JP2008067056A - Network system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a network system for providing a network service among a plurality of networks concerning a network configuration composed of the plurality of networks. <P>SOLUTION: When a content distribution flow from a content server CS_B to a user terminal CLN is formed, a policy server PS_A extracts a route candidate for securing request quality in the network MPLS_A. A policy server PS_B extracts a route candidate for securing the request quality in the network MPLS_B, so as to report it to the policy server PS_A. The policy server PS_A determines a transmission route to satisfy the prescribed request quality among the respective route candidate combinations, and sets the transfer rule of each router. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a network system including a plurality of network networks having different operation policies, and more particularly to a technique for ensuring transmission quality of a flow flowing through a plurality of network networks.

  With the recent development of information technology, distribution and broadcasting of contents via an IP (Internet Protocol) network are being realized. In order to spread such network services, it is necessary to maintain transmission quality (for example, transmission bandwidth, delay time, etc.) at a predetermined required value. In other words, multimedia content such as video and audio is often distributed by a streaming method in which data is received and played back at the same time, and the data is received due to a narrow transmission band or a long delay time. This is because video and audio cannot be reproduced smoothly if the time is delayed.

  Therefore, a traffic control technology (TE: Traffic Engineering) called QoS (Quality of Service) has been put into practical use. QoS is a technology that secures a transmission band for specific data transmission and guarantees a constant transmission speed.

  As an example of a configuration for realizing QoS on an IP network, MPLS (Multi-Protocol Label Switching) technology has been proposed. In a network to which MPLS is applied, a policy that controls each router according to a preset operation policy in addition to a router that sends the data packet to a transfer destination in accordance with destination information added to a received data packet Server (Policy Server) may be deployed. This operation policy is a network operation policy that defines the control contents of the network in association with the conditions that can occur.

As an example, Japanese Patent Laying-Open No. 2002-359639 (Patent Document 1) discloses an autonomous cooperative QoS control system that can improve the throughput of the entire network.
JP 2002-359639 A

  Along with an increase in network transmission capacity and the like, it has been proposed to provide a network service as described above by connecting a plurality of network networks to each other. That is, it is assumed that a content is distributed from a server arranged in a certain network to another network.

  In such a form, since the data flow of the content to be distributed flows through a plurality of network networks, it is necessary to ensure transmission quality in each network network.

  However, since each policy server placed in the network only selects the optimum route in its own network according to its own operation policy, the optimum transmission route is not always selected when viewed as a whole data flow. Not. As a result, necessary transmission quality cannot be ensured, and a flow may not be formed.

  For this reason, it has been impossible to reliably provide a network service satisfying a predetermined required quality across a plurality of network networks.

  Accordingly, the present invention has been made to solve such a problem, and an object of the present invention is to provide a network service that satisfies a predetermined required quality between a plurality of network networks in a network configuration including a plurality of network networks. Is to provide a network system.

  According to the present invention, the network system includes a plurality of network networks connected to each other. Each of the network networks includes, according to a predetermined transfer rule, at least one router that sends a received data packet to a transfer destination according to information indicating a destination added to the data packet, and a corresponding network network, A policy server that sets router forwarding rules so that a flow of data packets sent from a specific source to a specific destination flows through a desired transmission path, and the policy servers are configured to be communicable with each other Is done. Further, each of the policy servers includes request command receiving means for receiving a request command relating to formation of a flow having a predetermined required quality between a transmission source of its own network and a destination of another network, In a network, route candidate extraction means for extracting a route candidate capable of ensuring the required quality and a route candidate capable of ensuring the required quality in the network network corresponding to the other policy server are acquired from another policy server. Based on the destination route candidate acquisition unit, the route candidate extracted by the route candidate extraction unit, and the route candidate acquired by the destination route candidate acquisition unit, the transmission source of the own network network and the destination of another network network Route determining means for determining a transmission route capable of ensuring the required quality, and the route determining means. The flow is formed along the transmission path determined by the setting means for setting the transfer rule of the router existing in the transmission path of the own network and the path determination means so that the flow is formed along the transmission path. And a transmission path notification means for notifying other policy servers of information related to the determined transmission path in order to set a transfer rule of a router existing in the transmission path of the other network. .

  Preferably, the required quality includes a required transmission band, and the path determination unit determines a transmission path based on a delay time of each path candidate and a load factor indicating a ratio of a used band.

  Preferably, each of the routers includes a failure determination unit that determines whether or not a failure has occurred in the transmission path related to the transmission based on success or failure of transmission of the data packet to the transfer destination, and a failure determination unit When the occurrence of a failure is determined, a failure occurrence notifying unit that notifies the corresponding policy server of the occurrence of the failure is provided. Upon receiving a notification of the occurrence of a failure from the router, the request command accepting unit accepts it as a request command for assigning another transmission route to the flow formed including the route in which the failure has occurred.

  Preferably, the network networks are connected to each other via a plurality of inter-network paths, and the route determination means includes a fault flow among a plurality of inter-network paths connecting the transmission source network network and another network network. A route candidate including the same intra-network route as the intra-network route included in the passing route is preferentially determined as a transmission route.

  Also preferably, the network system according to the present invention is arranged in any one of the network networks and configured to be capable of distributing a plurality of contents, and is arranged in the same network network as the content server, and from the user terminal And a content request receiving server for receiving a content distribution request to the content server. When the content request reception server receives a distribution request from the user terminal, the content request reception server notifies the policy server of the same network network as the content server that can distribute the requested content of the flow for distributing the content. When the request instruction receiving unit receives a flow formation notification for distributing content from the content request receiving server, the request command receiving unit distributes the content to the user terminal that issued the distribution request from the content server. It is accepted as a request command to ensure the required quality.

  According to this invention, the policy servers arranged in each of the plurality of network networks connected to each other are configured to be able to communicate with each other. When a request command is given to one of the policy servers, the policy server extracts a route candidate capable of ensuring the required quality in its own network, and from the other policy server to the other policy server. The route candidate that can ensure the required quality in the network corresponding to is acquired. Then, the policy server determines a transmission path that can ensure the required quality between a transmission source existing in its own network and a destination existing in another network. Thereby, it is possible to reliably form a flow that spans a plurality of network networks.

  Therefore, in a network configuration composed of a plurality of network networks, it is possible to realize a network system for providing a network service that satisfies a predetermined required quality between the plurality of network networks.

  Embodiments of the present invention will be described in detail with reference to the drawings. Note that the same or corresponding parts in the drawings are denoted by the same reference numerals and description thereof will not be repeated.

  FIG. 1 is a schematic configuration diagram of a network system 100 according to the embodiment of the present invention.

  Referring to FIG. 1, the network system 100 includes three network networks MPLS_A, MPLS_B, and MPLS_C. The network networks MPLS_A, MPLS_B, and MPLS_C are connected to each other via a two-network path, and are configured to be able to transmit data between the networks.

  The network network MPLS_A includes an autonomous system router (ASR) ASR_A1, ASR_A2, ASR_A3, ASR_A4, a policy server PS_A, a content server CS_A, and a content server CS_A. And a content request receiving server CDS_A. A loop path is formed in the order of the router ASR_A1, the router ASR_A2, the router ASR_A3, and the router ASR_A4.

  Each of the routers ASR_A1 to ASR_A4 is connected to two or more paths, and sends a received data packet to a transfer destination according to destination information added to the data packet according to a predetermined transfer rule. As the destination information, a destination IP (Internet Protocol) address or a label added in advance is used.

  In the following, regarding the flow of data packets flowing between network networks, a router connected to a boundary with another network network, that is, an inter-network path for interconnecting network networks, particularly an autonomous system boundary router (ASBR: Autonomous). System Border Router (hereinafter simply referred to as “border router”).

  The policy server PS_A sets the transfer rule of each of the routers ASR_A1 to ASR_A4 so that the flow of the data packet transmitted from the specific transmission source to the specific destination flows through the desired transmission path in the network MPLS_A. .

  The content server CS_A is connected to the router ASR_A1, is disposed in the network MPLS_A, and is configured to be able to distribute a plurality of contents. Specifically, the content server CS_A stores a plurality of multimedia contents such as video and audio, and starts distributing the stored content in response to a content distribution request given from the content request receiving server CDS_A.

  The content request reception server CDS_A is arranged in the network MPLS_A in association with the content server CS_A, and receives a content distribution start request from the user terminal CLN. Specifically, the content request receiving server CDS_A obtains a list of contents stored in the content server CS_A, and provides the list as a Web page (html format electronic data) for access from the user terminal CLN. At the same time, in response to a specified operation of the list (for example, selection and click with the cursor), it is determined which content is a content distribution start request.

  The network MPLS_B includes routers ASR_B1, ASR_B2, ASR_B3, ASR_B4, a policy server PS_B, a content server CS_B, and a content request receiving server CDS_B. A loop path is formed in the order of the router ASR_B1, the router ASR_B2, the router ASR_B3, and the router ASR_B4.

  Since routers ASR_B1 to ASR_B4 are similar to routers ASR_A1 to ASR_A4 described above, detailed description will not be repeated. Policy server PS_B, content server CS_B, and content request accepting server CDS_B are also similar to policy server PS_A, content server CS_A, and content request accepting server CDS_A, respectively, and therefore detailed description thereof will not be repeated.

  The network MPLS_C includes routers ASR_C1, ASR_C2, ASR_C3, and a policy server PS_C. Then, a loop path is formed in the order of the router ASR_C1, the router ASR_C2, and the router ASR_C3.

  Routers ASR_C1 to ASR_C3 are similar to router ASR_A1 and the like described above, and detailed description thereof will not be repeated. Since policy server PS_C is similar to policy server PS_A, detailed description thereof will not be repeated.

  Further, the policy servers PS_A, PS_B, and PS_C are configured to be able to communicate with each other via the control network CNW. Then, each of the policy servers PS_A, PS_B, and PS_C cooperates with each other to determine a transmission path that can ensure a predetermined required quality in the network system 100, and according to the determined transmission path, as will be described later. Set the forwarding rules for each router.

  Further, content request receiving servers CDS_A and CDS_B are also connected to the control network CNW. The content request receiving servers CDS_A and CDS_B and the policy servers PS_A, PS_B and PS_C are configured to be capable of mutual communication.

  In the embodiment of the present invention, the case where content data is distributed from the content server CS_B to the user terminal CLN in response to a distribution request from the user terminal CLN connected to the router ASR_A1 of the network MPLS_A is illustrated. To do. Furthermore, in the embodiment of the present invention, processing when a network failure occurs in the transmission path during distribution of the content data is also exemplified.

  In the following description, network networks MPLS_A, MPLS_B, and MPLS_C are collectively referred to as “network network MPLS”, policy servers PS_A, PS_B, and PS_C are collectively referred to as “policy server PS”, and routers ASR_A1 to ASR_A4 and ASR_B1 to ASR_B4 , ASR_C1 to ASR_C3 are collectively referred to as “router ASR”, the content servers CS_A and CS_B are collectively referred to as “content server CS”, and the content request reception servers CDS_A and CDS_B are collectively referred to as “content request reception server CDS”.

FIG. 2 is a schematic configuration diagram of the policy server PS.
Referring to FIG. 2, as an example, each policy server PS includes a communication unit 14, an interface (I / F) unit 16, a network failure detection unit 18, a path control unit 10, a database 12, a router. A setting unit 20.

  The communication unit 14 mutually transmits a control command and the like between the other policy server PS and the content request receiving server CDS interconnected via the control network CNW (FIG. 1). That is, the communication unit 14 sends a control command to another policy server PS or content request receiving server CDS via the control network CNW, and receives a control command given from the other policy server PS or content request receiving server CDS. This is given to the route control unit 10.

  The interface unit 16 is connected to the router ASR existing in the corresponding network network MPLS, and transmits the transfer rule given from the router setting unit 20. Further, the interface unit 16 receives information on the occurrence of a network failure (data transmission failure or increase in data loss rate) from the router ASR existing in the corresponding network network MPLS, and the failure occurrence information is sent to the network failure detection unit 18. Give to.

  The network failure detection unit 18 determines a route (range) in which a network failure has occurred based on the occurrence information of the network failure received via the interface unit 16 and gives the determination result to the route control unit 10. .

  The path control unit 10 accepts a request command relating to formation of a flow having a predetermined required quality between a transmission source of its own network network MPLS and a destination of another network network MPLS. Specifically, when the route control unit 10 receives a flow formation notification related to content distribution from the content request reception server CDS via the communication unit 14, the route control unit 10 transmits the notification from the corresponding content server CS to the user terminal CLN. It operates to form a flow that can ensure the required quality for content distribution (content distribution flow formation). Further, when the route control unit 10 receives a notification of the occurrence of a network failure from any router via the network failure detection unit 18, the flow (hereinafter referred to as “failure flow”) along the route in which the failure has occurred. It operates so that another transmission path may be assigned to the other (also referred to as fault flow switching).

  Then, the path control unit 10 executes processing corresponding to each of (1) content distribution flow formation and (2) fault flow switching.

  (1) As a specific process for forming a content distribution flow, the path control unit 10 can ensure the required quality related to content distribution in the network network MPLS to which the route control unit 10 belongs based on information stored in the database 12 Extract route candidates. Further, the route control unit 10 notifies the route extraction request and the required quality to the other policy server PS via the communication unit 14, and from the other policy server PS to the network network MPLS to which the other policy server PS belongs. Then, route candidates that can ensure the same required quality are acquired.

  On the other hand, when the route control unit 10 of another policy server PS receives the route extraction request and the required quality, the route control unit 10 extracts a route candidate capable of ensuring the notified required quality in the network network MPLS to which the route control unit 10 belongs, Reply with route candidate.

  Then, the route control unit 10 determines a transmission route that can ensure the required quality between the content server CS and the user terminal CLN based on the extracted route candidate and the route candidate acquired from another policy server PS. decide.

  (2) As a specific process of fault flow switching, the path control unit 10 generates a network fault based on information stored in the database 12 according to fault occurrence information notified from the network fault detection unit 18. The fault flow formed including the route is extracted. Then, the path control unit 10 extracts path candidates that can be substituted in the network network MPLS to which the path control unit 10 belongs based on information stored in the database 12 according to the transmission source and destination of the fault flow. Further, the path control unit 10 notifies the network policy extraction request and the required quality to the other policy server PS via the communication unit 14, and the network network to which the other policy server PS belongs from the other policy server PS. In MPLS, route candidates that can be substituted are acquired.

  On the other hand, when the route control unit 10 of another policy server receives the inter-network route extraction request and the required quality, the route control unit 10 extracts a route candidate that can ensure the notified required quality in the network network MPLS to which the route control unit 10 belongs. It responds with the route candidate.

  Then, the path control unit 10 determines a transmission path that can ensure the required quality between the source and destination of the fault flow based on the extracted path candidate and the path candidate acquired from the other policy server PS. decide.

  As described above, the route control unit 10 outputs information on the transmission route determined in any one of (1) content distribution flow formation and (2) fault flow switching to the router setting unit 20. Further, the path control unit 10 notifies the information related to the determined transmission path to the other policy server PS.

  The router setting unit 20 sets a transfer rule for the router ASR existing in the transmission path of the network network MPLS to which the router setting unit 20 belongs so that a flow is formed along the transmission path determined by the path control unit 10.

  On the other hand, upon receiving information related to the determined transmission path, the router setting unit 20 of the other policy server PS sets a transfer rule for the router ASR existing in the determined transmission path in the network to which it belongs. .

FIG. 3 is a diagram illustrating a data configuration stored in the database 12.
Referring to FIG. 3, the database 12 includes network configuration information 12a indicating the configuration of the network network to which the database 12 belongs, and network operation management information 12b indicating the current status of the network network to which the database 12 belongs. The database 12 is configured so that data can be written to and read from the path control unit 10.

  As the contents of the network configuration information 12a, "static" configuration data of the network such as the topology of the network to which the network belongs, the route, the bandwidth of each link, the delay time of each node, and the like are stored.

  The network operation management information 12b stores the current status, that is, the “dynamic” status value of the network network, for each path of the network network to which the network operation management information 12b belongs.

FIG. 4 is a diagram illustrating an example of the contents of the network operation management information 12b.
Referring to FIG. 4, network operation management information 12b includes a number No. Corresponding to these, a path column 80, a bandwidth column 82, a network failure presence / absence column 84, a free bandwidth column 86, a delay time column 88, and a load factor column 90 are formed.

  In the route column 80, each route in the network network to which the route belongs is stored in advance. As an example, in the network network MPLS_A shown in FIG. 1, as a path through which the flow addressed to the user terminal CLN can flow from the network network MPLS_B, (1) a route that flows in the order of the router ASR_A4 and the router ASR_A1, A plurality of routes such as a route that flows in the order of the router ASR_A2 and the router ASR_A1 can be formed. Such a route is created in advance according to the configuration of the network.

  A bandwidth value is stored in the bandwidth column 82 in association with each route in the corresponding route column 80. This bandwidth is a design value (maximum value) of each route, and is calculated in advance for each route based on the network configuration information 12a.

  The network failure presence / absence column 84 stores data indicating whether or not the occurrence of a network failure has been detected in the corresponding route by the network failure detection unit 18. As an example, “present” is stored in a period in which the occurrence of a network failure is detected on the corresponding route, and “none” is stored in a period in which the network failure has not occurred or has been recovered.

  The free bandwidth column 86 stores a remaining bandwidth (margin bandwidth) obtained by subtracting the total amount of the flow bandwidth allocated to the corresponding route from the bandwidth stored in the corresponding bandwidth column 82. The value stored in the free bandwidth field 86 decreases when a new flow is assigned, and increases when the flow is released.

  The delay time column 88 stores a delay time corresponding to the flow assigned to the corresponding route. Specifically, in the delay time column 88, a value obtained by adding a dynamic delay time corresponding to the total amount of flows allocated to the corresponding route to a static delay time generated based on the network configuration in advance. Stored.

  In the load factor column 90, a value obtained by subtracting the value stored in the corresponding free bandwidth column 86 from the value stored in the corresponding bandwidth column 82, that is, the bandwidth used by the currently allocated flow is displayed. The ratio (percentage) divided by the value stored in the bandwidth column 82 is stored.

FIG. 5 is a schematic configuration diagram of the router ASR.
Referring to FIG. 5, router ASR includes reception buffer unit 42, route selection unit 44, transmission buffer unit 46, network failure determination unit 52, control unit 40, and interface (I / F) unit 50. Consists of.

  The reception buffer unit 42 receives data packets from one or more connected network nodes and passes them to the route selection unit 44. The reception buffer unit 42 temporarily stores (buffers) received data packets so as not to exceed the processing amount in the route selection unit 44 and adjusts the amount of data packets passed to the route selection unit 44.

  The route selection unit 44 includes a transfer rule storage unit 48, and selectively outputs a data packet passed from the reception buffer unit 42 to a corresponding port of the transmission buffer unit 46 in accordance with the transfer rule stored in the transfer rule storage unit 48. To do. That is, the route selection unit 44 transmits the data to the port of the transmission buffer unit 46 corresponding to the transfer destination of each data packet according to the destination information added to the received data packet.

  The transmission buffer unit 46 transmits the data packet to the corresponding network node according to the port to which the data packet is given from the route selection unit 44. The transmission buffer unit 46 temporarily stores (buffers) the data packet output from the route selection unit 44 so as not to exceed the reception amount at the destination node, and adjusts the data packet amount to be transmitted. .

  The network failure determination unit 52 determines whether a failure has occurred in the transmission path related to the transmission based on the success or failure of the transmission of the data packet to the transfer destination. Specifically, the network failure determination unit 52 monitors the transmission protocol in the transmission buffer unit 46, and detects a transmission failure of the data packet due to occurrence of an ACK (Acknowledgment) reception time over or the like. When a data packet transmission failure is detected, the network failure determination unit 52 identifies the transmission path and notifies the control unit 40 of it.

  The interface unit 50 is connected to the corresponding policy server PS and outputs a transfer rule given from the connection-destination policy server PS to the control unit 40, and the network failure occurrence information given from the control unit 40 is connected to the policy-destination policy. Output to server PS.

  When receiving the transfer rule via the interface unit 50, the control unit 40 writes the transfer rule into the transfer rule storage unit 48 of the route selection unit 44. In addition, when receiving a network failure occurrence notification from the network failure determination unit 52, the control unit 40 notifies the corresponding policy server PS of the occurrence information of the network failure via the interface unit 50.

FIG. 6 is a schematic configuration diagram of the content request receiving server CDS.
Referring to FIG. 6, content request receiving server CDS includes communication units 62 and 64, control unit 60, list display screen data storage unit 68, and content list storage unit 66.

  The communication unit 62 performs data communication with the user terminal CLN and the content server CS via the router ASR of the corresponding network.

  The communication unit 64 performs data communication with the policy server PS via the control network CNW.

  The content list storage unit 66 stores information such as content titles stored in the corresponding content server CS.

  The list display screen data storage unit 68 stores data for displaying the list stored in the content list storage unit 66 for the user terminal CLN. As an example, the list display screen data storage unit 68 stores electronic data in html format that is dynamically or statically generated so that a Web page can be provided to the user terminal CLN as described above.

  Upon receiving a screen display request from the user terminal CLN via the communication unit 62, the control unit 60 selects a content selection screen for the user terminal CLN based on the electronic data stored in the list display screen data storage unit 68. I will provide a. Upon receiving a content distribution start request from the user terminal CLN based on the content selection screen, the control unit 60 transfers the requested content to the policy server PS of the network network MPLS to which the content server CS that can distribute the requested content belongs. Notify the formation of a flow for distributing the content. That is, the control unit 60 notifies the formation of a delivery route from the content server CS storing the requested content to the user terminal CLN in order to realize the delivery of the content in response to the content delivery start request from the user terminal CLN. To do.

(Content distribution flow formation)
Hereinafter, detailed operations relating to the formation of the content distribution flow will be described.

  Referring to FIG. 1 again, as an example, a case where content is distributed from the content server CS_B of the network network MPLS_B to the user terminal CLN of the network network MPLS_A will be described in detail.

FIG. 7 is a sequence diagram relating to formation of a content distribution flow.
Referring to FIG. 7, when a content distribution start request is given from user terminal CLN to content request reception server CDS_A belonging to the same network MPLS_A (step ST10), content request reception server CDS_A is requested. The content is searched (step ST12). Here, since the content request receiving server CDS_A stores only the list of contents stored in the content server CS_A belonging to the same network, the content stored in the other content server CS cannot be searched. Therefore, if the requested content does not exist in the search range, the content request reception server CDS_A gives a content information search request to the content request reception server CDS_B belonging to another network network (step ST14).

  When the content request reception server CDS_B finds that the content requested by the content information search request is stored in the content server CS_B, the required bandwidth and route extraction request necessary for the distribution of the content are the same as the content server CS_B. To the policy server PS_B belonging to the network network MPLS_B (step ST16). That is, the content request reception server CDS_B notifies the policy server PS_B of the formation of a flow for distributing content.

  The policy server PS_B extracts a route candidate that can secure the given required bandwidth in the network MPLS_B (step ST18). Then, the policy server PS_B sets the optimum route candidate extracted from step ST18 as the determined route of the network MPLS_B. Here, the policy server PS_B sets the one with the smallest delay time and load factor among the extracted route candidates as the determined route. That is, the policy server PS_B sets the route corresponding to the smallest one of the values stored in the delay time column 88 and the load factor column 90 in the network operation management information 12b shown in FIG.

  Furthermore, the policy server PS_B gives the set determined route, the required bandwidth necessary for distribution of the content, and the route extraction request to the policy server PS_A belonging to the same network MPLS_A as the user terminal CLN (step ST20).

  The policy server PS_A extracts route candidates that can secure the given required bandwidth in the network MPLS_A (step ST22). Then, the policy server PS_A determines whether or not there is any one that can maintain connectivity between the route candidates extracted in step ST22 and the determined route of the network MPLS_B set by the policy server PS_B. To do. If there is a route candidate that maintains the connectivity, the policy server PS_A sets the optimum route candidate among the route candidates as the determined route of the network MPLS_A. Here, the policy server PS_A sets a route with a small delay time and load factor as a determined route from among route candidates whose connectivity is maintained. Further, the policy server PS_A gives the determined route to the policy server PS_B when the determined route is present and the determined route can be set (step ST24).

  When the policy server PS_B cannot set the determined path in the policy server PS_A (in the case of step ST26), the policy server PS_B repeatedly executes step ST18 and subsequent steps again. The policy server PS_B removes the route candidate selected in the previous process from the route candidates extracted in step ST18, and sets the next optimal route candidate as the determined route of the network MPLS_B.

  When the determined path can be set in the policy server PS_A, the policy server PS_B sets the transfer rule of the router ASR existing in the network MPLS_B so that the flow is formed along the determined determined path ( Step ST28). Similarly, the policy server PS_A sets a transfer rule for the router ASR existing in the network MPLS_A so that a flow is formed along the set determination path (step ST30). Further, when the transfer rule setting (flow setting) is completed, the policy server PS_A gives a path setting completion notification to the content request receiving server CDS_A (step ST32).

  Here, the policy servers PS_A and PS_B use the value in the load factor column 90 corresponding to the path included in the set determined path in the network operation management information 12b shown in FIG. Update to the load factor calculated from the value obtained by adding the required required bandwidth.

  When the content request reception server CDS_A recognizes that the distribution flow from the content server CS_B to the user terminal CLN has been formed, it gives a content distribution request to the content request reception server CDS_B (step ST34). Further, the content request receiving server CDS_B gives a content distribution request to the content server CS_B in response to the content distribution request (step ST36). Then, content distribution from the content server CS_B to the user terminal CLN is started (step ST38). A content distribution flow is formed according to the sequence as described above.

  Note that, even when the user terminal CLN and the content server CS belong to any network network, when the content is distributed over a plurality of network networks, the content distribution flow is performed in the same sequence as in FIG. It is formed.

  Further, when the user terminal CLN and the content server CS belong to the same network, the content distribution flow can be formed only by the processing in the content request receiving server CDS belonging to the network.

(Fault flow switching)
Hereinafter, a switching operation when a network failure occurs in a transmission path during content distribution will be described.

  FIG. 8 is a diagram for explaining an example in which a network failure has occurred in the intra-network route of the network network MPLS_A.

  Referring to FIG. 8, as an example, when a content distribution flow is flowing through route 1a, if a network failure occurs in the network route between router ASR_A4 and router ASR_A1 in network network MPLS_A, the failure point It is necessary to switch to a transmission path that bypasses the network. Thus, a network failure in the network network MPLS_A can be dealt with by switching a transmission path in the network network MPLS_A. That is, the influence of the network failure can be avoided by switching the content distribution flow to the route 1b that is transmitted in the order of the router ASR_A4, the router ASR_A3, the router ASR_A2, and the router ASR_A1.

  That is, when the policy server PS_A detects a network failure, the policy server PS_A sets a new transfer rule for the routers ASR_A1 to ASR_A4 in the network MPLS_A.

  FIG. 9 is a diagram for describing an example in which a network failure has occurred in an inter-network path between the network network MPLS_A and the network network MPLS_B.

  Referring to FIG. 9, as an example, when a content distribution flow is flowing through path 1, a network failure occurs in an inter-network path between router ASR_A4 of network network MPLS_A and router ASR_B4 of network network MPLS_B. It is necessary to switch to a transmission path that bypasses the failure point. In this way, for a network failure in the route between networks, (1) switch to another route between both networks, or (2) bypass (relay) another network other than both networks By switching to the route to be used, the influence of the network failure can be avoided.

  That is, (1) switching to another inter-network path between the two network networks means switching to the path 2 including the inter-network path between the router ASR_A3 of the network network MPLS_A and the router ASR_B3 of the network network MPLS_B. .

  Also, (2) switching to a route that bypasses (relays) another network other than both network networks means switching to route 3 or route 4 that bypasses the network MPLS_C. More specifically, in the route 3, the router ASR_A4 of the network network MPLS_A and the router ASR_B4 of the network network MPLS_B function as “boundary routers”, whereas in the route 4, the router ASR_A3 of the network network MPLS_A and the network network MPLS_B The router ASR_B3 functions as a “boundary router”. Here, since the router ASR_A4 and the router ASR_B4 function as “boundary routers” in the route 1, the “boundary router” is not changed when switching from the route 1 to the route 3, but from the route 1 to the route 4 At the time of switching, the “border router” is changed.

  As described above, when a network failure occurs in an inter-network path, switching to a plurality of transmission paths is possible. Therefore, an optimal switching destination route is determined according to the following switching logic.

  Of the policy servers PS of each network, the policy server PS that detects the occurrence of a network failure becomes the main policy server, and controls switching of transmission paths in the entire network system. Along with this switching process, the policy server PS (including the main policy server) of each network network extracts route candidates in the network network to which the network server belongs, notifies the extracted route candidates to the main policy server, The transfer rule of each router of the network network to which it belongs is set so that the flow flows along the switching destination route determined by the main policy server.

  First, a description will be given of the logic of each policy server extracting route candidates in the network to which it belongs.

  Based on the failure occurrence information from the router ASR, the route candidates in which the network failure has occurred are excluded from the route candidates that can be formed in the network. Then, route candidates with less free bandwidth than the requested bandwidth of the fault flow are excluded. Furthermore, route candidates whose delay time and load factor exceed predetermined reference delay times and reference load factors, respectively, are excluded. Of the route candidates narrowed down in this way, a route candidate including the same intra-network route as the intra-network route included in the route in which the fault flow is flowing is given priority first. Then, a route candidate with a small delay time is given priority next, and a route candidate with a small load factor is given further priority. In this state, priorities are assigned to network network route candidates.

  In the above description, the priority is determined in the order of delay time and load factor, but this order may be changed.

Next, the logic by which the main policy server determines the switching destination route will be described.
Transmission path candidates capable of forming a flow from the transmission source (content server CS) to the destination (user terminal CLN) are created by combining the path candidates in each network. In creating a transmission path candidate, a path including a path where a network failure has occurred and a path that cannot be physically connected are excluded. Furthermore, those that include route candidates with less free bandwidth between networks than the requested bandwidth of the fault flow are also excluded. That is, only transmission paths that can secure a band necessary for switching the fault flow are extracted.

  Then, based on the delay time and the load factor of the route candidate included in each transmission route, the total delay time and the maximum load factor of each transmission route are calculated. Specifically, the transmission path with the smallest total delay time is selected with the highest priority. When there are a plurality of transmission paths having the same total delay time, the one with the smallest maximum load factor is selected with higher priority. Further, when there are a plurality of devices having the same maximum load factor, a transmission route with a higher priority is selected according to the following network route change type.

FIG. 10 is a diagram showing the route change type between networks.
With reference to FIGS. 8 to 10, the route change between networks when a network failure occurs can be classified into five types. Therefore, priorities can be determined in advance between different types.

  The type with the highest priority is the case where there is no change in the route between networks and only the route switching within the network is performed. Specifically, as shown in FIG. 8, the route is switched from the route 1a to the route 1b.

  The second type with the highest priority is when the network is not detoured (relayed), and the same router as before switching functions as a border router (ASBR). For example, as shown in FIG. 9, the route switching from the route 3 to the route 1 is performed.

  The third type with the highest priority is when the network network is not detoured (relayed), and a router different from that before switching functions as a border router. For example, as shown in FIG. 9, the route switching from route 1 to route 2 is performed.

  The fourth type with the highest priority is when the network network is detoured (relayed), and the same router as before switching functions as a border router (ASBR). For example, as shown in FIG. 9, the route switching from the route 1 to the route 3 is performed.

  The type with the lowest priority is a case where a network network is detoured (relayed) and a router different from that before switching functions as a border router. For example, as shown in FIG. 9, the route is switched from route 1 to route 4.

  In this way, among the plurality of inter-network paths connecting the network network MPLS_B of the transmission source (content server CS) and the network network MPLS_A of the destination (user terminal CLN), the intra-network path included in the path through which the failure flow passes A route candidate including the same intra-network route is preferentially determined as a switching destination route.

  In the above description, the priority is determined in the order of the total delay time, the maximum load factor, and the route change type between networks, but this order may be changed.

(Overall sequence)
FIG. 11 is a sequence diagram (part 1) relating to the switching of the fault flow. FIGS. 11 to 14 show, as an example, sequence diagrams when the policy server PS_A becomes the main policy server when a network failure occurs in the path 1 shown in FIG.

  Further, in the following sequence diagrams, in order to indicate the constituent elements of the policy server PS_A, the policy server PS_B, and the policy server PS_C, respectively, the network failure detection unit 18, the path control unit 10, the router setting unit 20, and the database 12 are shown. On the other hand, symbols “A”, “B”, and “C” indicating each policy server are added.

  Referring to FIG. 11, when router ASR in network network MPLS_A determines the occurrence of a network failure and gives a network failure occurrence notification to policy server PS_A (step ST50), network failure detection unit 18A of policy server PS_A A failure location notification for specifying which route the network failure has occurred is given to the route control unit 10A (step ST52). Based on the failure occurrence information from the router ASR, the route control unit 10A refers to the database 12A and extracts a failure flow including a route in which a network failure has occurred (step ST54). Then, the route control unit 10A extracts all alternative route candidates that can substitute for the route through which the extracted fault flow passes (step ST56). Subsequently, the path control unit 10A confirms the status (the presence / absence of a network failure, free bandwidth, delay time, load factor) of each extracted path candidate (step ST58). Furthermore, the route control unit 10A determines the priority of the extracted route candidate according to the logic described above (step ST60).

  The path control unit 10A determines the influence range of the network failure that has occurred (step ST62). That is, the route control unit 10A determines whether it is sufficient to change the route in the network network to which the route control unit 10A belongs or whether it is necessary to change the route between the networks with other network networks. judge.

  When it is determined that the change of the route in the network to which the route belongs is sufficient, the route control unit 10A gives a failure flow deletion request to the router setting unit 20A (step ST64). That is, the path control unit 10A gives a command for deleting the transfer rule stored in the router ASR existing in the transmission path of the fault flow. In response to the fault flow deletion request, the router setting unit 20A deletes and sets the transfer rule related to the fault flow for the target router ASR in the network MPLS_A (step ST66). When the transfer rule deletion setting is completed, the router setting unit 20A gives a failure flow deletion completion notification to the route control unit 10A (step ST68).

  In response to the failure flow deletion completion notification, the route control unit 10A gives a failure flow switching request to the router setting unit 20A (step ST70). That is, the path control unit 10A gives a command for setting a transfer rule of the router ASR existing in the failure flow switching destination path. In response to the fault flow switching request, the router setting unit 20 sets a new transfer rule related to fault flow switching for the target router ASR in the network MPLS_A (step ST72). When the setting of the new transfer rule is completed, the router setting unit 20A gives a failure flow switching completion notification to the route control unit 10A (step ST74). In this way, the transmission path of the fault flow is switched.

  On the other hand, when it is determined that it is necessary to change the route between the networks with other networks, the following sequence is executed.

FIG. 12 is a sequence diagram (part 2) relating to the switching of the fault flow.
Referring to FIG. 12, path control unit 10A extracts an alternative path candidate for another border router in its network network MPLS_A (step ST78). That is, the route control unit 10A extracts all alternative route candidates that can be alternatives to the route through which the extracted failure flow passes for each of the other border routers connected to the inter-network route in the network MPLS_A. Then, the path control unit 10A confirms the status (presence / absence of network failure, free bandwidth, delay time, load factor) of each path candidate extracted for each border router (step ST80). Furthermore, the route control unit 10A determines the priority of the extracted route candidate according to the logic described above (step ST82). Then, the route control unit 10A extracts a preset number of route candidates from the ones with high priority for each border router (step ST84).

  Furthermore, the route control unit 10A gives an inter-network failure flow extraction request to the route control unit 10B of the policy server PS_B (step ST86). In response to the inter-network fault flow extraction request, the path control unit 10B extracts the path in the network network MPLS_B of the fault flow notified from the path control unit 10A (step ST88). Then, the route control unit 10B notifies the route control unit 10A of the extracted route as an inter-network failure flow (step ST90).

  In response to the inter-network route flow notification, the route control unit 10A gives an inter-network route candidate extraction request to the route control unit 10B (step ST92). Then, the route control unit 10B extracts all alternative route candidates that can substitute for the route through which the extracted inter-network failure flow passes (step ST94). Subsequently, the path control unit 10B checks the status (presence / absence of network failure, free bandwidth, delay time, load factor) of each extracted path candidate (step ST96). Furthermore, the route control unit 10B determines the priority of the extracted route candidate according to the logic described above (step ST98). Then, the route control unit 10B extracts a preset number of route candidates from the highest priority for each border router (step ST100), and notifies the route control unit 10A of the extracted route candidates (step ST102). ). This notification includes information such as the delay time and load factor for the extracted route candidate.

  Similarly, the path control unit 10A gives an inter-network fault flow extraction request to the path control unit 10C of the policy server PS_C (step ST104). The route control unit 10C and the route control unit 10A perform the same processing (step ST106 to step ST120) as the above-described step ST88 to step ST102.

  Through the processing as described above, the policy server PS_A, which is the main policy server, collects information related to route candidates in each network. Then, an optimum switching destination route is determined from the route candidates collected in this way according to the following sequence.

FIG. 13 is a sequence diagram (No. 3) relating to the switching of the fault flow.
Referring to FIG. 13, the route control unit 10A searches for a transmission route that can be set as a switching destination from the transmission source (content server CS_B) to the destination (user terminal CLN) by combining route candidates in each network network. (Step ST122). Then, the path control unit 10A extracts the inter-network delay time and the inter-network load factor of the inter-network path from the network network MPLS_A to the network network MPLS_B and the inter-network path from the network network MPLS_A to the network network MPLS_C (step ST124). ).

  Further, the path control unit 10A acquires the inter-network delay time and the inter-network load factor of the inter-network path from the network network MPLS_B to the network network MPLS_A and the inter-network path from the network network MPLS_B to the network network MPLS_C. A network delay time / internet load factor extraction request is sent to the path control unit 10B (step ST126). The path control unit 10B responds to the inter-network delay time / internet load factor extraction request, the inter-network path from the network network MPLS_B to the network network MPLS_A, and the inter-network path network from the network network MPLS_B to the network network MPLS_C. The network delay time and the network load factor are extracted (step ST128), and the extracted network delay time and the network load factor are notified to the path control unit 10A (step ST130).

  Similarly, the path control unit 10A acquires the inter-network delay time and the inter-network load factor of the inter-network path from the network network MPLS_C to the network network MPLS_A and the inter-network path from the network network MPLS_C to the network network MPLS_B. The network delay time / internet load factor extraction request is sent to the path control unit 10C (step ST132). The path control unit 10C responds to the network delay time / internet load factor extraction request, the inter-network path from the network network MPLS_C to the network network MPLS_A, and the inter-network path network from the network network MPLS_C to the network network MPLS_B. The inter-network delay time and the load factor are extracted (step ST134), and the extracted inter-network delay time and the inter-network load factor are notified to the path control unit 10A (step ST136).

  Then, the path control unit 10A calculates the total delay time and the maximum load factor of each settable transmission path searched in step ST122, and determines a switching destination path (step ST138). That is, as described above, for each transmission path, the priority is evaluated in the order of total delay time, maximum load factor, and inter-network path change type, and the transmission path with the highest priority is determined as the switching destination path. . Then, the path control unit 10A executes the following sequence in order to switch to the switching destination path for which the failure flow has been determined.

  FIG. 14 is a sequence diagram (part 4) relating to the switching of the fault flow. FIG. 14 shows a case where the route 4 shown in FIG. 9 is determined as the switching destination route as an example.

  Referring to FIG. 14, path control unit 10A provides a flow setting request to path control unit 10C so that the fault flow flows through the switching destination path (step ST140). This flow setting request includes information related to the route in the network MPLS_C among the switching destination routes determined by the route control unit 10A.

  In response to the flow setting request, the path control unit 10C gives a fault flow switching request to the router setting unit 20C (step ST142). That is, the path control unit 10C gives a command for setting a transfer rule of the router ASR existing in the failure flow switching destination path. In response to the fault flow switching request, the router setting unit 20C sets a new transfer rule related to fault flow switching for the target router ASR in the network MPLS_C (step ST144). When the setting of the new transfer rule is completed, the router setting unit 20C gives a failure flow switching completion notification to the route control unit 10C (step ST146). Furthermore, in response to the failure flow switching completion notification, the path control unit 10C gives a switching completion notification to the path control unit 10A (step ST148).

  Since the failure flow before switching did not flow through the network network MPLS_C, the deletion setting for the transfer rule before switching to the router ASR of the network network MPLS_C is not performed.

  Similarly, the route control unit 10A gives a flow setting request to the route control unit 10B so that the failure flow flows through the switching destination route (step ST150). This flow setting request includes information related to the route in the network MPLS_B among the switching destination routes determined by the route control unit 10A.

  In response to the flow setting request, the path control unit 10B gives a fault flow deletion request to the router setting unit 20B (step ST152). That is, the path control unit 10B gives a command for deleting the transfer rule stored in the router ASR corresponding to the fault flow before switching. In response to the failure flow deletion request, the router setting unit 20B deletes and sets the transfer rule related to the failure flow before switching the target router ASR in the network MPLS_B (step ST154). When the transfer rule deletion setting is completed, the router setting unit 20B gives a failure flow deletion completion notification to the route control unit 10B (step ST156).

  Then, the path control unit 10B and the router setting unit 20B set a new transfer rule related to the switching of the fault flow for the target router ASR in the network MPLS_B, similarly to the above-described steps ST142 to ST148. (Steps ST158 to ST164).

  Furthermore, the route control unit 10A and the router setting unit 20A set a switching destination route in the network network MPLS_A to which the route control unit 10A and the router setting unit 20A belong. That is, the path control unit 10A and the router setting unit 20A set a new transfer rule related to the switching of the fault flow for the target router ASR in the network MPLS_A, similarly to the above-described steps ST152 to ST162. (Steps ST166 to ST176).

  Finally, the path control units 10A, 10B, and 10C store the history related to the path change of the fault flow (steps ST178, ST180, and ST182), and the fault flow switching process ends.

As described above, the failure flow is switched.
In the embodiment of the present invention, the route control unit 10 performs “request command receiving means”, “route candidate extracting means”, “destination route candidate acquiring means”, “route determining means”, “transmission route notifying means” and “transmission route notifying means”. "Distribution flow notification means" is realized, the network failure detection unit 18 realizes "failure occurrence notification means", and the router setting unit 20 realizes "setting means".

  According to the embodiment of the present invention, policy servers arranged in each of a plurality of network networks connected to each other are configured to be able to communicate with each other. When a request command is given to one of the policy servers, the policy server extracts a route candidate capable of ensuring the required quality in its own network, and from the other policy server to the other policy server. The route candidate that can ensure the required quality in the network corresponding to is acquired. Then, the policy server determines a transmission path that can ensure the required quality between a transmission source existing in its own network and a destination existing in another network. Thereby, it is possible to reliably form a flow that spans a plurality of network networks.

  In the embodiment of the present invention, a network system in which three network networks are connected to each other is illustrated. However, the present invention also applies to a network system in which two or more network networks are connected to each other. Applicable to.

  Further, in the embodiment of the present invention, the configuration in which the policy servers PS_A, PS_B, PS_C and the content servers CS_A, CS_B are connected to each other by a dedicated control network has been described. It may be connected.

  The embodiment disclosed this time should be considered as illustrative in all points and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

1 is a schematic configuration diagram of a network system according to an embodiment of the present invention. It is a schematic block diagram of a policy server. It is a figure which shows the data structure stored in a database. It is a figure which shows an example of the content of network operation management information. It is a schematic block diagram of a router. It is a schematic block diagram of a content request reception server. It is a sequence diagram concerning formation of a content distribution flow. It is a figure for demonstrating the example in which the network failure generate | occur | produced in the network path | route of a network network. It is a figure for demonstrating the example in which the network failure generate | occur | produced in the path | route between networks between network networks. It is a figure which shows the route change classification between networks. It is the sequence diagram (the 1) which concerns on switching of a failure flow. It is the sequence diagram (the 2) which concerns on switching of a failure flow. FIG. 10 is a sequence diagram (No. 3) relating to switching of a fault flow. FIG. 10 is a sequence diagram (No. 4) relating to the switching of the fault flow.

Explanation of symbols

  10 path control unit, 12 database, 12a network configuration information, 12b network operation management information, 14 communication unit, 16 interface (I / F) unit, 18 network failure detection unit, 20 router setting unit, 40 control unit, 42 reception buffer Unit, 44 route selection unit, 46 transmission buffer unit, 48 transfer rule storage unit, 50 interface (I / F) unit, 52 network failure determination unit, 60 control unit, 62, 64 communication unit, 66 content list storage unit, 68 List display screen data storage section, 80 route field, 82 bandwidth field, 84 network failure presence / absence field, 86 bandwidth field, 88 delay time field, 90 load factor field, 100 network system, ASR router (autonomous system router), CDS content request Reception server, CLN user Terminal, CNW control network, CS content server, MPLS network, PS policy server.

Claims (5)

A network system comprising a plurality of network networks connected to each other,
Each of the networks is
According to a predetermined transfer rule, at least one router for sending a received data packet to a transfer destination according to information indicating a destination added to the data packet;
A policy server configured to set the forwarding rule of the router so that a flow of a data packet sent from a specific transmission source to a specific destination flows through a desired transmission path in a corresponding network,
The policy servers are configured to communicate with each other,
Each of the policy servers
Request command receiving means for receiving a request command related to the formation of a flow having a predetermined required quality between a transmission source of the own network and a destination of another network,
Route candidate extraction means for extracting a route candidate capable of ensuring the required quality in its own network;
Destination route candidate acquisition means for acquiring a route candidate capable of ensuring the required quality in a network corresponding to the other policy server from another policy server;
Based on the route candidate extracted by the route candidate extraction unit and the route candidate acquired by the destination route candidate acquisition unit, between the transmission source of its own network and the destination of another network network, A route determination means for determining a transmission route capable of ensuring the required quality;
Setting means for setting the forwarding rule of the router existing in the transmission path of its own network so that a flow is formed along the transmission path determined by the path determination means;
The determined transmission path for setting the transfer rule of the router existing in the transmission path of another network so that a flow is formed along the transmission path determined by the path determination means. A transmission path notification means for notifying other policy server of the information relating to the network system. The required quality includes a required transmission band,
The network system according to claim 1, wherein the route determination unit determines the transmission route based on a delay time of each route candidate and a load factor indicating a ratio of a used band. Each of the routers
A failure determination unit that determines whether a failure has occurred in a transmission path related to the transmission based on success or failure of transmission of the data packet to a transfer destination;
When the occurrence of the failure is determined by the failure determination unit, failure occurrence notification means for notifying the occurrence of the failure to the corresponding policy server,
Upon receiving the notification of the occurrence of the failure from the router, the request command accepting unit accepts a request command for assigning another transmission path to a flow formed including the path where the fault has occurred. The network system according to claim 1 or 2. The network is connected to each other via a plurality of inter-network paths;
The route determination means includes an intra-network route that is the same as an intra-network route included in a route through which a failure flow passes among the plurality of inter-network routes that connect the network network of the transmission source and the other network network. The network system according to claim 3, wherein a route candidate is preferentially determined as the transmission route. The network system includes:
A content server arranged in any of the network networks and configured to be capable of delivering a plurality of contents;
A content request receiving server disposed on the same network as the content server, for receiving a content distribution request from the user terminal to the content server;
Upon receiving the distribution request from the user terminal, the content request receiving server forms a flow for distributing the content to the policy server in the same network network as the content server capable of distributing the requested content. Including a delivery flow notification means for notifying
When the request command accepting unit receives a notification of formation of a flow for delivering the content from the content request accepting server, the request command accepting unit relates to delivery of the content from the content server to the user terminal that has issued the delivery request. The network system according to any one of claims 1 to 4, which is accepted as a request command for ensuring required quality.

Images