliability interval elapses from the current time when setting a persistent connection with the server device 2. The second event is an extended interval elapsed event that occurs after the extended interval has elapsed since the present time.

The terminal identifier management unit 14 acquires a terminal identifier for uniquely identifying the terminal device 1.

The connection management unit 15 sets a TCP / IP connection with the server device 2 as a persistent connection. Further, the connection management unit 15 notifies the expansion interval to the server device 2 when the persistent connection is set.

The message receiving unit 16 receives a message transmitted from the server device 2 through a persistent connection with the server device 2, and executes a predetermined process.

The persistent connection manager 11, the duration calculator 12, the timeout processor 13, the terminal identifier manager 14, the connection manager 15, and the message receiver 16 are computers that operate according to a program (communication maintenance program). This is realized by the CPU. For example, the program is stored in a storage unit (not shown) of the terminal device 1, and the CPU reads the program, and according to the program, the persistent connection management unit 11, the duration calculation unit 12, the timeout processing unit 13, the terminal identifier The management unit 14, the connection management unit 15, and the message reception unit 16 may be operated.

In addition, the persistent connection management unit 11, the duration calculation unit 12, the timeout processing unit 13, the terminal identifier management unit 14, the connection management unit 15, and the message reception unit 16 are each realized by dedicated hardware. May be.

The server device 2 includes a persistent connection receiving unit 21, a keep alive processing unit 22, a message transmitting unit 23, an active connection managing unit 24, and a timer 25.

The persistent connection receiving unit 21 manages a persistent connection with the terminal device 1. Specifically, the persistent connection receiving unit 21 transmits, to the terminal device 1, a keep alive packet for trying to determine whether the persistent connection is conductive or transmits a message.

The keep alive processing unit 22 tries based on the extension interval transmitted from the terminal device 1 whether the persistent connection is conductive after the extension interval has elapsed since the present time. Specifically, when a persistent connection is set, the keep alive processing unit 22 performs a connection between the terminal device 1 and the server device 2 after the extended interval notified when the persistent connection is set from that point. An event (hereinafter referred to as a keep alive event) for confirming the continuity of is generated. At this time, the keep alive processing unit 22 transmits a packet (keep alive packet) used for confirmation of continuity between the terminal device 1 and the server device 2 to the terminal device 1.

The message transmission unit 23 receives a message transmission request from the other server device 3 to the terminal device 1 and instructs the persistent connection reception unit 21 to transmit a message to the terminal device 1.

The active connection manager 24 instructs the persistent connection receiver 21 to transmit a message using the last established connection from the terminal device 1.

The other server device 3 includes a message transmission application 31 (hereinafter referred to as a message transmission application 31) that instructs the terminal device 1 to execute some processing.

Next, each operation of the communication maintenance system of this embodiment will be described.

(Persistent connection setting)
FIG. 2 is a flowchart illustrating an operation example in which the terminal device 1 sets a persistent connection. FIG. 3 is a flowchart illustrating an operation example in which the server apparatus 2 sets a persistent connection. Hereinafter, the operation of setting a persistent connection between the terminal device and the server device will be described with reference to FIGS. 2 and 3. Here, it is assumed that the duration calculation unit 12 holds the high reliability interval and the expansion interval.

The persistent connection management unit 11 of the terminal device 1 starts setting a persistent connection with the server device 2 (step S101 in FIG. 2). For example, the persistent connection management unit 11 is a server device at the time when the terminal device 1 is activated, when the network device (not shown) included in the terminal device 1 is activated, or when the user operates the terminal device 1. 2. Setting up a persistent connection with 2.

The persistent connection management unit 11 inquires of the duration calculation unit 12 about the high reliability interval and the extension interval of the persistent connection. The duration calculation unit 12 acquires the retained high reliability interval. Then, the duration calculation unit 12 determines whether the change mode used for the current period calculation is the enlargement mode, the reduction mode, or the steady mode (step S102). Thereafter, the expansion interval is determined based on the determined mode and the high reliability interval.

When the change mode is the steady mode (“steady mode” in step S102), the duration calculation unit 12 notifies the persistent connection management unit 11 of a result indicating that the extension interval and the high-reliability interval are the same (step S102). S103). In the expansion mode or the reduction mode, the duration calculation unit 12 calculates the expansion interval by calculation. When the change mode is the expansion mode (“enlargement mode” in step S102), the duration calculation unit 12 calculates a value larger than the expansion interval calculated last time as the current expansion interval (step S104). On the other hand, when the change mode is the reduction mode (“reduction mode” in step S102), the duration calculation unit 12 calculates a value smaller than the previously calculated extension interval as the current extension interval (step S105).

The duration calculation unit 12 may hold, for example, the previously determined expansion interval, and in the expansion mode, the extended interval may be a period obtained by adding a highly reliable interval to the previous expansion interval. On the other hand, in the reduction mode, the duration calculation unit 12 may calculate a value obtained by subtracting the high reliability interval from the previous expansion interval, and may set a period obtained by adding half of the value to the high reliability interval as the expansion interval.

However, the method for calculating the expansion interval is not limited to the method described above. For example, in the expansion mode, the duration calculation unit 12 may set a period twice as long as the high-reliability interval as the extension interval. In the reduction mode, the duration calculation unit 12 may use a value obtained by adding half the value of the high-reliability interval to the original high-reliability interval (that is, 1.5 times the high-reliability interval) as the extended interval. .

Note that the expansion interval is calculated to be longer than the high reliability interval, and the calculation method is arbitrary. Further, when no persistent connection is set once and there is no highly reliable interval, an initial value specified in advance by the terminal device may be set as the extended interval. However, the initial value is set shorter than the period during which the connection is invalidated by the middle box device 4. The duration calculation unit 12 operates in the expansion mode in the initial state. The duration calculation unit 12 notifies the persistent connection management unit 11 of the determined high-reliability interval and extension interval.

The persistent connection manager 11 issues a unique connection identifier for managing the persistent connection (step S106). The persistent connection management unit 11 notifies the timeout processing unit 13 of the issued connection identifier, high reliability interval, and extended interval.

The timeout processing unit 13 sets the timer 17 attached to the terminal device 1 so as to generate a high-reliability interval elapse event including information indicating the connection identifier and the high-reliability interval after elapse of the high-reliability interval from the current time. . Further, the timeout processing unit 13 sets the timer 17 so as to generate an extended interval elapsed event including a connection identifier after the extended interval has elapsed from the present time (step S107).

Note that when the high-reliability interval is 0, the timeout processing unit 13 does not set a timer for generating a high-reliability interval elapse event. Also, the timing for generating the extended interval elapsed event may not be the same as the extended interval. For example, the timeout processing unit 13 may set the timer 17 so as to generate an extended interval elapsed event after a predetermined amount of time has elapsed from the extended interval.

Next, the persistent connection management unit 11 acquires a terminal identifier for identifying the terminal device 1 from the terminal identifier management unit 14 (step S108). The terminal identifier may be information that can identify the terminal, such as the IP address of the terminal device 1 or the user ID of the user who uses the terminal device 1.

Next, the persistent connection management unit 11 instructs the connection management unit 15 to set a persistent connection by passing the connection identifier, the terminal identifier, and the extension interval. That is, the persistent connection management unit 11 and the connection management unit 15 specify the extension interval and set the persistent connection with the server device 2. The connection management unit 15 checks whether there is a connection classified as unused at present among the persistent connections set with the server device 2 (step S109).

If there is no connection classified as unused (“No” in step S109), the connection management unit 15 starts setting a new connection from the terminal device 1 to the server device 2 (step S110). When the connection management unit 15 sets a connection with the persistent connection receiving unit 21 of the server device 2, the connection management unit 15 manages the connection in association with the connection identifier.

On the other hand, if there is a connection classified as unused (“Yes” in step S109), the connection management unit 15 manages the connection in association with the connection identifier (step S111).

Next, the connection management unit 15 writes a notification including the terminal identifier and the extension interval to the persistent connection receiving unit 21 of the server device 2 through the persistent connection associated with the connection identifier (hereinafter referred to as a new keepalive period update notification). ) Is transmitted (step S112).

When the persistent connection receiving unit 21 of the server device 2 receives the new keep-alive period update notification through the persistent connection (step S201 in FIG. 3), it issues a unique connection identifier for the persistent connection (step S202). Then, the persistent connection receiving unit 21 stores the connection identifier and the persistent connection in association with each other. Further, the persistent connection receiving unit 21 notifies the active connection managing unit 24 of the connection identifier and the terminal identifier included in the new keepalive period update notification.

The active connection management unit 24 stores the terminal identifier and the connection identifier in association with each other. At this time, the active connection management unit 24 checks whether or not another connection identifier is already stored in association with a certain terminal identifier (step S203). When these identifiers are already associated (“Yes” in step S203), the active connection management unit 24 deletes the existing association (step S204). On the other hand, if these identifiers are not associated (“NO” in step S203), the process proceeds to step S205.

Next, the active connection management unit 24 stores the terminal identifier and the new connection identifier in association with each other (step S205). The persistent connection receiving unit 21 notifies the keep alive processing unit 22 of the connection identifier and the extension interval included in the new keep alive period update notification.

The keep alive processing unit 22 is attached to the server apparatus 2 so as to generate a keep alive event including information indicating the connection identifier and the extension interval after elapse of the extension interval from the present time (that is, the persistent connection is set). The timer 25 is set (step S206).

(Send Messege)
FIG. 4 is a flowchart illustrating an example of processing in which the server apparatus 2 transmits a message. FIG. 5 is a flowchart illustrating an example of processing in which the terminal device 1 receives a message. Hereinafter, the operation when the server apparatus 2 transmits a message to the terminal apparatus 1 will be described with reference to FIGS. 4 and 5.

When the message transmission unit 23 of the server device 2 receives a message transmission request to the terminal device 1 from the other server device 3 or each application in the server device 2 (step S301), the message transmission unit 23 specifies a terminal that transmits the message (step S301). Step S302).

The message transmission unit 23 may specify the terminal from the terminal identifier included in the transmission request, for example. When the transmission request includes another identifier different from the terminal identifier, the message transmission unit 23, for example, sends the terminal identifier to an external device (not shown in FIG. 1) that manages the identifier and the terminal identifier in association with each other. The terminal may be specified from the terminal identifier received from the external device.

The message transmission unit 23 notifies the persistent connection reception unit 21 of the terminal identifier and the message to be transmitted. The persistent connection receiving unit 21 acquires the connection identifier associated with the terminal identifier from the active connection management unit 24. Specifically, the active connection management unit 24 specifies a connection identifier from the association between the terminal identifier and the connection identifier held as active connection information in response to a request from the persistent connection receiving unit 21 (step S303). The active connection management unit 24 notifies the identified connection identifier to the persistent connection receiving unit 21, and cancels the generation of the keep alive event including the connection identifier (step S304).

Next, the persistent connection receiving unit 21 acquires a persistent connection associated with the connection identifier (step S305), and transmits a message to the connection management unit 15 of the terminal device 1 through the persistent connection (step S306).

When the connection management unit 15 of the terminal device 1 receives a message through a certain persistent connection (step S401 in FIG. 5), it notifies the message reception unit 16 of the message. The message receiving unit 16 executes processing according to the content of the message (step S402). Note that the processing performed by the message receiving unit 16 is uniquely determined by each terminal device 1, and thus detailed description thereof is omitted here.

Next, the connection management unit 15 acquires a connection identifier stored in association with the persistent connection that has received the message (step S403). The connection management unit 15 notifies the persistent connection management unit 11 that the message has been received via the persistent connection indicated by the connection identifier, and classifies the persistent connection as unused (step S404).

When the persistent connection management unit 11 is notified that the message has been received, it notifies the timeout processing unit 13 of the connection identifier. The timeout processing unit 13 cancels the occurrence of the high reliability interval elapsed event and the extended interval elapsed event including the notified connection identifier (S405). Further, the persistent connection management unit 11 executes the processing after step S101 immediately or after a predetermined time has elapsed. That is, the terminal device 1 sets up a persistent connection with the server device 2 again.

(High reliability interval elapses)
Next, a case where a highly reliable interval elapsed event occurs in the terminal device 1 will be described with reference to FIG. FIG. 6 is a flowchart illustrating an operation example when a highly reliable interval elapse event occurs in the terminal device 1.

When a highly reliable interval elapsed event occurs in the terminal device 1 (step S501), that is, when a message using the persistent connection indicated by the connection identifier included in the highly reliable interval elapsed event is not received during the highly reliable interval. The timeout processing unit 13 notifies the persistent connection management unit 11 of the elapse of the high reliability interval of the persistent connection indicated by the connection identifier.

The persistent connection manager 11 checks the current operation mode held by the duration calculator 12 (step S502). When the mode held by the duration calculation unit 12 is the steady mode (“Yes” in Step S502), the persistent connection management unit 11 acquires the current high-reliability interval from the duration calculation unit 12 (Step S503). Then, the persistent connection management unit 11 causes the timeout processing unit 13 to cancel the occurrence of the extended interval elapse event of the persistent connection (step S504).

If the high reliability interval included in the high reliability interval elapsed event is shorter than the current high reliability interval (“Yes” in step S505), the persistent connection management unit 11 ends the process. On the other hand, when the high-reliability interval included in the high-reliability interval elapsed event is equal to or greater than the current high-reliability interval (“No” in step S505), the processing from step S101 illustrated in FIG. 2 is executed. That is, a persistent connection is set up with the server device.

On the other hand, if the mode held by the duration calculation unit 12 is other than the steady mode in Step S502 (“No” in Step S502), the processing from Step S101 illustrated in FIG. 2 is executed. That is, a persistent connection is added to the server device. At this time, since the high-reliability interval is equal to or smaller than the extended interval, a plurality of persistent connections are set between the terminal device and the server device when the high-reliability interval is not equal to the extended interval.

(Keep-alive period elapses)
FIG. 7 is a flowchart illustrating an example of processing when a keep-alive event occurs in the server device 2. FIG. 8 is a flowchart illustrating an example of processing when the terminal device 1 receives a keep-alive packet. Hereinafter, the operation when the keep alive event occurs in the server device 2 will be described with reference to FIGS. 7 and 8.

When the keep alive event by the timer 25 occurs in the server device 2 (step S601 in FIG. 7), that is, the message transmission using the persistent connection indicated by the connection identifier included in the keep alive event has not occurred during the extended interval. In this case, the keep alive processing unit 22 receives a keep alive event. Then, the keep alive processing unit 22 instructs the persistent connection receiving unit 21 to transmit a keep alive packet using the persistent connection indicated by the connection identifier. The keep alive packet includes information indicating an extension interval included in the keep alive event.

The persistent connection receiving unit 21 acquires the persistent connection indicated by the connection identifier (step S602), and transmits a keep-alive packet (step S603). At this time, if the extension interval set for the persistent connection is the same as or shorter than the connection invalidation deadline set in the middle box device 4, the keep-alive packet reaches the terminal device 1. On the other hand, when the extended interval set for the persistent connection is longer than the connection invalidation deadline set by the middle box device 4, the persistent connection is invalidated, and the keep-alive packet does not reach the terminal device 1.

When the keep-alive packet arrives at the terminal device 1 and is received by the connection management unit 15 (step S701 in FIG. 8), the connection management unit 15 obtains a persistent identifier associated with the persistent connection (step S702). . Then, the connection management unit 15 notifies the persistent connection management unit 11 that the keepalive is received, and classifies the persistent connection as unused (step S703).

The persistent connection management unit 11 instructs the timeout processing unit 13 to cancel the occurrence of the high-reliability interval elapsed event and the extended interval elapsed event including the connection identifier (step S704).

Next, the persistent connection management unit 11 confirms the mode of the duration calculation unit 12 (step S705). When the mode held by the duration calculation unit 12 is the steady mode (“steady mode” in step S705), the persistent connection management unit 11 executes the processing after step S101 illustrated in FIG. That is, a persistent connection is set up with the server device.

Further, when the mode held by the duration calculation unit 12 is the reduction mode (“reduction mode” in step S705), the persistent connection management unit 11 changes the mode of the duration calculation unit 12 to the steady mode (step S706). The process is terminated.

In addition, when the mode held by the duration calculation unit 12 is the expansion mode (“expansion mode” in step S705), the persistent connection management unit 11 sets the extension interval included in the received keepalive packet as the high reliability interval. The calculation unit 12 is notified (step S707). The duration calculation unit 12 holds the received high reliability interval. This high-reliability interval is used when the expansion interval is determined next time in steps S103 to S105 illustrated in FIG.

In this way, the persistent connection management unit 11 and the connection management unit 15 are determined according to the reception state (for example, the change mode) of the keep alive packet when the high reliability interval has elapsed after setting the persistent connection. A new persistent connection designating the extension interval is set between the server apparatus 2 and the server apparatus 2. If the connection with the server device 2 is performed at a highly reliable interval, the middle box device 4 does not invalidate the persistent connection. Therefore, the connection between the terminal device 1 and the server device 2 can be maintained.

Furthermore, when the persistent connection management unit 11 receives a keep-alive packet based on a new persistent connection from the server device 2, the extended interval specified by the setting of the persistent connection is set as a highly reliable interval. By setting the extended interval to a highly reliable interval, the processing interval for maintaining the connection can be lengthened. That is, since the processing for maintaining the connection between both devices can be reduced, the communication load can be suppressed.

(Expansion interval elapses)
Next, an operation when an extended interval elapsed event occurs in the terminal device 1 will be described with reference to FIG. FIG. 9 is a flowchart illustrating a processing example when an extended interval elapsed event occurs in the terminal device 1.

When the extended interval elapsed event occurs in the terminal device 1 (step S801), that is, the timeout processing unit 13 does not receive a message using the persistent connection indicated by the connection identifier included in the extended interval elapsed event, and When the keep alive packet from the server apparatus 2 is not received before the extension interval elapses, the timeout processing unit 13 notifies the persistent connection management unit 11 of the timeout of the persistent connection indicated by the connection identifier. The persistent connection management unit 11 notifies the connection management unit 15 of the destruction of the persistent connection having the connection identifier.

The connection management unit 15 discards the persistent connection associated with the connection identifier (step S802). Next, the persistent connection management unit 11 instructs the duration calculation unit 12 to operate as the reduction mode (step S803). That is, the persistent connection management unit 11 changes the mode held by the duration calculation unit 12 to the reduction mode.

As described above, according to the present embodiment, the terminal device 1 holds the high-reliability interval and the extension interval, and specifies the extension interval to set a persistent connection with the server device 2. The server device 2 transmits a keep-alive packet to the terminal device 1 when the extended interval specified by the setting of the persistent connection has elapsed after the persistent connection is set with the terminal device 1.

In addition, the terminal device 1 sets a new persistent connection that specifies an extended interval that is determined according to the reception state (for example, change mode) of the keep-alive packet when the high-reliability interval elapses after the persistent connection is set. Is set with the server device 2. When the terminal device 1 receives a keep-alive packet based on a new persistent connection from the server device 2, the terminal device 1 sets the extended interval specified by the setting of the persistent connection as a highly reliable interval.

Therefore, even in a communication network in which the connection between devices becomes invalid in a predetermined period (for example, including the middle box device 4), the communication load for maintaining the connection can be suppressed while maintaining the connection between the devices.

That is, in the present embodiment, communication is performed at a highly reliable interval using any one of persistent connections set between devices. Therefore, it is possible to guarantee the arrival of packets between devices. Specifically, even if the extension interval of a certain persistent connection is set long, no response period is generated.

Further, even if the keep alive interval of the server device 2 at a certain time is smaller than the connection invalidation time limit of the middle box device 4, the connection interval of the middle box device 4 is invalidated earlier by setting the extended interval longer. You can approach the deadline. Therefore, useless load on the network can be reduced.

Hereinafter, the present invention will be described with reference to specific examples, but the scope of the present invention is not limited to the contents described below. FIG. 10 is an explanatory diagram illustrating an operation example of the communication maintaining system of the present embodiment. In FIG. 10, the horizontal axis indicates the elapsed time, and the vertical axis indicates the order in which the persistent connection is set in the terminal device 1.

In this embodiment, the connection invalidation deadline of the middle box device 4 is 9 minutes, and the initial expansion interval is 2 minutes. In addition, as the extension interval, a value obtained by adding the current high-reliability interval to the extension interval when the last persistent connection is set is used. In FIG. 10, the high-reliability interval when setting each persistent connection is indicated by “high =” and a white rectangle, and the extension interval designated when setting each persistent connection is indicated by “expansion =”. Further, in FIG. 10, the difference between the extension interval and the high-reliability interval specified when setting each persistent connection is indicated by a shaded rectangle.

In the initial state, since the high-reliability interval is not set, first, the terminal device 1 sets the persistent connection 1 having the extension interval of 2 minutes with the server device 2. In the following description, “after the passage” indicates the passage of time after the persistent connection 1 is set.

After 2 minutes, the server device 2 generates a keep alive event for the persistent connection 1. Then, the keep alive packet in which the keep alive period is set to 2 minutes reaches the terminal device 1. Therefore, in the terminal device 1, the high reliability interval is set to 2 minutes. At this timing, the terminal device 1 is connected to the server device 2 using the persistent connection 2 having a high reliability interval of 2 minutes and an expansion interval of 4 minutes.

After the elapse of 4 minutes (that is, when 2 minutes of the high-reliability interval has elapsed since the elapse of 2 minutes), the terminal device 1 generates a high-reliability interval elapse event for the persistent connection 2. Then, the terminal device 1 sets a persistent connection 3 with a high reliability interval of 2 minutes and an expansion interval of 6 minutes with the server device 2.

After 6 minutes, the server device 2 generates a keep alive event for the persistent connection 2. Then, the keep alive packet in which the keep alive period is set to 4 minutes arrives at the terminal device 1. Since the extension interval of the persistent connection 2 is 4 minutes, the high reliability interval is set to 4 minutes in the terminal device 1. At this timing, the terminal device 1 is connected to the server device 2 using the persistent connection 4 having a high reliability interval of 4 minutes and an expansion interval of 10 minutes.

After 10 minutes, the server device 2 generates a keep alive event for the persistent connection 3. Then, the keep alive packet in which the keep alive period is set to 6 minutes reaches the terminal device 1. Since the expansion interval of the persistent connection 3 is 6 minutes, the high reliability interval is set to 6 minutes in the terminal device 1. At this timing, the terminal device 1 is connected to the server device 2 using the persistent connection 5 having a high reliability interval of 6 minutes and an expansion interval of 16 minutes.

After 16 minutes, the server device 2 generates a keep alive event for the persistent connection 4. Then, the server device 2 transmits a keep alive packet in which the keep alive period is set to 10 minutes toward the terminal device 1. However, since the connection invalidation deadline of the middle box device 4 is 9 minutes, this keep-alive packet does not reach the terminal device 1. Therefore, at this timing, the terminal device 1 generates an extended interval elapsed event for the persistent connection 4, and the duration calculation unit 12 sets the mode to be held to the reduced mode.

In addition, after 16 minutes have elapsed, the terminal device 1 generates a highly reliable interval elapsed event for the persistent connection 5. Then, the terminal device 1 sets the persistent connection 6 with the high reliability interval of 6 minutes and the expansion interval of 11 minutes with the server device 2. Here, 11 minutes of the expansion interval is obtained by adding 6 minutes of the high reliability interval to half of the value (10 minutes) obtained by subtracting 6 minutes of the high reliability interval from the previous expansion interval of 16 minutes (that is, (Previous expansion interval 16 minutes-high reliability interval 6 minutes) ÷ 2 + high reliability interval 6 minutes) is calculated.

After 22 minutes, the terminal device 1 generates a highly reliable interval elapsed event for the persistent connection 6. Then, the terminal device 1 sets a persistent connection 7 having a high reliability interval of 6 minutes and an expansion interval of 9 minutes with the server device 2. Here, 9 minutes of the expansion interval is obtained by adding 6 minutes of the high reliability interval to half of the value (5 minutes) obtained by subtracting 6 minutes of the high reliability interval from 11 minutes that is the previous expansion interval (that is, (Previous expansion interval 11 minutes−high reliability interval 6 minutes) ÷ 2 + high reliability interval 6 minutes) is calculated. Note that 0.5 minutes is calculated by rounding up to 1 minute.

After 26 minutes, the server device 2 generates a keep alive event for the persistent connection 5. Then, the server device 2 transmits a keep alive packet in which the keep alive period is set to 16 minutes toward the terminal device 1. However, since the connection invalidation deadline of the middle box device 4 is 9 minutes, this keep-alive packet does not reach the terminal device 1. Therefore, at this timing, the terminal device 1 generates an extended interval elapsed event for the persistent connection 4. However, since the mode held by the duration calculation unit 12 is already the reduction mode, nothing is particularly performed.

After 27 minutes, the server device 2 generates a keep alive event for the persistent connection 6. Then, the server device 2 transmits a keep alive packet in which the keep alive period is set to 11 minutes toward the terminal device 1. However, since the connection invalidation deadline of the middle box device 4 is 9 minutes, this keep-alive packet does not reach the terminal device 1. Therefore, at this timing, the terminal device 1 generates an extended interval elapsed event for the persistent connection 4. However, since the mode held by the duration calculation unit 12 is already the reduction mode, nothing is particularly performed.

After 28 minutes, the terminal device 1 generates a highly reliable interval elapsed event for the persistent connection 7. Then, the terminal device 1 sets a persistent connection 8 with a high reliability interval of 6 minutes and an expansion interval of 8 minutes with the server device 2. Here, 8 minutes of the expansion interval is obtained by adding 6 minutes of the high reliability interval to half of the value (3 minutes) obtained by subtracting 6 minutes of the high reliability interval from 9 minutes that is the previous expansion interval (that is, (Previous expansion interval 9 minutes−high reliability interval 6 minutes) ÷ 2 + high reliability interval 6 minutes) is calculated. Note that 0.5 minutes is calculated by rounding up to 1 minute.

After 31 minutes, the server device 2 generates a keep alive event for the persistent connection 7. Then, the keep alive packet in which the keep alive period is set to 9 minutes reaches the terminal device 1. Since the connection invalidation deadline of the middle box device 4 is 9 minutes, this keep alive packet reaches the terminal device 1. Since the extension interval of the persistent connection 7 is 9 minutes, the high reliability interval is set to 9 minutes in the terminal device 1. At this time, the terminal device 1 generates an extended interval elapsed event for the persistent connection 4, and the duration calculation unit 12 sets the held mode to the steady mode. Thereafter, the terminal device 1 operates in a state where the high reliability interval is set to 9 minutes.

After 34 minutes have elapsed, the terminal device 1 generates a highly reliable interval elapsed event for the persistent connection 8. At this time, the mode held by the duration calculation unit 12 is a steady mode, and since the high-reliability interval of the event is 6 minutes and the current high-reliability interval is 9 minutes, nothing is performed.

After 36 minutes, the server device 2 generates a keep alive event for the persistent connection 8. Then, the server device 2 transmits a keep alive packet in which the keep alive period is set to 8 minutes toward the terminal device 1. In the terminal device 1, since the mode held by the duration calculation unit 12 is the steady mode, the terminal device 1 establishes a persistent connection 9 between the server device 2 and the high-reliability interval and the extension interval of 9 minutes. Set.

Thereafter, the server device 2 transmits a keep alive packet to the terminal device 1 every nine minutes, and the terminal device 1 sets a persistent connection with the server device 2 every time the keep alive packet is received.

Next, the outline of the present invention will be described. FIG. 11 is a block diagram showing an outline of a communication maintenance system according to the present invention. The communication maintenance system according to the present invention includes a terminal device 90 (for example, a terminal device) that communicates with the server device 80 using a persistent connection set between the server device 80 (for example, the server device 2) and the server device 80. 1).

The server device 80 and the terminal device 90 each have at least one middle box device (for example, invalidating the persistent connection) when communication using the persistent connection is not performed during a keep-alive interval that is a predetermined period. Are connected to each other via a middle box device 4).

The terminal device 90 is longer than the high-reliability interval, which is the maximum period of the keep-alive interval when it can be confirmed that the persistent connection is valid, and the high-reliability interval used for confirming whether the persistent connection is valid. A connection management unit 91 (for example, the persistent connection management unit 11 and the connection management unit 15) that holds the expansion interval of the period and sets the persistent connection with the server device 80 by specifying the expansion interval is included.

The server device 80 transmits a keep alive packet to the terminal device 90 when the extended interval specified by the setting of the persistent connection has elapsed since the persistent connection was set up with the terminal device 90. (For example, the persistent connection receiver 21).

The connection management unit 91 of the terminal device 90 creates a new persistent connection that specifies an extended interval determined according to the reception state of the keep-alive packet when the high-reliability interval elapses after setting the persistent connection. When the keep alive packet based on the new persistent connection is received from the server apparatus 80, the extended interval specified by the setting of the persistent connection is set as the high reliability interval.

With such a configuration, even in a communication network in which a connection between devices becomes invalid during a predetermined period, it is possible to suppress a communication load for maintaining the connection while maintaining the connection between the devices.

Specifically, the terminal device 90 has, as a change mode indicating a method for changing the extension interval, an extension mode in which the extension interval of the newly set persistent connection is gradually increased, and a reduction mode in which the extension interval is gradually reduced. Alternatively, it may include a duration calculation unit that holds any change mode of the steady mode that does not change the extension interval and determines the extension interval based on the change mode and the high-reliability interval. Then, the connection management unit 91 determines the change mode according to the reception state of the keep alive packet, and assigns a new persistent connection that specifies the extension interval determined based on the change mode and the high reliability interval to the server device 80. You may set between.

In addition, the duration calculation unit may hold the extended mode as the change mode in the initial state. Specifically, the connection management unit 91 changes when the duration calculation unit holds the extended mode as the change mode and does not receive a keep-alive packet based on a new persistent connection from the server device 80. The mode may be changed from the extended mode to the reduced mode, and a new persistent connection specifying the extended interval determined based on the changed mode and the high-reliability interval may be set between the server device 80 and the server device 80.

Furthermore, when the connection management unit 91 receives a keep-alive packet based on a new persistent connection from the server device 80 in a state where the duration calculation unit holds the reduced mode as the change mode, the connection management unit 91 changes the change mode from the reduced mode. The mode may be changed to the steady mode, and a new persistent connection may be set between the server apparatus and the high-reliability interval set based on the extended interval specified by the setting of the persistent connection.

In addition, when the change mode to be held is the steady mode, the duration calculation unit determines the high-reliability interval as the extension interval, and when the change mode to be held is the extension mode, the interval indicating a period longer than the previously determined extension interval If the extension mode is determined and the change mode to be held is the reduction mode, the extension interval may be determined as an interval indicating a period shorter than the previously determined extension interval.

FIG. 12 is a block diagram showing an outline of the terminal device according to the present invention. The terminal device according to the present invention is the same as the terminal device 90 illustrated in FIG. By using the terminal device illustrated in FIG. 12, the communication load for maintaining the connection can be suppressed while maintaining the connection between the devices even in the communication network in which the connection between the devices is invalidated for a predetermined period.

Some or all of the above embodiments may be described as follows, but are not limited to the following descriptions.

The communication maintenance system provides a service to a terminal device operated by a user, a server device that mediates message transmission to the terminal device, and a user who operates the terminal device by transmitting a message to the terminal device. You may provide the middle server apparatus which monitors, relays, and cancels communication between a terminal device and a server apparatus by another server apparatus and a firewall function or a NAT function. At this time, the terminal device starts setting a persistent connection with the server device when the terminal is activated, when the network device on the terminal device is activated, or when the user operates the server device. The maximum period that can be confirmed that the persistent connection maintained between the server and the server is valid is the high reliability interval, and the interval for confirming whether the persistent connection set with the server is valid is the extended interval. If the maximum value of the high-reliability interval is determined, a persistent connection is newly established with the server device after the high-reliability interval elapses, and the maximum value of the high-reliability interval is determined. When a keep-alive packet is received that confirms whether or not the persistent connection is valid, a new persistent connection is set up with the server device. With the persistent connection management unit that maintains the persistent connection with the server device, the extended mode that maintains the high reliability interval and the expansion interval and gradually increases the expansion interval of the persistent connection that is newly set, Between the server unit and the duration calculation unit that retains the reduced mode that gradually decreases the expansion interval or the steady mode that does not change the expansion interval and determines the expansion interval based on that mode and the high-reliability interval A time-out processing unit that sets a timer associated with the terminal to generate a high-reliability interval elapse event after elapse of a high-reliability interval from the current time, and an extended interval elapse event after elapse of the extended interval from the current time, A TCP / IP connection between the terminal identifier management unit that acquires a terminal identifier for uniquely identifying the device and the server device The message is sent from the server device through the persistent connection between the server and the connection management unit that notifies the server device of the expansion interval when it is set, and executes the default processing. And a message receiving unit. In addition, the server device manages a persistent connection with the terminal device, and transmits a keep alive packet to try whether or not the persistent connection is conductive, or transmits a message to the terminal device. Based on the extended connection received from the persistent connection receiving unit, the terminal device, a keep alive processing unit that tries to determine whether the persistent connection is conductive after the extended interval from the current time, and a message transmission from the other server device to the terminal device A message transmission unit that accepts the request and instructs the persistent connection reception unit to transmit a message to the terminal device;
An active connection management unit that instructs the persistent connection receiving unit to transmit a message using the last-established persistent connection from the terminal device may be included.

In addition, the terminal device of the communication maintenance system, when the terminal device is activated, when the network device on the terminal device is activated, or when the user operates, the persistent connection management unit establishes the persistent connection with the server. Start setting, inquire the duration calculation unit about the high-reliability interval and extended interval of the persistent connection, the duration calculation unit acquires the high-reliability interval held, and the current period calculation is expanded Determine whether the mode, reduced mode or steady mode. In the steady mode, the expansion interval is the same as the reliable interval and is passed to the persistent connection manager, and the expansion interval calculated in the expansion mode is larger than the previously calculated expansion interval. The extension interval calculated in the reduced mode is calculated as a value smaller than the previously calculated extension interval, and the persistent connection management unit A unique connection identifier for management is issued, the connection identifier, the high-reliability interval, and the extended interval are passed to the timeout processing unit, and the time-out processing unit connects the connection identifier and the high-reliability interval after the high-reliability interval elapses from the current time. A high-reliability interval elapse event that includes a connection identifier and an extended interval elapse event that includes a connection identifier after the elapse of an extension interval from the current time, a timer attached to the terminal is set, and the persistent connection management unit sends the terminal identifier management unit to the terminal A terminal identifier for identifying the terminal is acquired, the connection identifier, the terminal identifier, and the extension interval are passed to the connection management unit, and the setting of the persistent connection is instructed. The connection management unit is set with the server device. If there is a connection that is classified as currently unused, and if there is no unused connection, the terminal device Starts a new connection setting for the server device, sets a connection with the persistent connection receiver of the server device, manages the connection in association with the connection identifier, and if there is an unused connection, The connection may be managed in association with the connection identifier, and a new keep alive period update notification including the terminal identifier and the extension interval may be transmitted to the persistent connection receiver of the server device through the persistent connection associated with the connection identifier.

The server device of the communication maintenance system issues a unique connection identifier to the persistent connection when the persistent connection receiver receives the new keep-alive period update notification through the persistent connection, and the connection identifier and the persistent connection And the connection identifier and the terminal identifier included in the new keep-alive period update notification are passed to the active connection management unit, and the active connection management unit Check whether another connection identifier is stored in association with it, and if so, delete the existing association, store the terminal identifier and the new connection identifier in association with each other, and the persistent connection reception unit performs the keep-alive process. Pass the connection identifier and the extended interval included in the new keep-alive period update notification. Live processing unit, connected from the current time after a lapse of diastolic interval identifier may be set to the accompanying timer device to generate a keep-alive events including the diastolic interval.

In addition, when the connection management unit receives a message through a certain persistent connection, the terminal device of the communication maintenance system passes the message to the message reception unit, and the message reception unit executes processing according to the content of the message, The management unit acquires a connection identifier stored in association with the persistent connection that has received the message, notifies the persistent connection management unit that the message has been received through the persistent connection indicated by the connection identifier, and When it is marked as used and the persistent connection manager is notified of the receipt of the message, the connection identifier is passed to the timeout processing unit, and the timeout processing unit includes a reliable interval elapsed event including the passed connection identifier and Cancel the occurrence of extended interval elapsed event, default It may be set persistent connection between the server again device after elapse.

In addition, when a highly reliable interval elapse event occurs, the terminal device of the communication maintenance system causes the time-out processing unit to indicate the elapse of the reliable interval of the persistent connection indicated by the connection identifier of the highly reliable interval elapse event to the persistent connection management unit. The persistent connection management unit confirms the current operation mode of the duration calculation unit, and if the operation mode of the duration calculation unit is in a steady state, obtains the current high-reliability interval from the duration calculation unit, Cancels the occurrence of the extended interval elapsed event of the persistent connection for the timeout processing unit, and terminates the process if the high-reliability interval included in the high-reliability interval elapse event is shorter than the current high-reliability interval, and the high-reliability interval elapses If the high-confidence interval included in the event is the same or longer than the current high-reliability interval, or the operation mode of the duration calculation unit is other than steady state If it may be added persistent connection with the server device.

In addition, in the server device of the communication maintenance system, when a keep alive event occurs, the keep alive processing unit is included in the keep alive event with respect to the continuous connection indicated by the connection identifier of the keep alive event with respect to the continuous connection receiving unit. It may be instructed to transmit a keep alive packet having the extended interval as a content, and the persistent connection receiving unit may acquire the persistent connection indicated by the connection identifier and transmit the keep alive packet.

In addition, when the terminal device of the communication maintenance system receives the keep alive packet at the connection management unit, the connection management unit acquires the persistent identifier associated with the persistent connection, and receives the keep alive from the persistent connection management unit. , Mark the persistent connection as unused, and the persistent connection manager cancels the occurrence of the reliable interval elapsed event and extended interval elapsed event including the connection identifier to the timeout processing unit, and calculates the duration. If the mode of the duration calculation unit is the steady mode, a persistent connection is set again with the server device, and if the mode of the duration calculation unit is the reduced mode, the persistent connection management unit Sets the mode of the duration calculation unit to the steady mode, and the mode of the duration calculation unit is the expansion mode If, notified to the duration calculator extended intervals included in keepalive packet received as reliable interval, duration calculator may hold a reliable interval received.

In addition, when an extended interval elapsed event occurs, the terminal device of the communication maintenance system notifies the persistent connection manager of the timeout of the persistent connection indicated by the connection identifier to the persistent connection manager, and the persistent connection manager Notify the connection manager of the destruction of the persistent connection with the connection identifier, the connection manager discards the persistent connection associated with the connection identifier, and the persistent connection manager reduces the duration calculation unit It may be instructed to operate as a mode.

As mentioned above, although this invention was demonstrated with reference to embodiment and an Example, this invention is not limited to the said embodiment and Example. Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

This application claims priority based on Japanese Patent Application 2012-158664 filed on July 17, 2012, the entire disclosure of which is incorporated herein.

The present invention is preferably applied to a communication maintenance system that maintains communication between devices.

DESCRIPTION OF SYMBOLS 1 Terminal apparatus 2 Server apparatus 3 Other server apparatus 4 Middle box apparatus 11 Persistent connection management part 12 Persistence period calculation part 13 Timeout processing part 14 Terminal identifier management part 15 Connection management part 16 Message receiving part 21 Persistent connection receiving part 22 Keep alive process Unit 23 Message transmission unit 24 Active connection management unit 31 Message transmission application

Claims

A server device;
A terminal device that communicates with the server device using a persistent connection set up with the server device;
The server device and the terminal device may include at least one middle box device that invalidates the persistent connection when communication using the persistent connection is not performed during a keep-alive interval that is a predetermined period. Connected to each other through
The terminal device
A high-reliability interval that is the maximum period of keep-alive intervals when it is confirmed that the persistent connection is valid, and a period that is longer than the reliable interval that is used to confirm whether the persistent connection is valid A connection management unit for holding an extension interval, and setting a persistent connection with the server device by specifying the extension interval;
The server device
A packet transmission unit that transmits a keep-alive packet to the terminal device when an extended interval specified in the setting of the persistent connection has elapsed since a persistent connection was set up with the terminal device;
The connection management unit of the terminal device creates a new persistent connection that specifies an extension interval determined according to a reception state of the keep-alive packet when the high-reliability interval elapses after the persistent connection is set. Set with the server device, and when the keep-alive packet based on the new persistent connection is received from the server device, the extended interval specified by the setting of the persistent connection is set to a highly reliable interval. A communication maintenance system.
As a change mode that indicates a method for changing the extension interval, the terminal device uses an expansion mode that gradually increases the extension interval of the newly set persistent connection, a reduction mode that gradually decreases the extension interval, or an extension interval. A duration calculation unit that holds any change mode of the steady mode that does not change and determines an expansion interval based on the change mode and the high-reliability interval,
The connection management unit determines the change mode according to the reception state of the keep-alive packet, and establishes a new persistent connection between the server device and the extended interval determined based on the change mode and the high-reliability interval. The communication maintenance system according to claim 1 to be set.
The duration calculation unit maintains the extended mode as the change mode in the initial state,
The connection management unit changes the change mode from the extended mode to the reduced mode when a keepalive packet based on a new persistent connection is not received from the server device while the duration calculation unit holds the extended mode as the change mode. The communication maintenance system according to claim 2, wherein a new persistent connection that specifies an extension interval determined based on the change mode and the high-reliability interval is set with the server device.
The connection management unit changes the change mode from the reduced mode to the steady mode when a keep alive packet based on a new persistent connection is received from the server device while the duration calculation unit holds the reduced mode as the change mode. The communication maintenance system according to claim 3, wherein a new persistent connection that specifies a high-reliability interval that is set based on an extension interval that is specified by the setting of the persistent connection is set between the server apparatus and the server apparatus.
When the change mode to be held is the steady mode, the duration calculation unit determines the high-reliability interval as the extension interval, and when the change mode to be held is the extension mode, the duration calculation unit extends to an interval that indicates a period longer than the previously determined extension interval. The interval is determined, and when the change mode to be held is the reduction mode, the expansion interval is determined to be an interval that indicates a period shorter than the previously determined expansion interval. Communication maintenance system.
A terminal device that communicates with the server device using a persistent connection set with the server device,
The server device is connected via at least one middle box device that invalidates the persistent connection when communication using the persistent connection is not performed during a predetermined period of a keep-alive interval. And
A high-reliability interval that is the maximum period of keep-alive intervals when it is confirmed that the persistent connection is valid, and a period that is longer than the reliable interval that is used to confirm whether the persistent connection is valid A connection management unit for holding an extension interval, and setting a persistent connection with the server device by specifying the extension interval;
The connection management unit, when the high-reliability interval elapses after setting the persistent connection, the extended interval specified by the setting of the persistent connection since the persistent connection is established with the server device A new persistent connection that specifies an extension interval determined according to a reception state of a keep-alive packet transmitted from the server device is set between the server device and the keep based on the new persistent connection. When the alive packet is received from the server device, the extended interval designated by the setting of the persistent connection is set as a highly reliable interval.
Communication with the server apparatus is performed using a persistent connection set with the server apparatus, and communication using the persistent connection is not performed with the server apparatus during a predetermined period. A terminal device connected via at least one middle box device that invalidates the persistent connection is a maximum period of a keep-alive interval when it is confirmed that the persistent connection is valid. Holding a certain reliable interval and an extended interval of a longer period than the reliable interval used for checking whether the persistent connection is valid;
The terminal device sets a persistent connection with the server device by specifying the extension interval,
The server device transmits a keep-alive packet to the terminal device when an extended interval specified in the setting of the persistent connection has elapsed since a persistent connection was set up with the terminal device,
The terminal device establishes a new persistent connection specifying an extension interval determined according to a reception state of the keepalive packet with the server device when the high-reliability interval elapses after setting the persistent connection. Set between
When the terminal device receives a keep-alive packet based on the new persistent connection from the server device, the terminal device sets the extended interval specified by the setting of the persistent connection as a highly reliable interval. Method.
As a change mode that indicates a method for changing the extension interval, the terminal device uses an expansion mode that gradually increases the extension interval of the newly set persistent connection, a reduction mode that gradually decreases the extension interval, or an extension interval. Hold one of the change modes of the steady mode that does not change,
The terminal device determines an expansion interval based on the change mode and the high reliability interval,
The terminal device determines the change mode according to the reception state of the keep alive packet,
The communication maintenance method according to claim 7, wherein the terminal device sets a new persistent connection that specifies an extension interval determined based on the change mode and the high-reliability interval with the server device.
Communication with the server apparatus is performed using a persistent connection set with the server apparatus, and communication using the persistent connection is not performed with the server apparatus during a predetermined period. A terminal device connected via at least one middle box device that invalidates the persistent connection is a maximum period of a keep-alive interval when it is confirmed that the persistent connection is valid. Holding a certain reliable interval and an extended interval of a longer period than the reliable interval used for checking whether the persistent connection is valid;
The terminal device sets a persistent connection with the server device by specifying the extension interval,
When the high-reliability interval has elapsed since the terminal device set the persistent connection, the extended interval specified in the persistent connection setting has elapsed since the persistent connection was established with the server device. A new persistent connection that specifies an extension interval determined according to the reception state of a keep-alive packet transmitted from the server device is sometimes set between the server device and
When the terminal device receives a keep-alive packet based on the new persistent connection from the server device, the terminal device sets the extended interval specified by the setting of the persistent connection as a highly reliable interval. How to maintain connection with.
Communication with the server apparatus is performed using a persistent connection set with the server apparatus, and communication using the persistent connection is not performed with the server apparatus during a predetermined period. A connection maintenance program applied to a computer connected via at least one or more middlebox devices to invalidate the persistent connection,
In the computer,
A high-reliability interval that is the maximum period of keep-alive intervals when it is confirmed that the persistent connection is valid, and a period that is longer than the reliable interval that is used to confirm whether the persistent connection is valid Holding an extension interval, specifying the extension interval and executing a connection management process for setting a persistent connection with the server device;
In the connection management process, when the high-reliability interval elapses after setting the persistent connection, the extended interval specified in the setting of the persistent connection has elapsed since the persistent connection was established with the server device. A new persistent connection that specifies an extension interval determined according to a reception state of a keep-alive packet transmitted from the server device is set between the server device and the keep based on the new persistent connection. A connection maintenance program that, when receiving an alive packet from the server device, causes the extended interval specified by the setting of the persistent connection to be set to a highly reliable interval.