CN106803803A - Virtual local area network restoration method, system and device - Google Patents

Abstract

A virtual local area network recovery method, system and device are used as a virtual local area network recovery mechanism in a hybrid software-defined network (Hybrid SDN) architecture. In this virtual local area network recovery mechanism, the centralized management feature of SDN is used to centrally process the events of the link (Link) or switch (Switch), and the centralized management controller calculates each virtual local area network (VLAN) path (Path). ), backup paths (Backup Paths) are calculated in advance for all link and switch failure events and stored in the controller. When a link or switch failure event occurs, the controller can receive the failure event and build a corresponding backup path, eliminating the need to repair the convergence of the virtual local area network (VLAN) path. time to improve the reliability of data transmission.

Description

Restoration method, system and device for virtual local area network

technical field

The invention relates to a virtual local area network recovery method, system and device of a hybrid software-defined network.

Background technique

The setting of the path (Path) in the virtual local area network (VLAN) of the traditional network relies on the calculation of distributed communication protocols, such as the use of Spanning Tree Protocol (Spanning Tree Protocol, STP) or VLAN trunk communication protocol (VLAN Trunking Protocol, VTP), so When a link (Link) or switch (Switch) failure event occurs, the switches need to exchange information with each other to recalculate the path, so that when a link failure (Link Failure) occurs, it takes several seconds Waiting for convergence, which is a big problem for data centers that emphasize high reliability.

Contents of the invention

In multiple implementation examples of the present disclosure, a virtual local area network recovery method is provided, which is applicable to a network in a hybrid software-defined network (Hybrid SDN) architecture. The network includes at least one controller and multiple switches. In this method, a backup path table is pre-calculated based on the network and an event notification message is generated according to the connection failure of the virtual local area network, and a link failure processing process is started based on the backup path table corresponding to the event notification message, and at the same time Carry out an inspection process. If the inspection result obtained by the inspection process indicates that it is a switch failure event, then restore the setting changed through the link failure handling process, and perform a switch failure handling process, so that according to a connection failure corresponding to the virtual local area network in the backup path table path to re-establish the virtual local area network. If the inspection result obtained by the inspection process indicates that it is not the above-mentioned switch failure event, then the link failure processing flow will re-establish the virtual local area network according to a path corresponding to the connection failure of the virtual local area network in the backup path table.

In a plurality of implementation examples of the present disclosure, a controller is provided for performing virtual local area network restoration function in a hybrid software-defined network (Hybrid SDN) architecture network. The controller includes a processor and a memory. The memory is used to store the backup path table, and the processor is configured to receive an event notification message. The processor is configured to start a link failure processing process based on the backup path table, and re-establish the virtual local area network according to a path in the backup path table corresponding to the connection failure of the virtual local area network. The processor is configured to perform a check process, wherein if the check result obtained by the check process is judged to be a switch fault event, then restore the setting changed through the link fault processing process, and perform a switch fault processing process, and The virtual local area network is re-established according to a path corresponding to the connection failure of the virtual local area network in the backup path table. If the inspection result obtained by the inspection process indicates that it is not a switch failure event, the execution of the brief inspection process is stopped.

In multiple implementation examples of the present disclosure, a virtual local area network recovery function system is provided, which is applicable to a network of a hybrid software-defined network (Hybrid SDN) architecture. The system includes at least one controller and multiple switches. The switch is used for generating an event notification message according to a connection failure of the virtual local area network. The controller is configured to simultaneously execute a link fault processing procedure and a checking procedure when receiving an event notification message, wherein the controller executing the link fault processing procedure includes corresponding to the virtual local area network based on a backup path table For a path of connection failure, the virtual local area network is re-established on the switch corresponding to this path. When the controller executes the inspection process, if the obtained inspection result indicates a switch failure event, it will restore the setting changed through the link failure processing flow, and perform a switch failure processing flow, and according to the backup path table A path corresponding to the connection failure of the virtual local area network re-establishes the virtual local area network on the switch of the corresponding path, and if the inspection result obtained by the inspection process indicates that it is not a switch failure event, the execution of the inspection process is stopped.

In order to make the above-mentioned features and advantages of the present invention more comprehensible, the following specific embodiments are described in detail with reference to the accompanying drawings.

Description of drawings

FIG. 1A is a schematic diagram illustrating a hybrid software-defined network (HybridSDN) architecture of one of multiple implementation examples of the present invention.

FIG. 1B is used to illustrate the notification method of a link failure event in the recovery mechanism of the exemplary embodiment of the present disclosure.

FIG. 1C is used to illustrate the notification method of a switch (Switch) failure event in the recovery mechanism of the implementation example of the present disclosure.

FIG. 1D is a schematic flowchart of a virtual local area network recovery method in a hybrid software-defined network (Hybrid SDN) architecture according to one of the exemplary embodiments of the present disclosure.

FIG. 2A is a schematic flowchart of a virtual local area network recovery method in a hybrid software-defined network (Hybrid SDN) architecture according to another exemplary embodiment of the present disclosure.

FIG. 2B is a schematic flowchart of a virtual local area network recovery method in a hybrid software-defined network (Hybrid SDN) architecture according to yet another exemplary embodiment of the present disclosure.

FIG. 3A is a schematic diagram illustrating an example of deployed virtual local area network (VLAN) routes and a pre-calculated backup path table between multiple switches in one of the network architectures of the present disclosure.

FIG. 3B is used to illustrate one of the multiple implementation examples of the present disclosure, an example schematic diagram of switching the route to a pre-calculated backup path when an error event occurs in a virtual local area network (VLAN) route deployed between multiple switches.

FIG. 4A is a schematic diagram of an example of a virtual local area network (VLAN) routing architecture that has been deployed among multiple switches in the network shown in FIGS. 3A-3B .

4B and 4C respectively illustrate that the controller pre-calculates backup paths (Backup Path) for all link (Link) and switch (Switch) failure events.

FIG. 5A is used to illustrate one of the multiple implementation examples of the present disclosure. A link failure event backup path table is used for re-deployment, for example, a schematic diagram of the link failure event backup path table during deployment must be completed.

FIG. 5B is used to illustrate one of the multiple implementation examples of the present disclosure. When a switch failure backup path table is used for re-deployment, for example, a schematic diagram of a switch failure backup path table during deployment must be completed.

FIG. 6 is a schematic diagram illustrating a packet format used as a Simple Network Management Protocol (SNMP) in an embodiment of the present disclosure.

FIG. 7 is a schematic diagram of a controller according to an exemplary embodiment of the disclosure.

【Symbol Description】

100: network

CTR: Controller

H1, H2: Server

ES1～ES7, 110: switches

SDNS1~SDNS2: software-defined switches

210: controller

220: switch

221: Link Down Notification

S110-S122: Steps of the virtual local area network restoration method

S230～S246: Steps of the virtual local area network restoration method

300: network

310, 312, 314, 316, 318, 320: switches

340: Controller

342: backup path table

331, 333, 335: paths

410: Link failure event backup path table

420: Switch failure backup path table

510: Link failure event backup path table

520: Switch failure event backup path table

610: version information (Version)

612: Community Name

614: Protocol Data Unit (Protocol Data Unit, PDU)

621: PDU category (PDU Type)

622: Request ID (Request ID)

623: Error Status

624: Error Index

625: multiple object identifiers (Object Identifiers, OIDs)

631~635: object identification code (OID)

700: Controller

702: Processor

704: memory

detailed description

Among the multiple implementation examples proposed in this disclosure, at least one example proposes to use the technology of software-defined network (Software-defined Network, SDN) to restore a kind of virtual local area network applied in hybrid software-defined network (Hybrid SDN) architecture. mechanism. The virtual local area network recovery mechanism in at least one implementation example utilizes the characteristics of SDN centralized management to centrally process the events of the link (Link) or switch (Switch), and calculate the value of each virtual local area network in the centralized management controller (Controller). When the network (VLAN) path (Path) is used, the backup path (Backup Path) is calculated in advance for all link (Link) and switch (Switch) failure events and stored in the controller. When a link (Link) or switch (Switch) failure event occurs, the controller can deploy the corresponding backup path by receiving the failure event, eliminating the need to repair the convergence required for the virtual local area network (VLAN) path time, improve the reliability of data transmission, and achieve a rapid recovery mechanism in a hybrid software-defined network architecture.

In this disclosure, "software-defined network (SDN)" refers to a network. This architecture modifies the control mode of the traditional network architecture and divides the network into a control plane (Control Plane) and a data plane (Data Plane). The management authority of the network is handed over to the controller (Controller) software of the control layer, and a centralized control method is adopted. A controller (which may be a server or any other device with this function) coupled to the software-defined network provides definition of transmission information to a corresponding switch in the software-defined network. This definition can include priority values, rules specifying traffic flow, and an action (eg, forward or "drop") for data flow transmission. This rule can specify, for example, input ports, virtual local area network (VLAN) tags (Tags), media access control (MAC) addresses and destination addresses, Ethertype, Internet Protocol (IP) source and destination addresses, Internet Protocol (IP) IP), Transmission Control Protocol (TCP) source port (port) and destination port and other information. However, fields (fields) of other packet headers in the transmission information can also be included in the rules, depending on their characteristics. After matching at least one of the rules, the switch in the software-defined network takes an action included in the corresponding information flow definition. One example of software-defined networking includes, but is not limited to, the OpenFlow protocol as described in the "OpenFlow Switch Specification" specified by the Open Networking Foundation (ONF).

In this disclosure, a switch capable of processing software-defined information flows in a software-defined network is called a "software-definable" switch. Such a software-definable switch can include ports (Ports) capable of processing software-defined information flow and traditional packet forwarding (such as Ethernet switching or IP routing). A flow definition usually includes one or more software-definable input ports.

In some embodiments, a software-defined network constructed by a hybrid software-defined network (Hybrid SDN) may be implemented by including multiple physical switches and at least one software-defined switch. The physical switch can be an Ethernet switch (Ethernet Switch). In an Ethernet switch, any number of switches coupled in any topology can logically operate as a single switch. And a software-defined switch such as a software-defined networking (SDN) switch.

In at least one example proposed in this disclosure, in the proposed hybrid software-defined network (Hybrid SDN) architecture including virtual local area network recovery mechanism, the controller (which can be a server or any other device with this function) can be implemented through a simple A network management protocol (Simple Network Management Protocol, SNMP) or a command-line interface (Command-Line Interface, CLI) command is used to control a physical switch (such as an Ethernet switch), and an OpenFlow module is used to control a software-defined switch. The controller under the hybrid software-defined network (Hybrid SDN) architecture has path restoration that simultaneously handles link (Link) or switch (Switch) failure events.

The Simple Network Management Protocol (SNMP) proposed in this disclosure refers to a protocol used in network management systems. The network management system includes a group of network management stations (Network Management System) and network elements (Network Element). The network management station can be a server or computer with information processing capability, on which the network management software (SNMPManager) is executed to monitor the managed network elements. The network elements are hardware devices such as hosts, bridges, routers, terminals, servers, etc., on which there is an intermediary software (SNMP Agent) responsible for executing the commands issued by the network management station, while SNMP is management software and intermediary software. A communication protocol used to exchange network management messages. The intermediary software (SNMP Agent) will automatically issue an event notification (EventReport) under certain special circumstances, such as using SNMP Trap to inform the network management software that certain conditions have occurred. The command-line interface (CLI) proposed in the present disclosure is an interface that generally uses text to directly issue commands between communication devices.

A virtual local area network recovery mechanism in a hybrid software-defined network (HybridSDN) architecture of the present disclosure will be illustrated below with an example, but not limited thereto.

Please refer to FIGS. 1A to 1D, wherein FIG. 1A is a schematic diagram of a hybrid software-defined network (Hybrid SDN) architecture illustrating one of the multiple implementation examples of the present invention; FIG. 1B is used to illustrate the restoration mechanism of the implementation example of the present disclosure. The notification method of the Road (Link) failure event; FIG. 1C is used to illustrate the notification method of the switch (Switch) failure event in the restoration mechanism of the implementation example of the present disclosure; and FIG. 1D is one of a plurality of exemplary embodiments according to the disclosure A schematic flow chart of a virtual local area network recovery method in a hybrid software-defined network (HybridSDN) architecture is shown.

The network 100 under the hybrid software-defined network (Hybrid SDN) architecture includes a controller CTR, a first server (Server) H1 and a second server H2, a plurality of physical switches and a plurality of software-defined switches, such as the illustrated Ethernet The first switch ES1, the second switch ES2, the third switch ES3, the fourth switch ES4, the fifth switch ES5, the sixth switch ES6 and the seventh switch ES7 of the network switch (Ethernet Switch), and the first software-defined switch SDNS1 With a second software-defined switch SDNS2. For the convenience of description, in this exemplary embodiment, a controller, two servers, seven physical switches and two software-defined switches are configured to illustrate a hybrid software-defined network (Hybrid SDN). However, the present disclosure is not limited thereto. In another exemplary embodiment, more servers and switches may be configured in the network 100 .

In this implementation example, the controller CTR, the first to the second servers (Server) H1 and H2, the first to the seventh switches ES1-ES7, and the first to the second software-defined switches SDNS1-SDNS2 are connected to each other to form a network 100. For example, the network 100 is a layer two network. Here, the controller CTR, the first to the second servers H1 and H2, the first to the seventh switches ES1-ES7, and the first to the second software-defined switches SDNS1-SDNS2 can also be regarded as nodes in the network 100 ( nodes).

The controller CTR is used to manage all physical machines (Physical Machines), virtual machines (Virtual Machines) and switches connected in the network 100 . For example, the controller CTR is a server, and stores related management information (Management Information), wherein the management information includes information about virtual machines running on the physical machines and information about switches connected to the physical machines. In at least one example proposed in this disclosure, the controller CTR can control the physical switch through the instructions of the Simple Network Management Protocol (SNMP) or the Command Line Interface (CLI), and the controller CTR can control the Ethernet Switch (Ethernet Switch) ) of the first to seventh switches ES1-ES7. The controller CTR can use the OpenFlow module to control software-defined switches, such as the first to second software-defined switches SDNS1 - SDNS2 .

The first to second servers H1 and H2 in the network 100 or other servers not shown but built under the network 100 are all physical hosts. The first to the second servers H1 and H2 or other servers can respectively run one or more virtual machines to provide different services. For example, the first to the second servers H1 and H2 may be equipped with a virtual bridge, and the virtual bridge can enable/disable STP function, setting (configure) related options of spanning tree protocol (option), set firewall rules (firewall rules), and implant (populate) forwarding table (forwarding table).

The first to seventh switches ES1-ES7 and the first to second software-defined switches SDNS1-SDNS2, or other switches not shown but built under the network 100 are deployed on the controller CTR, the first to the second servers H1 and H2 and other servers not shown but built under the network 100 are used for forwarding data packets. For example, the first to seventh switches ES1-ES7 and the first to second software-defined switches SDNS1-SDNS2 are layer two switches (layer two switches), and can enable/disable Spanning Tree Protocol (STP) Functions, setting related options of spanning tree protocol, allowing/blocking broadcast (broadcast), multicast (multicast) and unknown unicast (unicast) data packets, implanting transfer table and can be accessed via command line interface (CLI) or Simple network management protocol interface (SNMP interface) for remote (remote) setting.

In this exemplary embodiment, the controller CTR or another routing element is configured to calculate routing paths (also referred to as “predetermined routing paths” herein) of the network 100 . For example, the predetermined routing path is calculated according to a routing algorithm to more efficiently utilize all the bandwidth of the network 100 . The calculated routing path is transmitted by the controller CTR to each switch, such as a physical switch (such as an Ethernet switch), and the switch can be set through a communication protocol such as SNMP. And if it is a software-defined switch, the switch can be configured through, for example, the OpenFlow communication protocol. The controller CTR may use, for example, Dijkstra's Algorithm to calculate the shortest or best paths from this node to all other nodes starting from a certain node. But not as a limitation.

In the recovery mechanism in multiple exemplary embodiments, the controller CTR can calculate the backup path ( Backup Path) and save to the controller CTR. When a link (Link) or switch (Switch) failure event occurs, the controller CTR can deploy the corresponding backup path by receiving the failure event, eliminating the need to repair the virtual local area network (VLAN) path. Convergence time, improve the reliability of data transmission.

Please refer to FIG. 1B , which is used to illustrate the notification method of a link failure event in the restoration mechanism of the embodiment example of the present disclosure. In this embodiment, for example, when a link (Link) failure event occurs between the fourth switch ES4 and the sixth switch ES6 belonging to the Ethernet switch (EthernetSwitch), the fourth switch ES4 and the sixth switch ES6 will send a link The interruption (also referred to as "interruption") notification message notifies the controller CTR, for example, in one embodiment, the notification method such as SNMP Trap may be used to notify the controller CTR. The controller CTR can quickly adopt the pre-calculated backup path to quickly restore the transmission of the network 100 .

Please refer to FIG. 1C , which is used to illustrate the notification method of a switch (Switch) failure event in the recovery mechanism of the implementation example of the present disclosure. In this embodiment, due to the sixth switch ES6 failure event, it may be because the sixth switch is damaged or the sixth switch is overloaded, which causes the sixth switch ES6 to fail to transfer packets, resulting in the fourth switch ES4 and the seventh switch ES7 , the link (Link) established between the two and the sixth switch ES6 cannot communicate. Therefore, the fourth switch ES4 and the seventh switch ES7 will send a link interruption notification message to notify the controller CTR. In one embodiment, the notification method such as an SNMP abnormal status notification signal SNMP Trap can be used to notify the controller CTR. According to the SNMP Trap notifications from the fourth switch ES4 and the seventh switch ES7, the controller CTR judges that it may be a switch failure event, and can quickly adopt the pre-calculated backup path to quickly restore the transmission of the network 100 .

How does the controller CTR determine that it is a switch (Switch) failure event of the sixth switch ES6? In one implementation example, the protocol (Protocols) that forms a connection between the controller CTR and the sixth switch ES6 can be used to determine the sixth switch ES6 Whether it is interrupted, such as Internet Control Management Protocol (Internet Control Management Protocol, ICMP), OpenFlow protocol, Telnet communication protocol, SSH (Secure Shell) remote login protocol application, Simple Network Management Protocol (SNMP) or Address Resolution Protocol (Address Resolution) Protocol, ARP) and so on. In this embodiment, the Internet Control Management Protocol (ICMP) is used.

The virtual local area network restoration method proposed in the exemplary embodiments of the present disclosure in a hybrid software-defined network (Hybrid SDN) architecture can be divided into a physical switch or a software-defined switch for handling error events. In terms of physical switches (such as Ethernet switches), as long as the SNMP abnormal status notification signal SNMP Trap is set in the physical switch in advance, when a link (Link) failure event occurs, the physical switch will use SNMP Trap to notify the controller. And if a physical switch (Switch) failure event occurs, because the physical switch itself cannot send any notification to the controller, it must rely on the neighbors near the physical switch to tell the controller that a link (Link) failure has occurred, and the controller To determine whether the physical switch is alive, for example, directly use ICMP to determine whether there is a response.

And above-mentioned link (Link) failure event or switch (Switch) failure event, if occur in software-defined switch, when for example the first to second software-defined switch SDNS1～SDNS2 in the figure, then software-defined The switch can, for example, use the OpenFlow module to send a link interruption notification message to notify the controller. When a link failure event occurs, the software-defined switch will directly notify the controller by using, for example, a link failure notification message (Link Failure Notification). As for the failure event of the software-defined switch, the controller and the software-defined switch periodically exchange information to confirm whether the software-defined switch is still operating normally, that is, keep alive information. If a software-defined switch failure event is detected, that is, when the confirmation information for keeping alive cannot be obtained, the controller will automatically identify the failure of the software-defined switch (Switch). The virtual local area network restoration method implemented by the example of the present disclosure will be described below with a practical example.

FIG. 1D is a schematic flowchart of a virtual local area network restoration method in a hybrid software-defined network (Hybrid SDN) architecture according to one of the exemplary embodiments of the present disclosure.

The virtual local area network recovery method of this embodiment is applicable to the network 100 of the hybrid software-defined network (Hybrid SDN) architecture in FIG. 1A , for example. The network 100 includes a controller CTR and a switch 110. The switch 110 can be, for example, one of the physical switches ES1-ES7 in FIG. 1A , or one of the software-defined switches SDNS1-SDNS2 or any switch in the network 100 . In the virtual local area network recovery method implemented in this embodiment, firstly, in step S110, the controller CTR pre-calculates a backup path table based on the network. The switch 110 generates an event notification message 111 and sends it to the controller CTR according to the connection failure of the virtual local area network. Corresponding to the event notification message 111, the controller CTR starts a link fault processing process based on the pre-calculated backup path table, such as step S114, and simultaneously performs a checking process, such as step S116, to check whether it is a switch fault event.

If the inspection result obtained by the inspection process of step S116 indicates that it is a switch failure event, then as in step S118, the setting changed through the link failure processing flow is returned, and a switch failure processing process is carried out as in step S120, so that according to the backup path table A path corresponding to the connection failure of the virtual local area network is re-established in the virtual local area network. In one embodiment, if the link fault handling process in step S114 has not been carried out, and the inspection result obtained in the checking process in step S116 indicates that it is a switch failure event, then the link fault handling process can also be stopped, and there is no need to perform step S118 Reverts the settings changed via the link failure handling process. If the inspection result obtained in the inspection process of step S116 indicates that it is not the above-mentioned switch failure event, the inspection process is stopped, as in step S124.

FIG. 2A is a schematic flowchart of a virtual local area network recovery method in a hybrid software-defined network (Hybrid SDN) architecture according to an exemplary embodiment of the present disclosure. When the hybrid software-defined network (Hybrid SDN) is initially established, the controller 210 calculates the routing (route) path between the nodes in the network, and passes the calculated routing path through, for example, the SNMP communication protocol and/or the OpenFlow communication protocol The settings are made separately for a physical switch (such as an Ethernet switch) or for a software-defined switch. In the virtual local area network restoration method disclosed in this embodiment, the controller 210 pre-calculates the backup path table and stores it (such as step S230). The backup path table may include, for example, a link failure backup path table and a switch failure backup path table. The controller 210 will continuously monitor whether there is a link failure event in the network. If the switch 220 encounters a link failure event, it will send an event notification message to notify the controller 210 . For example, if the switch 220 encounters a link failure event, it will send a link down notification message (Link DownNotification) 221 to notify the controller 210 . . If the link down notification message (Link DownNotification) 221 is sent by a physical switch (such as an Ethernet switch), the notification can be sent through, for example, an SNMP Trap. If it occurs in a software-defined switch, a link interruption notification message can be sent through, for example, an OpenFlow module.

The controller 210, which is constantly listening in the network, can know whether there is a link failure event. After receiving the link interruption notification message 221 (such as step S232), the link interruption notification message 221 will be filtered by a filter (such as step S234). This filter (Filter) can be a software module in the controller 210, and the reason for filtering the link interruption notification message 221 is because if it is the same link failure event or it may be a switch failure event, if it is a switch failure event then The switches at both ends of the link will send a link interruption notification message, which can avoid repeated processing of the same link failure event or misjudgment.

Then, carry out the link interruption processing flow (such as step S236), for example, reconfigure the port of transmission for the virtual local area network (VLAN) in the switch, that is, redistribute the port for the VLAN transmission used by the switch on the path establish. In order to save the processing time of the link failure event, after receiving the notification of the link failure event, immediately deploy the corresponding backup path to the network.

However, this link failure event may also be generated by a switch failure event. Therefore, in addition to the original execution thread (Thread) that handles link failure events, this method will also open another execution thread (Thread). At the same time, confirm whether the switch is alive.

In the step of confirming whether the switch is alive (such as step S238), the controller 210 will confirm whether the switch is alive. For a physical switch (such as an Ethernet switch), because the physical switch itself cannot send any notification to the controller 210, it must rely on the switches near the physical switch to tell the controller 210 that a link (Link) failure has occurred, and the control The controller 210 is used to determine whether the physical switch is alive. For example, the controller 210 sends an Internet Control Management Protocol (ICMP) packet to confirm whether the switches at both ends of the link failure respond to the ICMP packet. The above-mentioned ICMP packets can also be detected by packets of other types of protocols, such as OpenFlow protocol, Telnet communication protocol, SSH (SecureShell) remote login protocol application, Simple Network Management Protocol (SNMP) or Address Resolution Protocol (ARP), etc. Grouping of protocols. For the software-defined switch, the controller 210 periodically exchanges information with the software-defined switch to confirm whether it is still in normal operation, that is, keep alive information. If a software-defined switch failure event is detected, that is, when the confirmation information of keep alive cannot be obtained, the controller 210 will automatically identify the software-defined switch (Switch) failure.

Next, the controller 210 will confirm whether the switch is disconnected (such as step S240), and confirm whether the switch is alive. If the controller 210 finds that it is a switch failure, the system will reply (Roll Back) the change of the link interruption (such as step S242), that is, the change of the setting in the processing flow for the link interruption. A reply will also be given. Because if it is for the failure of the switch, the deployment of the backup path previously performed for the link interruption will generate an error, so recovery needs to be performed. Then perform the process of switch interruption processing (such as step S244), in the case of switch failure, find an appropriate routing path from the pre-calculated switch failure backup path table and deploy the backup path to the network. For example, the virtual local area network (VLAN) in the switch reconfigures the transmission ports so as to troubleshoot the switch, that is, reconfigures the VLAN transmission ports used by other switches on the path. Then recalculate a new backup path (such as step S246).

If the controller 210 finds that it is determined to be a link failure, then the system will then recalculate a new backup path (such as step S246), including the backup paths of link failure and switch failure, and re-update the link failure backup path table respectively And switch failure backup path table.

FIG. 2B is a schematic flowchart of a virtual local area network recovery method in a hybrid software-defined network (Hybrid SDN) architecture according to yet another exemplary embodiment of the present disclosure. Basically, the same or similar steps are performed with the same reference numerals as in FIG. 2A , and will not be repeated here, but the order of execution is different. The main difference is that after the step of confirming whether the switch is alive (such as step S238), if the controller 210 finds that the switch is faulty, the system will roll back (Roll Back) the change of the settings made when the link is interrupted (such as step S242). Then perform the process of switch interruption processing (such as step S244), in the case of switch failure, find an appropriate routing path from the pre-calculated switch failure backup path table and deploy the backup path to the network. Then recalculate a new backup path (such as step S246), and return to step S232 to detect whether the link interruption notification message 221 is received. In the embodiment of FIG. 2B, if the link interruption processing flow (step S236) has not been carried out, and the check result obtained in the step of confirming whether the switch is alive (such as step S238) indicates that the switch is faulty, then the link interruption processing can also be stopped In the process (step S236), there is no need to change the setting of the Roll Back (Roll Back) link interruption (step S242). If the checking result obtained in the step of confirming whether the switch is alive (such as step S238) indicates that it is not the above-mentioned switch failure event, then stop the checking process.

3A-3C will be used in conjunction with actual examples to illustrate the implementation example of the virtual local area network recovery method between the controller and multiple switches in the hybrid software-defined network (Hybrid SDN) architecture of the present disclosure, but it is not intended to be used as an example. limit.

Please refer to FIGS. 3A to 3C. In this embodiment, a network 300 constructed under a hybrid software-defined network (HybridSDN) architecture includes at least a first switch 310, a second switch 312, a third switch 314, a fourth switch 316, The fifth switch 318 , the sixth switch 320 , and the controller 340 . The controller 340 can control the physical switch through commands of Simple Network Management Protocol (SNMP) or Command Line Interface (CLI), and use the OpenFlow module to control the software-definable switch in the physical machine. For example, in this embodiment, the controller 340 can control the first switch 310 , the second switch 312 , the third switch 314 , the fourth switch 316 , the fifth switch 318 , and the sixth switch 320 , but it is not limited thereto.

Please refer to FIG. 3A , which is used to illustrate a schematic diagram of an example of a virtual local area network (VLAN) route that has been deployed among multiple switches under the network 300 architecture of one of the multiple implementation examples of the present disclosure, and a pre-calculated backup path table. . A virtual local area network (VLAN) routing main path 331 is constructed among the first switch 310 , the second switch 312 , the third switch 314 , the fourth switch 316 , the fifth switch 318 and the sixth switch 320 . The primary path 331 of this VLAN is at least to the sixth switch 320 via the first switch 310 , the third switch 314 , the fifth switch 318 , and to the fourth switch 316 via the first switch 310 , the third switch 314 . In the implementation example in this disclosure, the controller 340 pre-calculates the backup path (Backup Path) for all link (Link) and switch (Switch) failure events, completes the pre-calculated backup path table 342 and saves it to the controller 340 storage device or component.

For example, as shown in FIG. 3A , the pre-calculated backup path table 342 includes backup path information of all link failures and all switch failures, and these information can be stored as a link failure backup path table and a switch failure backup path table. In one of the multiple implementation examples shown in the figure, the backup path table 342 includes the backup path 333 corresponding to the failure or damage of the link (Link) 332 and the backup path 335 corresponding to the failure or damage of the fifth switch 318. This is an implementation example only, not a limitation. The backup path table 342 includes the backup path information of all link failures and all switch failures in the network 300 under the hybrid software-defined network (Hybrid SDN) architecture, and after the failure event occurs, after re-establishing the topology in the network 300 , the controller 340 will also recalculate the backup path (Backup Path) for failure events of all links (Link) and switches (Switch) at the same time, and update the backup path table 342 .

When, for example, the link (Link) 332 between the illustrated first switch 310 and the third switch 314 fails or is damaged, the controller 340 is notified through the first switch 310 and/or the third switch 314, and the controller 340 After confirmation, the path 333 is directly deployed, that is, to the sixth switch 320 via the first switch 310 , the second switch 312 , the fourth switch 316 , the third switch 314 , and the fifth switch 318 . In the deployment procedure, the port for the virtual local area network (VLAN) in the switch will be re-deployed, that is, the VLAN transmission port used by the switch on the original main path 331 will be re-deployed, and the path 333 will be used instead. The VLAN transmission ports used in each switch planned in . Therefore, after obtaining the backup path information, a backup path table for link failure or switch failure during deployment will be generated, and based on this backup path table, for example, remove the VLAN transmission port in the switch or add a new switch. Route 333 has been deployed on the VLAN transmission port.

Please refer to FIG. 3B , which is used to illustrate one of the multiple implementation examples of the present disclosure. When an error event occurs in a virtual local area network (VLAN) route that has been deployed between multiple switches, the route is switched to a pre-calculated backup path. Example schematic. If the fifth switch 318 fails or is damaged, the path 331 of the original virtual local area network (VLAN) route will have problems. Therefore, after the controller 340 may be notified via the third switch 314, the sixth switch 320 and/or other switches, the controller 340 confirms that the switch 318 is faulty or damaged, and directly deploys the path 335 according to the pre-calculated backup path table 342, that is, via the first switch 310 , the third switch 314 , the fourth switch 316 to the sixth switch 320 . In the deployment procedure, the VLAN transmission port used by the switch on the original path 331 will be re-deployed, and the VLAN transmission port used by the path 335 will be used instead.

Please refer to FIGS. 4A-4C , which are used to illustrate one of the multiple implementation examples of the present disclosure. An error event occurs in a virtual local area network (VLAN) route that has been deployed between multiple switches and the route is switched to a pre-calculated backup path. , the switches on the path are switched to the corresponding VLAN transmission ports, wherein FIG. 4A is a virtual local area network (VLAN) routing architecture that has been deployed between multiple switches under the network 300 in the implementation example of FIGS. 3A-3B For illustration, FIG. 4B and FIG. 4C illustrate that the controller pre-calculates backup paths (Backup Path) for all link (Link) and switch (Switch) failure events. The same components in FIG. 4A and FIGS. 3A-3B are described with the same reference numerals, and will not be repeated here. In this network 300, a virtual local area network (VLAN) Vlan10 has been deployed with a routing primary path 331, including port 1 and port 3 of the third switch 314, port 1 and port 3 of the fifth switch 318 via the first switch 310. port 2, and port 2 of the sixth switch 320 .

Please refer to FIG. 4A and FIG. 4B first, when the link (Link) 332 between the first switch 310 and the third switch 314 fails or is damaged, that is, for example, the virtual local area network (VLAN) Vlan10 cannot be used. Notify the controller 340 (please refer to FIG. 3B ) through the first switch 310 and/or the third switch 314, and then, the controller 340 confirms that, for example, the virtual local area network Vlan10 is reconfigured to the path 333, that is, via the first switch 310, the second switch The second switch 312 , the fourth switch 316 , the third switch 314 , the fifth switch 318 to the sixth switch 320 . In the deployment procedure, the switch reconfigures the transmission port for Vlan10, that is, the Vlan10 transmission port used by the switch on the original path 331 is reconfigured, and the VLAN used in each switch planned in the path 333 is used instead. transport port.

In one embodiment, please refer to FIG. 4B, according to the pre-calculated link (Link) failure event backup path table 410, the content corresponding to the backup path 333 of the link (Link) 332 of Vlan10 includes (1) from the first Port 2 of a switch 310 is switched to port 2 of the second switch 312; (2) switched from port 1 of the second switch 312 to port 1 of the fourth switch 316; (3) switched from port 2 of the fourth switch 316 to Port 2 of the third switch 314; (4) switch from port 3 of the third switch 314 to port 1 of the fifth switch 318; and (5) switch from port 2 of the fifth switch 318 to port 2 of the sixth switch 320 .

Please refer to FIG. 4A and FIG. 4C , if the fifth switch 318 fails or is damaged, the original primary path 331 of the virtual local area network Vlan10 route will have problems. Therefore, after the controller 340 may be notified via the third switch 314, the sixth switch 320 and/or other switches, the controller 340 confirms that the fifth switch 318 is faulty or damaged, and directly deploys the path 335 according to the pre-calculated backup path table 342, That is, to the sixth switch 320 via the first switch 310 , the third switch 314 , and the fourth switch 316 . In the deployment procedure, the VLAN transmission port used by the switch on the original path 331 will be re-deployed, and the Vlan10 transmission port used by the path 335 will be used instead.

In one embodiment, please refer to FIG. 4C, according to the pre-calculated switch failure backup path table 420, the content in the backup path 335 corresponding to the failure event of the fifth switch 318 of Vlan10 includes (1) from the port of the first switch 310 1 is switched to port 1 of the third switch 314; (2) switched from port 2 of the third switch 314 to port 2 of the fourth switch 316; and (3) switched from port 3 of the fourth switch 316 to the sixth switch 320 port 1.

According to the implementation example of the virtual local area network recovery method between the controller and multiple switches in the hybrid software-defined network (Hybrid SDN) architecture proposed in this disclosure, the characteristics of SDN centralized management are used to pre-calculate each virtual local area network Calculate the backup path (Backup Path) based on the failure events of all links (Link) and switches (Switch) of the network (VLAN). When a link (Link) or switch (Switch) failure event occurs, the controller can deploy a corresponding backup path by receiving the failure event, for example, using the link failure event backup path shown in Figure 4B and Figure 4C The table 410 or the switch failure backup path table 420 is re-deployed to avoid the convergence time required to repair the virtual local area network (VLAN) path, achieve a fast recovery mechanism in the hybrid software-defined network architecture, and improve the reliability of data transmission.

In the virtual local area network recovery method proposed in this disclosure, according to the link failure event backup path table 410 or the switch failure backup path table 420, when the link failure or the switch failure re-deploys the routing path, it must Pay attention to the order of removing or adding VLAN transmission ports, so as to avoid the risk of network loops (loop). Therefore, for re-deployment using the link failure event backup path table 410 in FIG. 4B , for example, the link failure event backup path table 510 (as shown in FIG. 5A ) must be completed during deployment. When using the switch failure backup path table 420 in FIG. 4C for re-deployment, for example, the switch failure backup path table (as shown in FIG. 5B ) must be completed during deployment.

Please refer to FIG. 5A, the link fault event backup path table 510 during deployment, when using the link fault event backup path table 410 in FIG. Except Vlan10; (2) remove Vlan10 from port 1 of the third switch 314; (3) add Vlan10 from port 2 of first switch 310; (4) add Vlan10 from port 2 of second switch 312; (5) add Vlan10 from port 2 of first switch 310; Port 1 of the second switch 312 is added to Vlan10; and (6) Port 1 of the fourth switch 316 is added to Vlan10.

Please refer to FIG. 5B, the switch fault event backup path table 520 during deployment. In this embodiment, the fifth switch 318 is faulty or damaged. When using the switch fault event backup path table 420 in FIG. 4C for re-deployment, it includes ( 1) remove Vlan10 from port 3 of the third switch 314; (2) remove Vlan10 from port 2 of the sixth switch 320; (3) add Vlan10 from port 3 of the fourth switch 316; Port 1 of switch 320 joins Vlan10.

The virtual local area network recovery method of the implementation example proposed in this disclosure is applicable to the architecture of a hybrid software-defined network (Hybrid SDN). The network built under the hybrid software-defined network (Hybrid SDN) architecture can use Simple Network Management Protocol (SNMP) or Command Line Interface (CLI) commands to control physical switches, and use OpenFlow modules to control physical switches. Software-definable switches. For the grouping format of Simple Network Management Protocol (SNMP), please refer to FIG. )614. The protocol data unit (PDU) 614 includes a PDU type (PDU Type) 621, a request ID (Request ID) 622, an error status (Error Status) 623, an error index (Error Index) 624, and a plurality of object identification codes (ObjectIdentifiers, OIDs) 625. The plurality of OIDs 625 may include, for example, individual OIDs 631-635 and the like. The administrator of Simple Network Management Protocol (SNMP) will inform the corresponding switch how to set according to individual object identification codes, such as the object identification codes (OID) 631~635 shown in the figure. Different switches have different setting methods. Instead, the deployment process for restoration of the virtual local area network proposed in the present disclosure can be achieved.

FIG. 7 is a schematic diagram of a controller according to an exemplary embodiment of the disclosure. Please refer to FIG. 7 , the controller 700 includes a processor 702 and a memory 704 . According to the present disclosure, in a hybrid software-defined network (Hybrid SDN) architecture, a virtual local area network recovery method is performed between the controller 700 and multiple switches. The controller 700 utilizes the feature of SDN centralized management to pre-calculate the backup path table 724 for all links (Links) and switches (Switches) of each virtual local area network (VLAN) to calculate the backup path table 724 and store it in the memory 704 . When a failure event of a link (Link) or a switch (Switch) occurs, the controller 700 can deploy a corresponding backup path by receiving the failure event, for example, the link failure event backup path shown in FIG. 4B and 4C is used. The table 410 or the switch failure backup path table 420 is re-deployed to avoid the convergence time required for repairing the virtual local area network (VLAN) path and improve the reliability of data transmission.

The processor 702 is used to control all operations of the controller 700 . For example, the processor 702 is a central processing unit (central processing unit, CPU), but the disclosure is not limited thereto.

The memory 704 is used for storing data. For example, the memory 704 is a static random-access memory (static random-access memory, SRAM), a dynamic random-access memory (dynamic random access memory), a flash memory (flash memory) or other memories or combinations thereof, but The present disclosure is not limited thereto. In particular, the memory 704 stores a plurality of instructions, and the processor 702 executes these instructions to complete the virtual local area network restoration method proposed in this disclosure.

Specifically, the above instructions may include a routing path calculation module 712, a firewall enabling module 714, a spanning tree protocol closing module 716, a forwarding table updating module 718, a firewall clearing module 720, and a node merging/removing module 722 in an implementation example . Here, the processor 702 will execute the routing path calculation module 712 to form a network topology and calculate the routing path according to the nodes in the network; execute the firewall enabling module 714 to enable the firewall of each node to block the routes between nodes; execute generating Tree protocol shutdown module 716 is to close the spanning tree protocol function of each node; Execute transfer table update module 718 to implant the transfer table of each node; Remove module 722 to detect the addition or removal of nodes.

In addition, the above-mentioned instructions may be stored in a computer-readable recording medium. For example, the computer-readable recording medium is a compact disc (CD-ROM), magnetic tape (magnetic tape), floppy disc (floppy disc) or optical data storage device (optical data storage device).

Although the present invention has been disclosed as above with the embodiments, it is not intended to limit the present invention. Those skilled in the art can make some changes and modifications without departing from the spirit and scope of the present invention, so the protection scope of the present invention The scope defined by the appended claims shall prevail.

Claims (32)

1. a kind of Virtual Local Area Network restored method, it is characterised in that fixed suitable for a kind of hybrid software Network in the adopted network architecture, the network at least includes controller and multiple switch, wherein, the method Including：

Backup path table is precalculated based on the network；And

Event notification message is produced according to the connecting fault of the Virtual Local Area Network, it is logical corresponding to the event Know message and be based on the backup path table and start a link failure handling process, and carry out check process simultaneously, Wherein

If the inspection result that the check process is obtained points out to be exchange fault event, reply via the chain The setting of road troubleshooting process change, and exchange fault handling process is carried out, so as to according to the backup In routing table to should the path of connecting fault of Virtual Local Area Network re-establish the VLAN Network,

If the inspection result that the check process is obtained points out the not exchange fault event, by the chain Road troubleshooting process according in the backup path table to should Virtual Local Area Network connecting fault path Re-establish the Virtual Local Area Network.

2. Virtual Local Area Network restored method as claimed in claim 1, if wherein the check process is obtained Inspection result point out to be the exchange fault event, stop the link failure handling process, be not required to reply Via the setting that the link failure handling process is changed.

3. Virtual Local Area Network restored method as claimed in claim 1, wherein the backup for precalculating Routing table includes link failure backup path table and exchange fault backup path table.

4. Virtual Local Area Network restored method as claimed in claim 1, also including based on the network again Set up another backup path table.

5. Virtual Local Area Network restored method as claimed in claim 1, wherein leading to the event is received Knowing before start the link failure handling process after message, further including to filter the event notification message, It is confirmed whether that being directed to the event notification message is processed.

6. Virtual Local Area Network restored method as claimed in claim 1, wherein fixed in the hybrid software These interchangers of adopted network include multiple physical switches, and the wherein event notification message is included by the thing The unusual condition notification signal that reason interchanger sends.

7. Virtual Local Area Network restored method as claimed in claim 6, wherein physical switches send The unusual condition notification signal for simple network management protocol unusual condition notification signal.

8. Virtual Local Area Network restored method as claimed in claim 1, wherein fixed in the hybrid software These interchangers of adopted network include the interchanger of multiple software definables, wherein the event notification message bag Include the linkdown notification message that is sent by the interchanger of the software definable or the controller is soft with these The holding operation information that the interchanger of part definable is exchanged periodically.

9. Virtual Local Area Network restored method as claimed in claim 1, the wherein check process include hair Go out detection signal one by one in the network to should these interchangers of event notification message detect, and According to should detection signal responsive status, whether the inspection result for confirmation is the exchange fault Event.

10. Virtual Local Area Network restored method as claimed in claim 9, the wherein detection signal are to meet The packet of internet control management protocol.

11. Virtual Local Area Network restored methods as claimed in claim 9, the wherein detection signal are to meet OpenFlow agreements, Telnet communication protocols, the application of SSH distal ends login protocols, simple network management association The packet of one of view or address resolution protocol.

12. a kind of controllers, it is characterised in that in a kind of net of the hybrid software definition network architecture In network, Virtual Local Area Network return function is carried out, wherein the controller includes processor and memory, Wherein the memory is used to store backup path table, and the processor is configured to reception event and notifies news Breath, wherein,

The processor is based on the backup path table and starts link failure handling process, and according to the backup path In table to should the path of connecting fault of Virtual Local Area Network re-establish the Virtual Local Area Network；

The processor carries out check process, if wherein the inspection result that the check process is obtained is judged as exchanging Machine event of failure, then reply the setting changed via the link failure handling process, and carry out interchanger event Barrier handling process, and according in the backup path table to should Virtual Local Area Network connecting fault path The Virtual Local Area Network is re-established, if the inspection result that the check process is obtained points out the not exchange Machine event of failure, then stop performing the check process.

13. controllers as claimed in claim 12, if wherein the inspection result that the check process is obtained refers to Go out to be judged as the exchange fault event, stop the link failure handling process, be not required to reply via the chain The setting of road troubleshooting process change.

The backup path table that 14. controllers as claimed in claim 12, the wherein memory are stored It is that the processor is configured to be precalculated under the network according to the hybrid software definition network architecture Produce, and store in the memory.

15. controllers as claimed in claim 13, wherein the backup path table are backed up including link failure Routing table and exchange fault backup path table, reason device is configured to the chain for performing in this place respectively Used in road troubleshooting process and the check process.

16. controllers as claimed in claim 12, wherein and the processor is configured to perform the chain After road troubleshooting process and the check process, be further configured to based on re-establish this is virtual Another backup path table is set up in region, and updates to the memory, is used to replace the backup path table.

17. controllers as claimed in claim 12, the wherein memory be static RAM, Other memories of dynamic random access memory, flash memory or its combination.

18. controllers as claimed in claim 12, the wherein processor are configured to perform the link During troubleshooting process, before the link failure handling process is started, further include to filter the event Notification message, is confirmed whether that being directed to the event notification message is processed.

19. controllers as claimed in claim 12, the wherein event notification message are managed for simple network The unusual condition notification signal of agreement.

20. controllers as claimed in claim 12, the wherein event notification message are the processor through matching somebody with somebody Put the holding operation information periodically to be swapped to the network.

21. controllers as claimed in claim 12, the wherein processor are configured to perform the inspection In flow the check process include sending detection signal one by one in the network to should event notification message These interchangers are detected, and according to should detection signal responsive status, the inspection for confirmation Whether result is the exchange fault event.

22. controllers as claimed in claim 12, the wherein detection signal are to meet internet-based control pipe Manage the packet of agreement, OpenFlow agreements, Telnet communication protocols, the application of SSH distal ends login protocols, The packet of one of simple network management protocol or address resolution protocol.

23. a kind of systems of Virtual Local Area Network return function, it is characterised in that be applicable a hybrid software The network of the network architecture is defined, the wherein system includes at least including controller and multiple switch,

Wherein, the interchanger is used to be produced event to notify news according to the connecting fault of the Virtual Local Area Network Breath,

The controller is used to when event notification message is received, at the same perform link failure handling process with Check process, wherein,

The controller performs the link failure handling process to be included based on backup path table to should virtual local area The path of the connecting fault of network on these interchangers in path to that should re-establish the VLAN Network；And

When the controller performs the check process, if the inspection result for obtaining points out to be an exchange fault thing Part, then reply the setting changed via the link failure handling process, and carry out exchange fault processing stream Journey, and according in the backup path table to should Virtual Local Area Network connecting fault path to should The Virtual Local Area Network is re-established on these interchangers in path, if the inspection that the check process is obtained Result points out the not exchange fault event, then stop performing the check process.

The system of 24. Virtual Local Area Network return functions as claimed in claim 23, if wherein obtaining Inspection result points out to be judged as the exchange fault event, stops the link failure handling process, is not required to back The multiple setting changed via the link failure handling process.

The system of 25. Virtual Local Area Network return functions as claimed in claim 23, wherein the backup road Footpath table is to precalculate generation under the network of the controller according to the hybrid software definition network architecture, And stored.

The system of 26. Virtual Local Area Network return functions as claimed in claim 25, wherein the backup road Footpath table includes link failure backup path table and exchange fault backup path table, respectively in the controller Used in the link failure handling process for performing and the check process.

The system of 27. Virtual Local Area Network return functions as claimed in claim 23, wherein and the control After device performs the link failure handling process and the check process, this for being based further on re-establishing is virtual Another backup path table is set up in region, is used to replace the backup path table.

The system of 28. Virtual Local Area Network return functions as claimed in claim 23, the wherein controller When performing the link failure handling process, before the link failure handling process is started, further include The event notification message is filtered, is confirmed whether that being directed to the event notification message is processed.

The system of 29. Virtual Local Area Network return functions as claimed in claim 23, the wherein event are led to Know the unusual condition notification signal that message is a simple network management protocol.

The system of 30. Virtual Local Area Network return functions as claimed in claim 23, the wherein event are led to Know that message is the holding operation information that the controller is periodically swapped to these interchangers of the network.

The system of 31. Virtual Local Area Network return functions as claimed in claim 23, the wherein controller Perform the check process in the check process include sending detection signal one by one in the network to should event These interchangers of notification message are detected, and according to should detection signal responsive status, so as to Confirm whether the inspection result is the exchange fault event.

The system of 32. Virtual Local Area Network return functions as claimed in claim 23, the wherein detection are believed Number it is to meet packet, OpenFlow agreements, Telnet communication protocols, the SSH of internet control management protocol The packet of one of the application of distal end login protocol, simple network management protocol or address resolution protocol.