WO2011144082A2 - 虚拟网络单元中的路由计算方法和主节点设备 - Google Patents

虚拟网络单元中的路由计算方法和主节点设备 Download PDF

Info

Publication number
WO2011144082A2
WO2011144082A2 PCT/CN2011/074664 CN2011074664W WO2011144082A2 WO 2011144082 A2 WO2011144082 A2 WO 2011144082A2 CN 2011074664 W CN2011074664 W CN 2011074664W WO 2011144082 A2 WO2011144082 A2 WO 2011144082A2
Authority
WO
WIPO (PCT)
Prior art keywords
node
path
topology
master node
primary
Prior art date
Application number
PCT/CN2011/074664
Other languages
English (en)
French (fr)
Other versions
WO2011144082A3 (zh
Inventor
朱明明
李宇华
Original Assignee
华为技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Priority to PCT/CN2011/074664 priority Critical patent/WO2011144082A2/zh
Priority to EP20110782999 priority patent/EP2621130B1/en
Priority to CN2011800006322A priority patent/CN102204188B/zh
Publication of WO2011144082A2 publication Critical patent/WO2011144082A2/zh
Publication of WO2011144082A3 publication Critical patent/WO2011144082A3/zh
Priority to US13/930,856 priority patent/US9077608B2/en

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00Routing or path finding of packets in data switching networks
    • H04L45/02Topology update or discovery
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00Routing or path finding of packets in data switching networks
    • H04L45/58Association of routers
    • H04L45/586Association of routers of virtual routers
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00Routing or path finding of packets in data switching networks
    • H04L45/64Routing or path finding of packets in data switching networks using an overlay routing layer
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00Routing or path finding of packets in data switching networks
    • H04L45/22Alternate routing

Definitions

  • the embodiments of the present invention relate to the field of communications technologies, and in particular, to routing in a virtual network unit.
  • VNE Virtual Network Element
  • a single-image cluster router The control plane and management plane of the VNE are concentrated on one master router, that is, the master node, and the other router acts as the standby node, and the other routers are slave nodes. Traffic forwarding is available for all nodes in the VNE.
  • the primary node and the standby node are usually assumed by a large-capacity, high-performance, high-performance central processing unit (CPU) and a router with a high network location.
  • the master node collects VNE internal extensions. Pu, the corresponding calculation is performed, and the internal forwarding information base (hereinafter referred to as iFIB) is generated, and then the iFIB is sent to other nodes to complete the route calculation in the VNE.
  • the iFIB is queried first, and the VNE outbound interface is obtained, and then forwarded to the corresponding VNE node.
  • the external forwarding table is further queried (external forwarding information Base; Abbreviation: eFIB), obtain the next hop, and then forward the message from the VNE.
  • eFIB internal forwarding information Base
  • VNE shields the internal interconnection details and presents it as a super-node with a single image. From the perspective of the overall topology of the network, it is a routing node that provides a unified eFIB packet. Forward.
  • VNE technology simplifies network topology, improves route convergence and stability, and reduces management complexity. Compared with multi-box routers, there is no need to introduce dedicated interconnect devices and interconnect interface technologies. Protected existing investments and reduced investment and operation costs. And according to the network application, devices of different models and different forms can be aggregated to reduce the maintainability of the device.
  • An existing VNE internal topology calculation scheme is an On Demand Routing (ODR) technology, but the ODR technology is only applicable to a hub-spoke shape topology, and cannot be applied to dual-homing. Other topologies such as trees and rings have a narrow range of applications. Summary of the invention
  • the embodiment of the invention provides a route calculation method and a master node device in a virtual network unit, so as to implement calculation of internal routes of the virtual network unit under various topology structures.
  • An embodiment of the present invention provides a route calculation method in a virtual network unit, including: a primary node in a virtual network unit receives and saves topology information reported by a node other than the primary node in a virtual network unit;
  • the master node performs topology verification on all the topology information locally saved by the master node according to the topology type of the virtual network unit, where the saved topology information includes topology information and a location of the master node.
  • the primary node calculates a weight value of the internal path of the virtual network unit and the internal path according to the topology type and the topology information;
  • the master node generates and saves an internal route entry according to the internal path and the weight value of the internal path, and sends the internal routing entry to a node other than the primary node in the virtual network unit.
  • the embodiment of the present invention further provides a master node device in a virtual network unit, including: a receiving module, configured to receive topology information reported by a node device of the virtual network unit except the master node device;
  • a saving module configured to save topology information received by the receiving module
  • a verification module configured to save the save module according to a topology type of the virtual network unit All the topology information is topologically verified, and all the topology information saved by the saving module includes topology information of the master node device and topology information received by the receiving module;
  • a calculation module configured to calculate, according to the topology type and the topology information saved by the save module, a weight value of the internal path of the virtual network unit and the internal path after the topology verification is passed;
  • a generating module configured to generate and save an internal routing entry according to the internal path and the weight value of the internal path calculated by the computing module;
  • a sending module configured to send an internal routing entry generated by the generating module to a node device other than the primary node device in the virtual network unit.
  • the primary node After the primary node receives and saves the topology information reported by the node in the virtual network unit, the primary node performs topology verification on the saved topology information according to the topology type of the virtual network unit, and performs topology verification.
  • the master node calculates the weight values of the internal path and the internal path of the virtual network unit according to the topology type and the topology information. Then, the master node can generate and save the internal route entry according to the weight values of the internal path and the internal path. And send the internal routing entry to a node other than the primary node in the virtual network unit.
  • FIG. 1 is a flowchart of an embodiment of a route calculation method according to the present invention.
  • FIG. 2 is a schematic diagram of an embodiment of the topology collection of the present invention
  • 3 is a schematic diagram of an embodiment of a dual-homing tree extension 41 of the present invention
  • FIG. 4 is a schematic view of an embodiment of a ring topology of the present invention.
  • Figure 5 is a schematic view of an embodiment of a semi-circular topology of the present invention.
  • FIG. 6 is a schematic diagram of an embodiment of a primary path and a secondary path in a dual-homing tree topology according to the present invention
  • FIG. 7 is a schematic diagram of an embodiment of a primary path and a secondary path in a ring topology of the present invention
  • FIG. 9 is a schematic diagram of an embodiment of a primary path and a secondary path in a semi-circular topology according to the present invention
  • FIG. 10 is a schematic diagram of a primary path and a secondary path in a semi-circular topology according to the present invention
  • FIG. 11 is a schematic structural diagram of an embodiment of a master node device in a virtual network unit according to the present invention
  • FIG. 12 is a schematic structural diagram of another embodiment of a master node device in a virtual network unit according to the present invention.
  • the route calculation method may include:
  • Step 101 The master node in the VNE receives and saves the topology information reported by the node other than the master node in the VNE.
  • the VNE is a virtual heterogeneous cluster router, and the VNE includes a primary node, an alternate primary node, and a secondary node.
  • the master node is a large-capacity high-performance aggregation router at the aggregation layer. It is a node that manages all the nodes in the entire VNE. It is used to centrally process the control signaling of the VNE, and sends the calculated control information to the slave node, the control plane of the VNE.
  • On the primary node all nodes in the above VNE include the primary node, the standby primary node, and all secondary nodes. Inside a VNE only A primary node and an alternate primary node. The standby primary node is faulty on the primary node, and cannot be
  • the master node is taken over to manage all nodes in the VNE.
  • the slave node in the VNE is a small-capacity, low-performance router in the aggregation layer. It does not process control signaling and is only responsible for data forwarding.
  • Step 102 The master node performs topology verification on all the topology information locally saved by the master node according to the topology type of the VNE, and all the topology information locally saved by the master node includes the topology information of the master node and the VNE. Topological information reported by nodes other than the primary node.
  • Step 103 After the topology verification is passed, the primary node calculates the internal path of the VNE and the weight value of the internal path according to the topology type and the topology information.
  • the internal path refers to the path information between any two nodes in the VNE, including the node through which the path between the two nodes passes and the corresponding interface; the internal route refers to the primary node in the VNE according to the internal path. Standard routing information generated by each node.
  • Step 104 The master node generates and saves an internal routing entry according to the weight values of the internal path and the internal path, and sends the internal routing entry to a node other than the primary node in the VNE.
  • the above embodiment can implement the calculation of the internal path of the VNE under various topologies such as dual-tree, ring, and semi-ring, and thus can implement the simple and effective calculation of the primary path and the secondary path in the internal path of the VNE.
  • the topology collection needs to be performed before the VNE is managed and the internal route is calculated.
  • the topology collection may adopt dynamic collection and static configuration. .
  • the master node receives and saves the topology information reported by the node other than the master node in the VNE.
  • the master node can broadcast the topology collection request message of the VNE in a broadcast manner.
  • the basic collection request message carries the information of the primary node, the identifier, the name, and the topology type of the VNE.
  • the primary node may send the topology collection request message of the VNE in a Layer 2 or Layer 3 broadcast manner. Then, the primary node may receive the VNE.
  • topology collection response message sent by a node other than the primary node, where the topology collection response message carries topology information; wherein the topology collection response message is that the node other than the primary node in the VNE receives the topology
  • the request message it is determined that the VNE corresponding to the identifier of the VNE carried in the topology collection request message is sent to the master node by the VNE that belongs to the home node; if the node other than the master node in the VNE receives the topology collection After the request message, it is determined that the VNE corresponding to the identifier of the VNE carried in the topology collection request message is not the VNE to which the VNE belongs, but may not respond to the topology collection request message.
  • the master node After receiving the topology collection response message sent by the node other than the master node in the VNE, the master node may save the topology information carried by the topology collection response message, and include all the nodes in the VNE in the VNE cluster management. And then sending a registration success message to the node other than the master node in the VNE to indicate that the node other than the master node in the VNE is successfully registered in the VNE.
  • the primary node of the VNE After receiving and storing the topology information of a node other than the primary node in the VNE, the primary node of the VNE performs a topology check on all the topology information locally saved by the primary node. If the topology check is passed, the master node can determine that the entire VNE topology information is collected completely, and the VNE topology is established, and the internal path calculation can be started; if the topology check fails, the master node continues to periodically or The topology collection request message is sent in a broadcast manner, and the topology information reported by the node in the VNE is continuously waited until the entire VNE topology information is collected.
  • the master node may immediately perform a topology check on all the topology information locally saved by the master node; When a node other than the primary node goes offline or cannot detect the node, a waiting time can be set. After the waiting time is exceeded, the primary node performs topology verification again. If a node other than the primary node in the VNE frequently goes online or goes offline, causing the topology check to fail, the master node may not calculate the internal interior corresponding to the node that is frequently online or offline. Path Delete, but deactivate the internal path to ensure the normal forwarding of data. After the new topology check is passed and the internal paths of all the nodes in the VNE are calculated, the master node in the VNE generates a new internal route entry and sends it to all nodes in the VNE for routing update.
  • the step 101 may receive, by the primary node, topology information that is reported by the node other than the primary node in the VNE after the location of the primary node is obtained. That is to say, in the static configuration mode, after the node in the VNE learns the location of the primary node, it can actively report the topology information to the primary node.
  • the master node After receiving the topology information, the master node saves the topology information, and all the nodes in the VNE are included in the VNE cluster management.
  • the master node of the VNE After the topology information reported by one of the VNEs is received and saved, the master node of the VNE performs a topology check on all the topology information saved locally. If the topology check passes, the master node can determine that the entire VNE topology information is collected completely, and the VNE topology is established, and the internal path calculation can be started; if the topology check fails, the master node continues to wait for the VNE. The topology information reported by the node until the entire VNE topology information is collected.
  • the topology information in the embodiment of the present invention includes the identifier of the node in the VNE, the basic information of the device where the node is located, the external connection port information, and the internal connection port information.
  • the internal connection port refers to an interface that is connected between nodes in the VNE; the external connection port refers to an interface between a node inside the VNE and a device external to the VNE; that is, the interface of each node in the VNE is divided into an internal connection port and an external interface.
  • the connection port is connected to the node inside the VNE and the device outside the VNE.
  • the external connection port information includes an interface name of the external connection port, an Internet Protocol (IP) address, and an encapsulation information of the external connection port (for example, the external connection port is a Layer 2 interface or a Layer 3 interface);
  • the information of the internal connection port includes the interface name of the internal connection port, the IP address, and the encapsulation information of the internal connection port (for example, the internal connection port is a Layer 2 interface or a Layer 3 interface).
  • the master node may periodically or periodically send the topology message in the VNE, where the topology message carries the topology information of each node in the VNE, so as to receive the topology message.
  • the node After determining that the topology information of the node is different from the topology information carried in the topology message, the node sends its topology information to the master node.
  • the master node After receiving the topology information, the master node performs corresponding processing according to the actual situation, for example, performing path switching or performing new path calculation.
  • the extension 4 packet may be sent in a broadcast manner or in a directional manner.
  • the master node can receive and save the topology information of the newly added node reported by the newly added node in the VNE, and then re-store the saved information according to the topology type of the VNE.
  • the topology information is topologically verified, that is, the newly added node in the VNE actively reports its own topology information to the master node; or, the master node can receive and save the node in the VNE connected to the newly added node.
  • the topology information is verified, and then all the topology information locally saved by the master node is topologically verified according to the topology type of the VNE, where the topology information is that the node connected to the newly added node receives the extension sent by the master node.
  • the packet After the packet is sent, it is determined that the information of the external interface and the information of the internal interface have changed, that is, the topology information of the topology information is different from the topology information carried in the topology message, and then sent to the master node; or
  • the master node can receive and save the topology information reported by the newly added node after receiving the topology packet sent by the master node, and re-based the topology of the VNE. Type on all the topology information stored locally master node topology verification.
  • the master node can receive and save the changed topology information reported by the node directly connected to the removed node in the VNE, and then re-based the topology node according to the topology type. All the topology information saved locally is topologically verified; the changed topology information is automatically reported to the master node by the node directly connected to the removed node, or the changed topology information is directly connected to the removed node.
  • the node After receiving the topology message sent by the master node, the node determines that its topology information is different from the topology information carried in the topology message, and then sends it to the master node.
  • the nodes in the VNE except the master node can detect the faults of the nodes other than the master node in the VNE through a heartbeat mechanism. If the primary node does not receive a response from a node to the heartbeat, it can be determined that the node has failed.
  • the primary node in the topology After the verification is passed, when the control plane of the node other than the primary node in the VNE fails, and the forwarding plane does not fail, after the primary node discovers that the above node fails through the heartbeat mechanism, the primary node can delete the failed node.
  • the topology information is re-received and saved, and the topology information reported by the node other than the primary node and the failed node in the VNE is re-acquired, and all the extension information stored locally by the primary node is extracted according to the topology type of the VNE. check.
  • the primary node can receive and save the changed topology information reported by the failed node, and then re-localize the primary node according to the topology type of the VNE. All the saved topology information is topologically verified.
  • the master node can receive and save the changed topology information reported by the node on the failed link through the alternate path, and receive and save and occur.
  • the changed topology information reported by the node connected to the node on the faulty link and then the master node can perform topology verification on all the topology information locally saved by the master node according to the topology type of the VNE.
  • step 102 the master node performs a topology check on all the topology information locally saved by the master node according to the topology type of the VNE.
  • FIG. 3 is a schematic diagram of an embodiment of a dual-homing tree topology 41.
  • the master node may perform all topology information locally saved according to the master node. Determine whether the verification pass condition of the dual-homing tree topology is satisfied; at this time, the topology verification can be: The verification condition of the double-homing tree topology is satisfied; that is, when the double-homing tree is formed When the verification pass condition of the topology is satisfied, the master node can determine the dual-homing tree topology check.
  • the verification pass condition of the dual-homing tree topology includes: all nodes in the VNE have at least two internal ports available; and the primary node and the standby primary node are directly connected through the internal connection port, and the primary node and the standby primary The internal connection port and the link state between the nodes are normal; and all the slave nodes in the VNE have only two inner connection ports, which are respectively connected to the primary node and the standby primary node; and the slaves of the VNE have reported their own topology information.
  • the number of nodes is the same as the number of pre-planned slave nodes.
  • FIG. 4 is a schematic diagram of an embodiment of a ring topology according to the present invention, when the topology type of the VNE
  • the master node can determine whether the verification pass condition of the ring topology is satisfied according to all the topology information saved locally by the master node; at this time, the topology check can pass the check of the ring topology.
  • the condition is satisfied; that is, when the check pass condition of the ring topology is satisfied, the master node can determine that the ring topology check passes.
  • the verification pass condition of the ring topology includes: each node in the VNE has two inner connection ports with normal status; and the primary node and the standby primary node are directly connected through the internal connection port; and the nodes in the VNE are two or two Connected, starting from an internal connection port of the primary node, each node other than the primary node in the VNE can reach another internal connection port of the primary node; and two rings formed from the two internal connection ports of the primary node On the path, the order of the nodes is reversed.
  • FIG. 5 is a schematic diagram of an embodiment of a semi-ring topology according to the present invention.
  • the master node may determine the ring topology according to all topology information locally saved by the master node. Whether the verification pass condition has been met; at this time, the topology check can pass the verification pass condition that can be a semi-circular topology; that is, when the check pass conditions of the semi-circular topology are satisfied, the main The node can determine the circular topology check pass.
  • the verification pass condition of the semi-ring topology includes: Each node except the primary node and the standby primary node in the VNE has at least two inner connection ports with normal status; and both the primary node and the standby primary node have At least one internal connection port, and there is no directly connected internal link between the primary node and the standby primary node; and, in addition, the node in the VNE except the primary node and the standby primary node is connected to each other, and an internal connection port of the primary node Initially, each node other than the primary node and the standby primary node in the VNE can reach a normal internal connection port of the standby primary node; and a semi-annular path formed from an inner connection port of the primary node passes All nodes in the VNE except the primary node, the last node of the semi-ring path is the alternate primary node.
  • the master node can determine that the entire VNE topology information is collected completely, and the VNE topology is established, and the internal path calculation can be started; if the topology check fails, the master node continues to periodically or periodically.
  • the topology collection request message is sent in a broadcast manner, and the topology information reported by the node other than the master node in the VNE is continuously waited until the entire VNE topology information is collected. Complete.
  • step 103 after the topology check is passed, the master node performs a method of calculating the internal path of the VNE and the weight value of the internal path according to the topology type and the topology information.
  • the internal path of the VNE is centralized and calculated by the primary node.
  • the master node may calculate two paths between every two slave nodes in the VNE, and calculate the slave node to the master node, and the path from the slave node to the standby master node for each slave node in the VNE; Determining that the weight of the path passing through the primary node in the two paths is a weight value of a path between the first slave node and the master node, and a sum of weight values of a path between the master node and the second slave node, such as a formula (1) Determining; determining a weight value of a path passing through the alternate primary node in the two paths is a weight value of a path between the first slave node and the standby master node, and a weight of a path between the standby master node and the second slave node The sum of the values, as shown in equation (2); and determining that the path with the
  • first and second in the embodiments of the present invention are only for convenience of description, and are used to distinguish two different slave nodes, and are not limited to the slave nodes.
  • weight[i][main][j] is the weight value of the path between the i-th slave node and the j-th slave node through the master node, weight t .
  • the main [i] is the weight value of the path between the i-th slave node and the master node
  • weighty [j] is the weight value of the path between the j-th slave node and the master node, that is, the master node and the j-th slave node
  • the weight value of the inter-path in equation (2), weight[i] [ready] [j] is the weight value of the path between the i-th slave node and the j-th slave node through the alternate master node, weight t .
  • Ji] is the weight value of the path between the i-th slave node and the standby master node, weight t .
  • Jj] is the weight value of the path between the jth slave node and the standby master node, that is, the weight value of the path between the standby master node and the jth slave node.
  • the weight value of the path between the primary node and the first slave node or the second slave node may be allocated by the master node according to the number of the first slave node or the second slave node, and the standby master node and the first node
  • the weight value of the path between the slave node or the second slave node may also be assigned by the master node according to the number of the first slave node or the second slave node, for example: for the slave node numbered odd, with the node and the master node
  • the path between the paths is the primary path, and the weight value is assigned to 1; the path between the node and the alternate primary node is the secondary path, and the weight value is assigned as 99, as shown in equations (3) and (4);
  • the path between the node and the master node is the secondary path, and the weight value is assigned to 99;
  • the path between the node and the standby master node is the primary path, and the weight value is assigned to 1, such as 3) and formula (4)
  • FIG. 6 is the main path in the dual-homing tree topology of the present invention.
  • a schematic diagram of an embodiment with a slave path in which the dotted line shows the primary path between the slave node and the root node, and the dotted line shows the secondary path between the slave node and the root node.
  • the root node includes a primary node and an alternate primary node.
  • weight t . [i] is the weight value of the path between the i-th slave node and the master node
  • the primary path and the secondary path calculated by the primary node for each two secondary nodes in the VNE can be as shown in Table 1.
  • nodes 1 point 1 node 3 point - slave node 3 slave node from node 3 node from node 3 node from node 3 node from node 3
  • Section 3 Main Section - Alternate Master - Main Section - Alternate Master - Main Section - Alternate Master - Main Section - Alternate Master
  • the weight value of the path between the primary node and the first slave node or the second slave node may be pre-configured, and the weight value of the path between the standby primary node and the first slave node or the second slave node may also be pre-configured. That is to say, the weight value of the path between the primary node and the first slave node or the second slave node may not be allocated in the manner shown in the formula (3), but an arbitrary value is pre-configured; The weight value of the path between the first slave node or the second slave node may also be allocated in a manner not shown in the equation (4), but an arbitrary value is pre-configured.
  • the direct link between the primary node and the standby primary node is used for active/standby switchover. Generally, it is not used for data traffic. You can set the direct link between the primary node and the standby primary node.
  • the weight value is greater than the weight value of the path between the primary node and the secondary node, and the weight value is greater than the path between the alternate primary node and the secondary node, for example: can be set to 130.
  • the embodiment of the present invention sets different weight values for the internal path of the virtual network unit, and may Implement load balancing within the virtual network unit.
  • the master node can compare the first slave node with the master.
  • the weight value of the path between the nodes and the weight value of the path between the first slave node and the standby master node When the weight value of the path between the first slave node and the master node is small, the master node mainly uses the path through the master node.
  • the primary node takes the path through the alternate primary node as the primary path to pass the primary node
  • the path is the secondary path.
  • the master node needs to compare the weight value of the path between the slave node 1 and the master node and the slave node 1 and The weight value of the path between the alternate primary nodes, since the weight value of the path between the slave node 1 and the master node is 1, and the weight of the path between the slave node 1 and the standby master node is 99, the master node is the slave node 1
  • the path between the slave node 2 and the master node is taken as the master path between the slave node 1 and the slave node 2, and the path between the slave node 1 and the slave node 2 through the standby master node is used as the slave node 1 and the slave node 2.
  • the master node After the weight value of the internal path and the internal path of the VNE is calculated, the master node performs step 104 to generate and save an internal routing entry according to the weight value of the primary path, the primary path, the weight value of the secondary path, and the secondary path.
  • the generated internal routing entries are sent to nodes in the VNE other than the primary node.
  • the internal path of the VNE is relatively fixed.
  • the internal path of the VNE is relatively fixed.
  • the internal path calculation there is a master node as each node in the VNE, and two paths of the node to all nodes except the self in the VNE are calculated, and each time in the VNE is recalculated only when the topology changes. Nodes to all other sections of the VNE except themselves The two paths of the point can greatly simplify the calculation of the internal path and minimize the occupied system resources.
  • the master node may first calculate, for each node in the VNE, two paths of the node to all nodes in the VNE except the above node; then the master node may determine a path with fewer nodes in the two paths. As the primary path, the path with more nodes is the secondary path, and the primary path is assigned a smaller weight value, and the secondary path is assigned a larger weight value. As shown in FIG. 7, FIG. 7 is in the ring topology of the present invention.
  • the primary node may perform step 104 to generate and save an internal routing entry based on the primary path, the weight value of the primary path, the secondary path, and the weight value of the secondary path, and The internal routing entry is sent to a node in the VNE other than the primary node.
  • the primary and secondary paths from the primary node to the secondary node 1, and the primary and secondary paths from the primary node to the secondary node 2 are shown in FIG. As shown in FIG. 7, the weight value of the primary path can be assigned to 1 and the weight value of the secondary path can be assigned to 99. Specifically, in the internal path calculation, the master node may directly assign the weight value of the path according to the cyclic link information according to the equations (5) and (6).
  • leftpath[i,j] is the hop count of the path between node i and node j on the left-hand loop path
  • rightpath[i,j] is the hop of the path between node i and node j on the right-hand loop path.
  • Numbers; leftpos(i) and leftpos1 are the positions of node i and node j on the left-hand loop path, respectively
  • rightpos(i) and rightpos(j) are the positions of node i and node j on the right-hand loop path, respectively.
  • weight(leftpath[i,j]) is the weight of the path between node i and node j on the left-handed path
  • weight(rightpath[i,j]) is The weight value of the path between node i and node j on the right hand loop path.
  • the master node may also determine that the weight of each path of each node in the VNE to any other node in the VNE is the sum of the pre-configured weight values on the path between the two adjacent nodes on the two paths, such as FIG.
  • FIG. 8 is a schematic diagram of another embodiment of a primary path and a secondary path in a ring topology according to the present invention, that is, in a case where a path between two adjacent nodes is pre-configured with a weight value.
  • the node calculates the weight values of the above two paths, it only needs to superimpose the pre-configured weight values on the paths between the adjacent two nodes on the left-hand path and the right-hand path. Then, the master node can determine that the path with the value of ' ⁇ !, the weight value is the primary path, and the path with the larger weight value is the secondary path.
  • different weight values are set for the internal path of the virtual network unit, and load balancing within the virtual network unit can be implemented.
  • the topology type of the VNE is a semi-ring topology
  • only one path between each two nodes in the VNE is reachable. For the convenience of calculation, it can start from an internal connection port of the primary node, pass all the nodes on the ring, and reach the standby primary node to form a complete semi-circular path.
  • the master node may calculate the slave node to the master node and the path of the slave node to the standby master node for each slave node in the VNE; then the master node compares the path from the slave node to the master node to the slave node to Set the number of nodes through which the path of the standby primary node passes, and set a larger weight value for the path with more nodes.
  • the primary node may perform step 104 to generate and save the internal route according to the primary path, the weight value of the primary path, the secondary path, and the weight value of the secondary path.
  • An entry and sends the generated internal route entry to a node in the VNE other than the primary node.
  • FIG. 9 is a schematic diagram of an embodiment of a primary path and a secondary path in a semi-ring topology of the present invention, and FIG. 9 shows a path from node 1 to a primary node and an alternate primary node, and from node 2 to a primary node and an alternate primary.
  • the path to the node As shown in Figure 9, the path weight value with a large number of nodes can be set to 99, and the path weight value with a small number of nodes can be set to 1.
  • the master node may also determine that the weight of the path from the node to the master node is the sum of the pre-configured weight values on the path between the two nodes on the path from the node to the master node, and determine the The weight of the path from the node to the standby primary node is the sum of the pre-configured weight values on the path between the adjacent two nodes on the path from the node to the standby primary node; that is, the half-ring topology can be configured.
  • the weight value of each path is superimposed on the weight value of the path between the node to the primary node or the adjacent two nodes on the alternate primary node path, and is the weight value of the path from the node to the primary node or the alternate primary node.
  • the master node can determine that the path with the smaller weight value is the primary path, and the path with the larger weight value is the secondary path.
  • FIG. 10 is the primary path and the secondary path in the semi-ring topology of the present invention.
  • the master node may perform step 104, generating and saving an internal routing entry based on the primary path, the weight value of the primary path, the secondary path, and the weight value of the secondary path, and transmitting the generated internal routing entry Give the VNE a node other than the primary node.
  • different weight values are set for the internal path of the virtual network unit, and load balancing within the virtual network unit can be implemented.
  • FIG. 11 is a schematic structural diagram of an embodiment of a master node device in a virtual network unit according to the present invention.
  • the master node device in this embodiment can implement the process of the embodiment shown in FIG. 1 of the present invention.
  • the device can include:
  • the receiving module 1101 is configured to receive topology information reported by the node device other than the master node device in the VNE;
  • the saving module 1102 is configured to save the topology information received by the receiving module 1101.
  • the verification module 1103 is configured to perform topology verification on all the topology information saved by the saving module 1102 according to the topology type of the VNE.
  • the topology information saved by the saving module 1102 includes topology information and reception of the master node device. Topology information received by module 1101;
  • the calculating module 1104 is configured to: according to the topology type and the saving module, after the topology verification is passed 1102.
  • the saved topology information, the internal path of the VNE and the weight value of the internal path are calculated;
  • the generating module 1105 is configured to generate and save an internal routing entry according to the internal path calculated by the computing module 1104 and the weight value of the internal path;
  • the sending module 1106 is configured to send the internal routing entry generated by the generating module 1105 to a node device other than the master node device in the VNE.
  • the above-mentioned master node device can calculate the internal path of the VNE under various topological structures such as double-homed tree, ring and semi-ring, and thus can realize simple and effective calculation of the primary path and the secondary path in the internal path of the VNE. .
  • FIG. 12 is a schematic structural diagram of another embodiment of a master node device in a virtual network unit according to the present invention. Compared with the master node device shown in FIG. 11, the difference is that, in this embodiment, the sending module 1106 can also be broadcasted. Sending a topology collection request message of the VNE, where the topology collection request message carries information of the primary node device, an identifier, a name, and a topology type of the VNE;
  • the receiving module 1101 may receive a topology collection response message sent by a node device other than the master node device in the VNE, where the topology collection response message carries topology information; the topology collection response message is in addition to the master node device in the VNE.
  • the node device After receiving the topology collection request message, the node device determines that the VNE corresponding to the VNE identifier carried in the topology collection request message is sent to the master node device after being sent to the VNE.
  • the saving module 1102 can save the above topology information, and all the node devices in the VNE are included in the VNE cluster management.
  • the receiving module 1101 may also receive the topology information reported by the node device other than the master node device in the VNE after the location of the master node device is learned; at this time, the saving module 1102 may also save the topology information, and All node devices in the VNE are included in the VNE cluster management.
  • the sending module 1106 may also send the topology message periodically or periodically in the VNE after the topology check is passed, where the topology message carries the topology information of each node device in the VNE; After the node device that receives the topology message determines that its topology information is different from the topology information carried in the topology message, the node information is sent to the master node device.
  • the verification module 1103 may determine, according to the topology information saved by the save module 1102, whether the verification pass condition corresponding to the topology type has been met. Specifically, the verification module 1103 may determine whether the verification pass condition corresponding to the topology type has been met according to the method provided by the method embodiment of the present invention, and details are not described herein again.
  • the calculation module 1104 can include:
  • a first calculation sub-module 11041 configured to calculate two paths between each two slave node devices in the VNE when the topology type is a dual-homing tree topology, and calculate the slave node device in the VNE a path from the node device to the master node device and the slave node device to the standby master node device; the first determining submodule 11042, configured to determine that the weight of the path through the master node device in the two paths is a first slave node The weight value of the path between the device and the master node device, and the sum of the weight values of the path between the master node device and the second slave node device; determining the weight value of the path through the standby master node device in the two paths is first The weight value of the path between the slave node device and the standby master node device, and the weight value of the path between the slave master node device and the second slave node device; and determining that the path having the smaller weight value among the two paths is The primary path, the path with the larger weight value is the secondary path
  • the weight value of the path between the primary node device and the first slave node device or the second slave node device is pre-configured, or the primary node device is allocated according to the number of the first slave node device or the second slave node device;
  • the weight value of the path between the master node device and the first slave node device or the slave node device is pre-configured, or the master node device is allocated according to the number of the first slave node device or the second slave node device.
  • different weight values are set for the internal path of the virtual network unit, and load balancing within the virtual network unit can be implemented.
  • the first determining submodule 11042 may further compare the weight value of the path between the first slave node device and the master node device with the first slave node device and the standby master node device when the weight values of the two paths are the same.
  • the weight value of the inter-path when the weight value of the path between the first slave node device and the master node device is small, the path through the master node device is the main path to pass the alternate main section.
  • the path of the point device is the secondary path.
  • the path through the standby master node device is the primary path
  • the path through the master node device is the secondary path.
  • the generating module 1105 may generate an internal routing entry according to the primary path, the weight value of the primary path, the secondary path, and the weight value of the secondary path; the saving module 1102 may save the internal routing entry generated by the generating module 1105.
  • the calculation module 1104 can include:
  • a second calculation sub-module 11043 configured to calculate, when the topology type is a ring topology, a path of the node device to each node device in the VNE except the foregoing node device, for each node device in the VNE ;
  • the second determining sub-module 11044 is configured to determine that the path that is less than the node device in the two paths is the primary path, and the path that is more than the node device is the secondary path, and the primary path is assigned a smaller weight value.
  • the path assigns a larger weight value; or, determines that the weight values of the two paths are the sum of the pre-configured weight values on the paths between the adjacent two node devices on the two paths, and determines that the weight value has a smaller weight value.
  • the path is the primary path, and the path with the larger weight value is the secondary path;
  • the generating module 1105 is specifically configured to generate an internal routing entry according to the weight value of the primary path, the weight value of the primary path, the secondary path, and the secondary path.
  • the calculation module 1104 can include:
  • a third calculation sub-module 11045 configured to calculate a path from the node device to the master node device and the slave node device to the standby master node device for each slave node device in the VNE when the topology type is a semi-ring topology;
  • the third determining sub-module 11046 is configured to compare the path of the path from the node device to the master node device to the path of the slave node device to the standby master node device, and set a larger path for the path that passes through the node device.
  • the weight value is set to a smaller weight value for the path that has a smaller number of node devices; or the third determining sub-module 11046 is configured to determine a weight value of the path from the node device to the master node device from the node device to the master Between two adjacent node devices on the path of the node device The sum of the pre-configured weight values on the path, and the weight value of the path from the node device to the standby primary node device is determined as a pre-configured weight on the path between the adjacent two node devices on the path from the node device to the standby primary node device. The sum of the values;
  • the third determining submodule 11046 may further determine that the path having the smaller weight value is the primary path, and the path having the larger weight value is the secondary path;
  • the generation module 1105 can generate an internal routing entry based on the primary path, the weight value of the primary path, the secondary path, and the weight value of the secondary path.
  • the above-mentioned master node device can calculate the internal path of the VNE under various topological structures such as double-homed tree, ring and semi-ring, and thus can realize simple and effective calculation of the primary path and the secondary path in the internal path of the VNE. .
  • modules in the apparatus in the embodiments may be distributed in the apparatus of the embodiment according to the description of the embodiments, or may be correspondingly changed in one or more apparatuses different from the embodiment.
  • the modules of the above embodiments may be combined into one module, or may be further split into a plurality of sub-modules.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Description

虚拟网络单元中的路由计算方法和主节点设备
本发明实施例涉及通信技术领域, 尤其涉及一种虚拟网络单元中的路由
背景技术
虚拟网络单元( Virtual Network Element; 以下简称: VNE )是一种集 群路由器技术, 基于软件技术将两层或多层汇聚结构中的多台路由器设备 虚拟化聚合为一个虚拟汇聚节点, 对外呈现为一个单一映像的集群路由 器。 VNE的控制平面和管理平面集中在一台主路由器, 即主节点上, 由另 一台路由器充当备节点, 其他路由器为从节点。 VNE内所有节点都可进行 流量转发。
在 VNE 内部, 主节点和备节点通常由大容量、 高性能、 拥有高性能 中央处理器 (Central Processing Unit; 以下简称: CPU ) 、 网络位置较高 的路由器来承担, 主节点通过收集 VNE 内部拓朴, 进行相应的计算, 生 成内邵转发表 ( internal Forwarding Information Base; 以下简称: iFIB ) , 然后将 iFIB下发到其他各节点上, 从而完成 VNE内的路由计算。 当有报 文进入 VNE内部时, 先查询 iFIB , 得到 VNE出接口, 然后转发到相应的 VNE 节点; 然后当才艮文到达该出接口所在节点时再查询外部转发表 ( external Forwarding Information Base; 以下简称: eFIB ) , 获得下一跳 , 然后将该报文从 VNE转发出去。 从整个转发过程来看, VNE对外屏蔽了 内部的互连细节, 呈现为一个具有单映像的超级节点, 从网络整体拓朴的 视角来看, 就是一个路由节点, 对外提供统一的 eFIB进行报文转发。
VNE技术简化了网络拓朴,提高了路由收敛性和稳定性, 降低了管理 复杂性。 和多框路由器相比, 无需引入专用互连设备和互连接口技术, 保 护了现有投资, 降低了投资和运维成本。 并且可以根据网络应用, 把不同 型号、 不同形态的设备聚合起来, 减少了设备的维护性。
现有的一种 VNE 内部拓朴计算方案为按需路由选择 (On Demand Routing; 以下简称: ODR ) 技术, 但是 ODR 技术仅适用于中心 -分支 ( hub&spoke )形状的拓朴, 不能适用于双归树形和环形等其他拓朴结构, 适用范围较窄。 发明内容
本发明实施例提供一种虚拟网络单元中的路由计算方法和主节点设备, 以实现在多种拓朴结构下, 对虚拟网络单元的内部路由进行计算。
本发明实施例提供一种虚拟网络单元中的路由计算方法, 包括: 虚拟网络单元中的主节点接收并保存虚拟网络单元中除所述主节点之外 的节点上报的拓朴信息;
所述主节点根据所述虚拟网络单元的拓朴类型对所述主节点本地保存的 所有拓朴信息进行拓朴校验, 所述保存的拓朴信息包括所述主节点的拓朴信 息和所述虚拟网络单元中除所述主节点之外的节点上报的拓朴信息;
在所述拓朴校验通过之后, 所述主节点根据所述拓朴类型和所述拓朴信 息, 计算所述虚拟网络单元的内部路径和所述内部路径的权重值;
所述主节点根据所述内部路径和所述内部路径的权重值生成并保存 内部路由条目, 并将所述内部路由条目发送给所述虚拟网络单元中除所述 主节点之外的节点。
本发明实施例还提供一种虚拟网络单元中的主节点设备, 包括: 接收模块, 用于接收虚拟网络单元中除所述主节点设备之外的节点设备 上报的拓朴信息;
保存模块, 用于保存所述接收模块接收的拓朴信息;
校验模块, 用于根据所述虚拟网络单元的拓朴类型对所述保存模块保存 的所有拓朴信息进行拓朴校验, 所述保存模块保存的所有拓朴信息包括所述 主节点设备的拓朴信息和所述接收模块接收的拓朴信息;
计算模块, 用于在所述拓朴校验通过之后, 根据所述拓朴类型和所述保 存模块保存的拓朴信息, 计算所述虚拟网络单元的内部路径和所述内部路径 的权重值;
生成模块, 用于根据所述计算模块计算的所述内部路径和所述内部路径 的权重值生成并保存内部路由条目;
发送模块, 用于将所述生成模块生成的内部路由条目发送给所述虚拟 网络单元中除所述主节点设备之外的节点设备。
通过本发明实施例, 主节点接收并保存虚拟网络单元中的节点上报的拓 朴信息之后, 根据该虚拟网络单元的拓朴类型对保存的拓朴信息进行拓朴校 验, 在拓朴校验通过之后, 主节点根据拓朴类型和上述拓朴信息, 计算虚拟 网络单元的内部路径和内部路径的权重值; 然后, 主节点可以根据上述内部 路径和内部路径的权重值生成并保存内部路由条目, 并将该内部路由条目发 送给虚拟网络单元中除主节点之外的节点。 从而可以实现在双归树形、 环形 和半环形等多种拓朴结构下, 对虚拟网络单元的内部路径进行计算, 进而可 以实现简单有效的计算虚拟网络单元的内部路径中的主路径和次路径, 并且 可以通过对内部路径设置不同的权重值, 实现虚拟网络单元内部的负载均衡。 附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案, 下面将对实 施例或现有技术描述中所需要使用的附图作一简单地介绍, 显而易见地, 下 面描述中的附图是本发明的一些实施例, 对于本领域普通技术人员来讲, 在 不付出创造性劳动的前提下, 还可以根据这些附图获得其他的附图。
图 1为本发明路由计算方法一个实施例的流程图;
图 2为本发明拓朴收集一个实施例的示意图; 图 3为本发明双归树形拓 41、一个实施例的示意图;
图 4为本发明环形拓朴一个实施例的示意图;
图 5为本发明半环形拓朴一个实施例的示意图;
图 6为本发明双归树形拓朴中主路径与从路径一个实施例的示意图; 图 7为本发明环形拓朴中主路径与从路径一个实施例的示意图; 图 8为本发明环形拓朴中主路径与从路径另一个实施例的示意图; 图 9为本发明半环形拓朴中主路径与从路径一个实施例的示意图; 图 10为本发明半环形拓朴中主路径与从路径另一个实施例的示意图; 图 11为本发明虚拟网络单元中的主节点设备一个实施例的结构示意图; 图 12为本发明虚拟网络单元中的主节点设备另一个实施例的结构示意图。 具体实施方式
为使本发明实施例的目的、 技术方案和优点更加清楚, 下面将结合本发 明实施例中的附图, 对本发明实施例中的技术方案进行清楚、 完整地描述, 显然, 所描述的实施例是本发明一部分实施例, 而不是全部的实施例。 基于 本发明中的实施例, 本领域普通技术人员在没有做出创造性劳动的前提下所 获得的所有其他实施例, 都属于本发明保护的范围。
图 1为本发明路由计算方法一个实施例的流程图, 如图 1所示, 该路由 计算方法可以包括:
步骤 101 , VNE中的主节点接收并保存 VNE中除该主节点之外的节点上 报的拓朴信息。
本实施例中, VNE是一个虚拟异构的集群路由器, VNE内包括主节点、 备用主节点和从节点。 主节点为汇聚层大容量高性能汇聚路由器, 是对整个 VNE内部所有节点进行管理的节点, 用于集中处理 VNE的控制信令, 并将 计算得到的控制信息发给从节点, VNE 的控制平面在主节点上, 上述 VNE 内部所有节点包括主节点、 备用主节点和所有从节点。 一个 VNE 内部只有 一个主节点和一个备用主节点。 备用主节点是在主节点发生故障, 不能对
VNE内部所有节点进行管理时, 来接替主节点对 VNE内部所有节点进行管 理的。
VNE中的从节点, 为汇聚层小容量低性能路由器, 不处理控制信令, 只 负责数据转发。
步骤 102, 主节点根据 VNE的拓朴类型对该主节点本地保存的所有拓朴 信息进行拓朴校验, 该主节点本地保存的所有拓朴信息包括主节点的拓朴信 息和 VNE中除该主节点之外的节点上报的拓朴信息。
步骤 103 , 在拓朴校验通过之后, 主节点根据拓朴类型和拓朴信息, 计 算 VNE的内部路径和所述内部路径的权重值。
本实施例中, 内部路径是指 VNE 中任意两个节点之间的通路信息, 包 括两个节点之间的通路经过的节点和相应的接口; 内部路由是指 VNE 中的 主节点根据内部路径为每个节点生成的标准路由信息。
步骤 104, 主节点根据上述内部路径和内部路径的权重值生成并保存内 部路由条目, 并将该内部路由条目发送给 VNE中除主节点之外的节点。
上述实施例可以实现在双归树形、 环形和半环形等多种拓朴结构下, 对 VNE的内部路径进行计算, 进而可以实现简单有效的计算 VNE的内部路径 中的主路径和次路径。
下面对本发明实施例提供的路由计算方法进行详细介绍。
本发明实施例中,在 VNE中被指定为主节点的节点启动之后,在对 VNE 进行管理和内部路由计算之前, 需要先进行拓朴收集, 拓朴收集可以采用动 态收集和静态配置两种方式。
( 1 )动态收集方式:
图 2为本发明拓朴收集一个实施例的示意图, 如图 2所示, 本发明实施 例中, 在步骤 101 , 主节点接收并保存 VNE中除该主节点之外的节点上报的 拓朴信息之前, 主节点可以广播方式发送 VNE 的拓朴收集请求消息, 该拓 朴收集请求消息携带主节点的信息、 VNE的标识、 名称和拓朴类型等, 具体 地, 主节点可以二层或者三层广播方式发送 VNE的拓朴收集请求消息; 然后, 主节点可以接收 VNE 中除主节点之外的节点发送的拓朴收集响 应消息, 该拓朴收集响应消息携带拓朴信息; 其中, 该拓朴收集响应消息是 VNE中除主节点之外的节点接收到上述拓朴收集请求消息之后,确定该拓朴 收集请求消息中携带的 VNE的标识对应的 VNE为自身归属的 VNE之后发 送给上述主节点的; 如果 VNE 中除主节点之外的节点接收到上述拓朴收集 请求消息之后, 确定该拓朴收集请求消息中携带的 VNE的标识对应的 VNE 不是自身归属的 VNE, 则可以不响应上述拓朴收集请求消息。
主节点接收到 VNE 中除主节点之外的节点发送的拓朴收集响应消息之 后, 可以保存该拓朴收集响应消息携带的拓朴信息, 并将上述 VNE 中的所 有节点纳入到 VNE集群管理中; 然后向上述 VNE中除主节点之外的节点发 送注册成功消息, 以指示上述 VNE中除主节点之外的节点在该 VNE内注册 成功。
VNE的主节点每接收并保存上述 VNE中除主节点之外的一个节点的拓 朴信息之后, 就会对该主节点本地保存的所有拓朴信息进行一次拓朴校验。 如果拓朴校验通过,则该主节点可以确定整个 VNE拓朴信息收集完整, VNE 拓朴建立, 可以开始进行内部路径的计算; 如果拓朴校验未通过, 则主节点 会继续周期性或定时以广播的方式发送拓朴收集请求消息, 继续等待 VNE 中的节点上报的拓朴信息, 直至整个 VNE拓朴信息收集完整。 具体地, 主 节点可以在接收并保存 VNE 中除主节点之外的一个节点上报的拓朴信息之 后, 可以马上对该主节点本地保存的所有拓朴信息进行一次拓朴校验; 在 VNE中除主节点之外的一个节点下线或者检测不到该节点时,可以设置一个 等待时间,在超过该等待时间之后,主节点重新进行拓朴校验。如果出现 VNE 中除主节点之外的一个节点频繁上线或下线, 导致拓朴校验不能通过时, 则 主节点可以先不将之前计算好的与该频繁上线或下线的节点对应的内部路径 删除, 而是把该内部路径去激活, 以保证数据的正常转发。 只有在新的拓朴 校验通过, 且 VNE内所有节点的内部路径均计算完成后, VNE中的主节点 再生成新的内部路由条目, 下发给 VNE中的所有节点进行路由更新。
( 2 )静态配置方式:
具体地, 步骤 101 可以为主节点接收 VNE中除主节点之外的节点在获 知该主节点的位置后上报的拓朴信息。也就是说,在静态配置方式下,在 VNE 中的节点获知主节点的位置之后, 可以主动向主节点上报拓朴信息。
接收到上述拓朴信息之后, 主节点保存该拓朴信息, 并将上述 VNE 中 的所有节点纳入 VNE集群管理中。
与动态收集方式相同, VNE的主节点每接收并保存上述 VNE中的一个 节点上报的拓朴信息之后, 就会对该主节点本地保存的所有拓朴信息进行一 次拓朴校验。 如果拓朴校验通过, 则该主节点可以确定整个 VNE拓朴信息 收集完整, VNE拓朴建立, 可以开始进行内部路径的计算; 如果拓朴校验未 通过, 则主节点会继续等待 VNE中的节点上报的拓朴信息, 直至整个 VNE 拓朴信息收集完整。
需要说明的是, 本发明实施例中的拓朴信息包括 VNE 中的节点的标识 符、 该节点所在设备的基本信息、 外连接口信息和内连接口信息。
其中, 内连接口是指在 VNE 内部各节点间连接的接口; 外连接口是指 VNE内部的节点与 VNE外部的设备连接的接口;即 VNE内每个节点的接口 分为内连接口和外连接口, 分别连接 VNE内部的节点和 VNE外部的设备。
外连接口信息包括外连接口的接口名、 因特网协议 ( Internet Protocol; 以下简称: IP )地址和该外连接口的封装信息 (例如: 该外连接口是二层接 口或三层接口)等; 内连接口信息包括内连接口的接口名、 IP地址和该内连 接口的封装信息 (例如: 该内连接口是二层接口或三层接口)等。
本发明实施例中, 在拓朴校验通过之后, 即 VNE拓朴建立之后, 如果 VNE的拓朴结构发生变化, 则主节点需要进行拓朴更新。 具体地, 在拓朴校验通过之后, 主节点可以在 VNE 中定时或周期性发 送拓朴报文, 该拓朴报文携带 VNE 中各节点的拓朴信息; 以便接收到上述 拓朴报文的节点确定自身的拓朴信息与拓朴报文中携带的拓朴信息不同之 后, 将自身的拓朴信息发送给主节点。 主节点接收到该拓朴信息后根据实际 情况进行相应的处理, 例如: 进行路径切换或者进行新的路径计算等。 其中, 该拓 4卜报文可以采用广播方式发送, 也可以采用定向方式发送。
在拓朴校验通过之后, 如果 VNE 中新增节点, 则主节点可以接收并保 存 VNE中的新增节点上报的该新增节点的拓朴信息, 然后重新根据 VNE的 拓朴类型对保存的拓朴信息进行拓朴校验, 也就是说, 由 VNE 中的新增节 点主动向主节点上报自身的拓朴信息; 或者, 主节点可以接收并保存 VNE 中与新增节点相连的节点发送的拓朴信息, 然后重新根据 VNE 的拓朴类型 对该主节点本地保存的所有拓朴信息进行拓朴校验, 其中, 该拓朴信息是与 新增节点相连的节点接收到主节点发送的拓朴报文之后, 确定自身的外连接 口信息和内连接口信息已经发生变化, 即确定自身的拓朴信息与该拓朴报文 中携带的拓朴信息不同之后, 发送给主节点的; 或者, 主节点可以接收并保 存新增节点在接收到主节点发送的拓朴报文之后上报的拓朴信息, 重新根据 VNE的拓朴类型对该主节点本地保存的所有拓朴信息进行拓朴校验。
在拓朴校验通过之后, 如果 VNE中去除节点, 则主节点可以接收并保存 VNE中与去除的节点直连的节点上报的变化后的拓朴信息, 然后重新根据拓 朴类型对该主节点本地保存的所有拓朴信息进行拓朴校验; 上述变化后的拓 朴信息由与去除的节点直连的节点主动上报给主节点, 或者上述变化后的拓 朴信息是与去除的节点直连的节点接收到主节点发送的拓朴报文之后, 确定 自身的拓朴信息与拓朴报文中携带的拓朴信息不同之后, 发送给主节点的。
另外在控制平面,主节点与 VNE中除主节点之外的各节点间可以通过心 跳等机制, 来检测 VNE中除主节点之外的节点的故障。如果主节点收不到某 节点对心跳的响应, 就可以确定该节点发生故障。 本发明实施例中, 在拓朴 校验通过之后, 当 VNE中除主节点之外的节点的控制平面发生故障, 而转发 平面未发生故障时, 主节点通过心跳机制发现上述节点发生故障之后, 主节 点可以删除发生故障的节点的拓朴信息,重新接收并保存 VNE中除主节点和 发生故障的节点之外的节点上报的拓朴信息,并根据 VNE的拓朴类型对该主 节点本地保存的所有拓 4卜信息进行拓 卜校验。
而当 VNE 中除主节点之外的节点的数据平面发生故障时, 主节点可以 接收并保存发生故障的节点上报的变化后的拓朴信息, 然后重新根据 VNE 的拓朴类型对该主节点本地保存的所有拓朴信息进行拓朴校验。
在拓朴校验通过之后, 当 VNE中的内部链路发生故障时, 主节点可以接 收并保存发生故障链路上的节点通过备用路径上报的变化后的拓朴信息, 以 及接收并保存与发生故障链路上的节点相连的节点上报的变化后的拓朴信 息,然后主节点可以重新根据 VNE的拓朴类型对该主节点本地保存的所有拓 朴信息进行拓朴校验。
下面对步骤 102, 主节点根据 VNE的拓朴类型对该主节点本地保存的所 有拓朴信息进行拓朴校验的方法进行说明。
( 1 ) 图 3为本发明双归树形拓 41、一个实施例的示意图, 当 VNE的拓朴 类型为双归树形拓朴时, 主节点可以根据该主节点本地保存的所有拓朴信息 确定双归树形拓朴的校验通过条件是否已满足; 这时, 拓朴校验通过可以为: 双归树形拓朴的校验通过条件已满足; 也就是说, 当双归树形拓朴的校验通 过条件均被满足时, 主节点可以确定双归树形拓朴校验通过。
其中, 双归树形拓朴的校验通过条件包括: VNE内部所有节点均至少有 两个状态可用的内连接口; 并且主节点与备用主节点通过内连接口直连, 主 节点与备用主节点之间的内连接口和链路状态正常;并且 VNE中所有从节点 有且只有两个内连接口, 分别连接到主节点和备用主节点上; 并且 VNE中已 上报自身拓朴信息的从节点的个数与预先规划的从节点的个数相同。
( 2 ) 图 4为本发明环形拓朴一个实施例的示意图, 当 VNE的拓朴类型 为环形拓朴时, 主节点可以根据该主节点本地保存的所有拓朴信息确定环形 拓朴的校验通过条件是否已满足; 这时, 拓朴校验通过可以为环形拓朴的校 验通过条件已满足; 也就是说, 当环形拓朴的校验通过条件均被满足时, 主 节点可以确定环形拓朴校验通过。
其中, 环形拓朴的校验通过条件包括: VNE中的每个节点都有两个状态 正常的内连接口; 并且主节点与备用主节点通过内连接口直连; 并且 VNE中 的节点两两相连,从主节点的一个内连接口开始, 通过 VNE中除主节点之外 的各节点能够到达主节点的另一个内连接口; 并且从主节点的两个内连接口 出发形成的两条环状路径上, 经过节点的顺序正好相反。
( 3 ) 图 5为本发明半环形拓朴一个实施例的示意图, 当 VNE的拓朴类 型为半环形拓朴时, 主节点可以根据该主节点本地保存的所有拓朴信息确定 环形拓朴的校验通过条件是否已满足; 这时, 拓朴校验通过可以为半环形拓 朴的校验通过条件已满足; 也就是说, 当半环形拓朴的校验通过条件均被满 足时, 主节点可以确定环形拓朴校验通过。
其中, 半环形拓朴的校验通过条件包括: VNE中除主节点和备用主节点 之外的每个节点均有至少两个状态正常的内连接口; 并且, 主节点和备用主 节点均有至少一个内连接口, 且主节点与备用主节点之间没有直连的内部链 路; 并且, VNE中除主节点与备用主节点之外的节点两两相连, 从主节点的 一个内连接口开始, 通过 VNE中除主节点与备用主节点之外的各节点, 能够 到达备用主节点的一个状态正常的内连接口; 并且, 从主节点的一个内连接 口出发形成的半环状路径通过了 VNE中除所述主节点之外的所有节点,该半 环状路径的最后一个节点为备用主节点。
在拓朴校验通过之后, 主节点可以确定整个 VNE拓朴信息收集完整, VNE拓朴建立, 可以开始进行内部路径的计算; 如果拓朴校验未通过, 则主 节点会继续周期性或定时以广播的方式发送拓朴收集请求消息, 继续等待 VNE中除主节点之外的节点上报的拓朴信息, 直至整个 VNE拓朴信息收集 完整。
下面对步骤 103 , 在拓朴校验通过之后, 主节点根据拓朴类型和拓朴信 息, 计算 VNE的内部路径和该内部路径的权重值的方法进行说明。
1、 当 VNE的拓朴类型为双归树形拓朴时, VNE的内部路径是集中由主 节点计算的。具体地,主节点可以计算 VNE中每两个从节点之间的两条路径, 以及为 VNE中的每个从节点计算该从节点到主节点,以及该从节点到备用主 节点的路径; 然后确定上述两条路径中通过主节点的路径的权重值为第一从 节点与主节点之间路径的权重值, 与主节点与第二从节点之间路径的权重值 之和, 如式(1 )所示; 确定上述两条路径中通过备用主节点的路径的权重值 为第一从节点与备用主节点之间路径的权重值, 与该备用主节点与第二从节 点之间路径的权重值之和, 如式( 2 )所示; 以及确定两条路径中具有较 d、权 重值的路径为主路径, 具有较大权重值的路径为次路径。
需要说明的是, 本发明实施例中的 "第一" 与 "第二" 仅为描述方便, 用于区分两个不同的从节点, 并不作为对从节点的限定。
weight[i] [主] [j] = weightto主 [i] + weighty [j] ( 1 ) eight[i] [备] [ j] = weightto备 [i] + weightto备 [ j] ( 2 ) 式( 1 )中, weight[i] [主] [j]为第 i个从节点与第 j个从节点之间通过主节 点的路径的权重值, weightt。主 [i]为第 i个从节点与主节点之间路径的权重值, weighty [j]为第 j个从节点与主节点之间路径的权重值, 即主节点与第 j个 从节点之间路径的权重值; 式(2 ) 中, weight[i] [备] [j]为第 i个从节点与第 j 个从节点之间通过备用主节点的路径的权重值, weightt。Ji]为第 i个从节点 与备用主节点之间路径的权重值, weightt。Jj]为第 j个从节点与备用主节点 之间路径的权重值, 即备用主节点与第 j个从节点之间路径的权重值。
本发明实施例中, 主节点与第一从节点或第二从节点之间路径的权重值 可以是主节点根据第一从节点或第二从节点的编号分配的, 备用主节点与第 一从节点或第二从节点之间路径的权重值也可以是主节点根据第一从节点或 第二从节点的编号分配的, 例如: 对于编号为奇数的从节点, 以该节点与主 节点之间的路径为主路径, 权重值分配为 1; 以该节点与备用主节点之间的 路径为次路径, 权重值分配为 99, 如式(3)和式(4)所示; 对于编号为偶 数的从节点, 以该节点与主节点之间的路径为次路径, 权重值分配为 99; 以 该节点与备用主节点之间的路径为主路径, 权重值分配为 1, 如式(3)和式 (4)所示。 采用上述分配方式, 双归树形拓朴结构的 VNE中, 从节点与根 节点之间的主路径与从路径可以如图 6所示, 图 6为本发明双归树形拓朴中 主路径与从路径一个实施例的示意图, 图 6中, 虚线所示为从节点与根节点 之间的主路径, 点划线所示为从节点与根节点之间的次路径。 其中, 根节点 包括主节点和备用主节点。
1, i为奇数
weightt0主 [i] = (3)
99, i为偶数 99, i为奇数;
weightto备 [i] = (4)
1, i为偶数;
式(3) 中, weightt。 [i]为第 i个从节点与主节点之间路径的权重值; 式
(4) 中, weightt。Ji]为第 i个从节点与备用主节点之间路径的权重值。
以图 6所示 VNE为例, 通过上述方法, 主节点为 VNE中每两个从节点 计算的主路径与次路径可以如表 1所示。
表 1
Figure imgf000014_0001
节点 3 从 从节 点 从节点 1 从节 点 从节点 1 从节 点 从 节 节 1- 主 节 -备用主 1- 主 节 -备用主 1- 主 节 点 1 -
,#、 ,#、 节点 点 -从节 节点-从 点 -从节 备 用
1 点 2 节点 2 点 3 主 节 点-从 节点 3 从 从节 点 从节 点 从节点 2 从节 点 从节点 2 从 节 节 2-备用 2- 主 节 -备用主 2- 主 节 -备用主 点 2-
,#、 主节点 ,#、 节点-从 点 -从节 节点-从 主 节
2 节点 1 点 1 节点 3 点-从 节点 3 从 从节 点 从节点 3 从节 点 从节点 3 从节 点 从节点 3
节 3- 主 节 -备用主 3- 主 节 -备用主 3- 主 节 -备用主
,#、 ,#、 节点 点 -从节 节点-从 点 -从节 节点-从
3 点 1 节点 1 点 2 节点 2
另外主节点与第一从节点或第二从节点之间路径的权重值可以是预先配 置的, 同样备用主节点与第一从节点或第二从节点之间路径的权重值也可以 是预先配置的; 也就是说, 主节点与第一从节点或第二从节点之间路径的权 重值也可以不按照式(3 )所示的方式分配, 而是预先配置一任意数值; 同样 备用主节点与第一从节点或第二从节点之间路径的权重值也可以不按照式 ( 4 )所示的方式分配, 而是预先配置一任意数值。
双归树形拓朴中, 主节点与备用主节点之间的直连链路用于主备倒换时 使用, 一般不用于数据流量, 可以设置主节点与备用主节点之间直连链路的 权重值大于主节点与从节点之间路径的权重值, 以及大于备用主节点与从节 点之间路径的权重值, 例如: 可以设置为 130。
本发明实施例对上述虚拟网络单元的内部路径设置不同的权重值, 可以 实现虚拟网络单元内部的负载均衡。
当 VNE的拓朴类型为双归树形拓朴时,如果两个从节点之间通过主节点 的路径与通过备用主节点的路径的权重值相同, 则主节点可以比较第一从节 点与主节点之间路径的权重值与第一从节点与备用主节点之间路径的权重 值, 当第一从节点与主节点之间路径的权重值较小时, 主节点以通过主节点 的路径为主路径, 以通过所述备用主节点的路径为次路径; 当第一从节点与 备用主节点之间路径的权重值较小时, 主节点以通过备用主节点的路径为主 路径, 以通过主节点的路径为次路径。 以图 6所示 VNE为例, 当按照式( 3 ) 和式(4 )所示的方式分配从节点与主节点, 以及从节点与备用主节点之间路 径的权重值时, 从节点 1与从节点 2之间通过主节点的路径的权重值与通过 备用主节点的路径的权重值是相同的, 这时主节点需要比较从节点 1与主节 点之间路径的权重值以及从节点 1与备用主节点之间路径的权重值, 由于从 节点 1与主节点之间路径的权重值为 1 , 而从节点 1与备用主节点之间路径 的权重值为 99, 因此主节点以从节点 1与从节点 2之间通过主节点的路径作 为从节点 1与从节点 2之间的主路径, 以从节点 1与从节点 2之间通过备用 主节点的路径作为从节点 1与从节点 2之间的次路径。
通过上述方式, 计算 VNE的内部路径和内部路径的权重值之后 , 主节点 执行步骤 104, 根据上述主路径、 主路径的权重值、 上述次路径和次路径的 权重值生成并保存内部路由条目,并将生成的内部路由条目发送给 VNE中除 主节点之外的节点。
2、 当 VNE的拓朴类型为环形拓朴时, VNE的内部路径是相对比较固定 的, 对于环形拓朴来说, 每两个节点之间存在比较固定的两条路径, 环形拓 朴下的这两条路径, 可以称之为左手路径和右手路径, 合起来正好是一个完 整 VNE环状链路。 在内部路径计算时, 有主节点为 VNE中的每个节点, 计 算该节点到 VNE中除自身之外的其他所有节点的两条路径,只有在拓朴发生 变化时才重新计算 VNE中的每个节点到该 VNE中除自身之外的其他所有节 点的两条路径, 从而可以大大简化内部路径的计算, 使占用的系统资源最少。 具体地, 主节点可以先为 VNE中的每个节点, 计算该节点到 VNE中除 上述节点之外的其他所有节点的两条路径; 然后主节点可以确定两条路径中 经过节点较少的路径为主路径, 经过节点较多的路径为次路径, 并为主路径 分配较小的权重值, 为次路径分配较大的权重值, 如图 7所示, 图 7为本发 明环形拓朴中主路径与从路径一个实施例的示意图; 之后, 主节点可以执行 步骤 104, 根据上述主路径、 上述主路径的权重值、 次路径和上述次路径的 权重值生成并保存内部路由条目,并将该内部路由条目发送给 VNE中除主节 点之外的节点。
图 7中示出了主节点到从节点 1的主路径和次路径, 以及主节点到从节 点 2的主路径和次路径。 如图 7所示, 可以将主路径的权重值分配为 1 , 将 次路径的权重值分配为 99。 具体地, 在内部路径计算时, 主节点可以直接根 据环状链路信息, 按照式(5 )和式(6 )分配路径的权重值。
Figure imgf000017_0001
式( 5 ) 中, leftpath[i,j]为左手环路径上节点 i与节点 j之间路径的跳数, rightpath[i,j]为右手环路径上节点 i 与节点 j 之间路径的跳数; leftpos(i)与 leftpos①分别为节点 i与节点 j在左手环路径上的位置, rightpos(i)与 rightpos(j) 分别为节点 i与节点 j在右手环路径上的位置。
1; if ( leftpath≥ rightpath)
weight(leftpath[i,j])
99; if ( leftpath < rightpath)
J99; if ( leftpath≥ rightpath) weight(rightpath[i,j])
1; if ( leftpath < rightpath) 式( 6 ) 中, weight(leftpath[i,j])为左手环路径上节点 i与节点 j之间路径 的权重值, weight(rightpath[i,j])为右手环路径上节点 i与节点 j之间路径的权 重值。 另外, 主节点也可以确定 VNE中每个节点到 VNE中其他任一节点的两 条路径的权重值为上述两条路径上相邻两个节点之间路径上预先配置的权重 值之和, 如图 8所示, 图 8为本发明环形拓朴中主路径与从路径另一个实施 例的示意图, 也就是说, 在相邻两个节点间的路径都预先配置了权重值的情 况下, 主节点计算上述两条路径的权重值时, 只需要把左手路径、 右手路径 上相邻两个节点之间路径上预先配置的权重值叠加就可以了。 然后, 主节点 可以确定具有较'■!、权重值的路径为主路径,具有较大权重值的路径为次路径。
本发明实施例对上述虚拟网络单元的内部路径设置不同的权重值, 可以 实现虚拟网络单元内部的负载均衡。
3、 当 VNE的拓朴类型为半环形拓朴时, VNE中每两个节点之间只有一 条路径可达。 为方便计算, 可以从主节点的一个内连接口出发, 经过环上所 有节点, 到达备用主节点, 形成一个完整的半环形路径。 具体地, 主节点可 以为 VNE 中的每个从节点计算该从节点到主节点, 以及该从节点到备用主 节点的路径; 然后主节点比较上述从节点到主节点的路径与上述从节点到备 用主节点的路径所经过的节点数, 并为经过节点数较多的路径设置较大的权 重值, 为经过节点数较少的路径设置较小的权重值; 并确定具有较小权重值 的路径为主路径, 具有较大权重值的路径为次路径; 然后, 主节点可以执行 步骤 104, 根据主路径、 该主路径的权重值、 次路径和该次路径的权重值生 成并保存内部路由条目, 并将生成的内部路由条目发送给 VNE 中除主节点 之外的节点。
图 9为本发明半环形拓朴中主路径与从路径一个实施例的示意图, 图 9 中示出了从节点 1到主节点和备用主节点的路径, 以及从节点 2到主节点和 备用主节点的路径。 如图 9所示, 可以将经过节点数较多的路径权重值设置 为 99, 将经过节点数较少的路径权重值设置为 1。
另外, 主节点也可以确定从节点到主节点的路径的权重值为上述从节点 到主节点的路径上相邻两个节点之间路径上预先配置的权重值之和, 确定从 节点到备用主节点的路径的权重值为上述从节点到备用主节点的路径上相邻 两个节点之间路径上预先配置的权重值之和; 也就是说, 可以通过配置半环 形拓朴上每段路径的权重值, 将从节点到主节点或备用主节点路径上相邻两 个节点之间路径的权重值叠加, 作为从节点到主节点或备用主节点的路径的 权重值。 然后, 主节点可以确定具有较小权重值的路径为主路径, 具有较大 权重值的路径为次路径, 如图 10所示, 图 10为本发明半环形拓朴中主路径 与从路径另一个实施例的示意图; 之后, 主节点可以执行步骤 104, 根据主 路径、 该主路径的权重值、 次路径和该次路径的权重值生成并保存内部路由 条目, 并将生成的内部路由条目发送给 VNE中除主节点之外的节点。
本发明实施例对上述虚拟网络单元的内部路径设置不同的权重值, 可以 实现虚拟网络单元内部的负载均衡。
本领域普通技术人员可以理解: 实现上述方法实施例的全部或部分步骤 可以通过程序指令相关的硬件来完成, 前述的程序可以存储于一计算机可读 取存储介质中, 该程序在执行时, 执行包括上述方法实施例的步骤; 而前述 的存储介质包括: ROM、 RAM, 磁碟或者光盘等各种可以存储程序代码的介 质。
图 11为本发明虚拟网络单元中的主节点设备一个实施例的结构示意图, 本实施例中的主节点设备可以实现本发明图 1所示实施例的流程,如图 11所 示, 该主节点设备可以包括:
接收模块 1101 , 用于接收 VNE中除主节点设备之外的节点设备上报的 拓朴信息;
保存模块 1102, 用于保存接收模块 1101接收的拓朴信息;
校验模块 1103 , 用于根据 VNE的拓朴类型对保存模块 1102保存的所有 拓朴信息进行拓朴校验; 其中, 保存模块 1102保存的所有拓朴信息包括主节 点设备的拓朴信息和接收模块 1101接收的拓朴信息;
计算模块 1104, 用于在拓朴校验通过之后, 根据拓朴类型和保存模块 1102保存的拓朴信息, 计算 VNE的内部路径和所述内部路径的权重值; 生成模块 1105 , 用于根据计算模块 1104计算的内部路径和该内部路径 的权重值生成并保存内部路由条目;
发送模块 1106, 用于将生成模块 1105生成的内部路由条目发送给 VNE 中除主节点设备之外的节点设备。
上述主节点设备可以实现在双归树形、环形和半环形等多种拓朴结构下, 对 VNE的内部路径进行计算, 进而可以实现简单有效的计算 VNE的内部路 径中的主路径和次路径。
图 12 为本发明虚拟网络单元中的主节点设备另一个实施例的结构示意 图, 与图 11所示的主节点设备相比, 不同之处在于, 本实施例中, 发送模块 1106还可以广播方式发送 VNE的拓朴收集请求消息, 该拓朴收集请求消息 携带主节点设备的信息、 VNE的标识、 名称和拓朴类型;
接收模块 1101可以接收 VNE中除主节点设备之外的节点设备发送的拓 朴收集响应消息, 该拓朴收集响应消息携带拓朴信息; 该拓朴收集响应消息 是 VNE中除主节点设备之外的节点设备接收到拓朴收集请求消息之后,确定 拓朴收集请求消息中携带的 VNE的标识对应的 VNE为自身归属的 VNE之后 发送给该主节点设备的;
保存模块 1102可以保存上述拓朴信息, 并将 VNE中的所有节点设备纳 入 VNE集群管理中。
另外, 接收模块 1101也可以接收 VNE中除主节点设备之外的节点设备 在获知上述主节点设备的位置后上报的拓朴信息; 这时, 保存模块 1102也可 以保存上述拓朴信息, 并将 VNE中的所有节点设备纳入 VNE集群管理中。
进一步地, 本实施例中, 发送模块 1106还可以在拓朴校验通过之后, 在 VNE中定时或周期性发送拓朴报文, 该拓朴报文携带 VNE中各节点设备的 拓朴信息; 以便接收到上述拓朴报文的节点设备确定自身的拓朴信息与上述 拓朴报文中携带的拓朴信息不同之后,将自身的拓朴信息发送给主节点设备。 本实施例中, 校验模块 1103可以根据保存模块 1102保存的拓朴信息确 定拓朴类型对应的校验通过条件是否已满足。 具体地, 校验模块 1103可以按 照本发明方法实施例提供的方法确定拓朴类型对应的校验通过条件是否已满 足, 在此不再赘述。
本实施例中, 计算模块 1104可以包括:
第一计算子模块 11041 , 用于当拓朴类型为双归树形拓朴时, 计算 VNE 中每两个从节点设备之间的两条路径,以及为 VNE中的每个从节点设备计算 该从节点设备到主节点设备, 以及该从节点设备到备用主节点设备的路径; 第一确定子模块 11042, 用于确定上述两条路径中通过主节点设备的路 径的权重值为第一从节点设备与主节点设备之间路径的权重值, 与主节点设 备与第二从节点设备之间路径的权重值之和; 确定上述两条路径中通过备用 主节点设备的路径的权重值为第一从节点设备与备用主节点设备之间路径的 权重值, 与备用主节点设备与第二从节点设备之间路径的权重值之和; 以及 确定上述两条路径中具有较小权重值的路径为主路径, 具有较大权重值的路 径为次路径;
其中, 主节点设备与第一从节点设备或第二从节点设备之间路径的权重 值是预先配置的, 或者主节点设备根据第一从节点设备或第二从节点设备的 编号分配的; 备用主节点设备与第一从节点设备或从节点设备之间路径的权 重值是预先配置的, 或者主节点设备根据第一从节点设备或第二从节点设备 的编号分配的。
本发明实施例对上述虚拟网络单元的内部路径设置不同的权重值, 可以 实现虚拟网络单元内部的负载均衡。
进一步地, 第一确定子模块 11042还可以当上述两条路径的权重值相同 时, 比较第一从节点设备与主节点设备之间路径的权重值与第一从节点设备 与备用主节点设备之间路径的权重值, 当第一从节点设备与主节点设备之间 路径的权重值较小时, 以通过主节点设备的路径为主路径, 以通过备用主节 点设备的路径为次路径; 当第一从节点设备与备用主节点设备之间路径的权 重值较小时, 以通过备用主节点设备的路径为主路径, 以通过主节点设备的 路径为次路径。
具体地, 生成模块 1105可以根据主路径、 上述主路径的权重值、 上述次 路径和上述次路径的权重值生成内部路由条目;保存模块 1102可以保存生成 模块 1105生成的内部路由条目。
本实施例中, 计算模块 1104可以包括:
第二计算子模块 11043 , 用于当拓朴类型为环形拓朴时, 为 VNE中的每 个节点设备,计算该节点设备到 VNE中除上述节点设备之外的其他所有节点 设备的两条路径;
第二确定子模块 11044, 用于确定上述两条路径中经过节点设备较少的 路径为主路径, 经过节点设备较多的路径为次路径, 并为主路径分配较小的 权重值, 为次路径分配较大的权重值; 或者, 确定上述两条路径的权重值为 上述两条路径上相邻两个节点设备之间路径上预先配置的权重值之和, 并确 定具有较小权重值的路径为主路径, 具有较大权重值的路径为次路径;
这时, 生成模块 1105 , 具体用于根据主路径、 主路径的权重值、 次路径 和次路径的权重值生成内部路由条目。
本实施例中, 计算模块 1104可以包括:
第三计算子模块 11045 , 用于当拓朴类型为半环形拓朴时, 为 VNE中的 每个从节点设备计算从节点设备到主节点设备, 以及从节点设备到备用主节 点设备的路径;
第三确定子模块 11046, 用于比较从节点设备到主节点设备的路径与从 节点设备到备用主节点设备的路径所经过的节点设备数, 并为经过节点设备 数较多的路径设置较大的权重值, 为经过节点设备数较少的路径设置较小的 权重值; 或者, 第三确定子模块 11046, 用于确定从节点设备到主节点设备 的路径的权重值为从节点设备到主节点设备的路径上相邻两个节点设备之间 路径上预先配置的权重值之和, 确定从节点设备到备用主节点设备的路径的 权重值为从节点设备到备用主节点设备的路径上相邻两个节点设备之间路径 上预先配置的权重值之和;
进一步地, 第三确定子模块 11046还可以确定具有较小权重值的路径为 主路径, 具有较大权重值的路径为次路径;
生成模块 1105可以根据主路径、 主路径的权重值、 次路径和次路径的权 重值生成内部路由条目。
上述主节点设备可以实现在双归树形、环形和半环形等多种拓朴结构下, 对 VNE的内部路径进行计算, 进而可以实现简单有效的计算 VNE的内部路 径中的主路径和次路径。
本领域技术人员可以理解附图只是一个优选实施例的示意图, 附图中 的模块或流程并不一定是实施本发明所必须的。
本领域技术人员可以理解实施例中的装置中的模块可以按照实施例描述 进行分布于实施例的装置中, 也可以进行相应变化位于不同于本实施例的一 个或多个装置中。 上述实施例的模块可以合并为一个模块, 也可以进一步拆 分成多个子模块。
最后应说明的是: 以上实施例仅用以说明本发明的技术方案, 而非对 其限制; 尽管参照前述实施例对本发明进行了详细的说明, 本领域的普通 技术人员应当理解: 其依然可以对前述各实施例所记载的技术方案进行修 改, 或者对其中部分技术特征进行等同替换; 而这些修改或者替换, 并不 使相应技术方案的本质脱离本发明各实施例技术方案的精神和范围。

Claims

权利要求
1、 一种虚拟网络单元中的路由计算方法, 其特征在于, 包括:
虚拟网络单元中的主节点接收并保存虚拟网络单元中除所述主节点之外 的节点上报的拓朴信息;
所述主节点根据所述虚拟网络单元的拓朴类型对所述主节点本地保存的 所有拓朴信息进行拓朴校验, 所述保存的拓朴信息包括所述主节点的拓朴信 息和所述虚拟网络单元中除所述主节点之外的节点上报的拓朴信息;
在所述拓朴校验通过之后, 所述主节点根据所述拓朴类型和所述拓朴信 息计算所述虚拟网络单元的内部路径和所述内部路径的权重值;
所述主节点根据所述内部路径和所述内部路径的权重值生成并保存内部 路由条目, 并将所述内部路由条目发送给所述虚拟网络单元中除所述主节点 之外的节点。
2、 根据权利要求 1所述的方法, 其特征在于, 所述虚拟网络单元中的主 节点接收并保存虚拟网络单元中除所述主节点之外的节点上报的拓朴信息之 前, 还包括:
所述主节点以广播方式发送虚拟网络单元的拓朴收集请求消息, 所述拓 朴收集请求消息携带所述主节点的信息、 所述虚拟网络单元的标识、 名称和 拓朴类型;
所述虚拟网络单元中的主节点接收并保存虚拟网络单元中除所述主节点 之外的节点上报的拓 4卜信息包括:
所述主节点接收所述虚拟网络单元中除所述主节点之外的节点发送的拓 朴收集响应消息, 所述拓朴收集响应消息携带所述拓朴信息; 所述拓朴收集 响应消息是所述虚拟网络单元中除所述主节点之外的节点接收到所述拓朴收 集请求消息之后, 确定所述拓朴收集请求消息中携带的虚拟网络单元的标识 对应的虚拟网络单元为自身归属的虚拟网络单元之后,发送给所述主节点的; 所述主节点保存所述拓朴信息, 并将所述虚拟网络单元中的所有节点纳 入所述虚拟网络单元集群管理中。
3、 根据权利要求 1所述的方法, 其特征在于, 所述虚拟网络单元中的主 节点接收并保存虚拟网络单元中除所述主节点之外的节点上报的拓朴信息包 括:
所述主节点接收所述虚拟网络单元中除所述主节点之外的节点在获知所 述主节点的位置后上报的拓朴信息;
所述主节点保存所述拓朴信息, 并将所述虚拟网络单元中的所有节点纳 入所述虚拟网络单元集群管理中。
4、 根据权利要求 1所述的方法, 其特征在于, 还包括:
在所述拓朴校验通过之后, 所述主节点在所述虚拟网络单元中定时或周 期性发送拓朴报文, 所述拓朴报文携带所述虚拟网络单元中各节点的拓朴信 息; 以便接收到所述拓朴报文的节点确定自身的拓朴信息与所述拓朴报文中 携带的拓朴信息不同之后, 将自身的拓朴信息发送给所述主节点。
5、 根据权利要求 1所述的方法, 其特征在于, 还包括:
在所述拓朴校验通过之后, 如果所述虚拟网络单元中新增节点, 则所述 主节点接收并保存所述虚拟网络单元中的新增节点上报的拓朴信息, 重新根 据所述拓朴类型对所述主节点本地保存的所有拓朴信息进行拓朴校验;或者, 所述主节点接收并保存所述虚拟网络单元中与新增节点相连的节点发送的拓 朴信息, 重新根据所述拓朴类型对所述主节点本地保存的所有拓朴信息进行 拓朴校验, 所述虚拟网络单元中与新增节点相连的节点发送的拓朴信息是所 述与新增节点相连的节点接收到所述主节点发送的拓朴报文之后, 确定自身 的拓朴信息与所述拓朴报文中携带的拓朴信息不同之后, 发送给所述主节点 的; 或者, 所述主节点接收并保存新增节点在接收到所述主节点发送的拓朴 报文之后上报的拓朴信息, 重新根据所述拓朴类型对所述主节点本地保存的 所有拓 4卜信息进行拓 4卜校验。
6、 根据权利要求 1所述的方法, 其特征在于, 还包括:
在所述拓朴校验通过之后, 如果所述虚拟网络单元中去除节点, 则所述 主节点接收并保存所述虚拟网络单元中与去除的节点直连的节点上报的变化 后的拓朴信息, 重新根据所述拓朴类型对所述主节点本地保存的所有拓朴信 息进行拓朴校验; 所述变化后的拓朴信息由所述与去除的节点直连的节点主 动上报给所述主节点, 或者所述变化后的拓朴信息是所述与去除的节点直连 的节点接收到所述主节点发送的拓朴报文之后, 确定自身的拓朴信息与所述 拓朴报文中携带的拓朴信息不同之后, 发送给所述主节点的。
7、 根据权利要求 1所述的方法, 其特征在于, 还包括:
在所述拓朴校验通过之后, 当所述虚拟网络单元中除所述主节点之外的 节点的控制平面发生故障, 而转发平面未发生故障时, 所述主节点通过心跳 机制发现所述节点发生故障之后, 所述主节点删除发生故障的节点的拓朴信 息, 重新接收并保存所述虚拟网络单元中除所述主节点和所述发生故障的节 点之外的节点上报的拓朴信息, 并根据所述拓朴类型对所述主节点本地保存 的所有拓 4卜信息进行拓 4卜校验。
8、 根据权利要求 1所述的方法, 其特征在于, 还包括:
在所述拓朴校验通过之后, 当所述虚拟网络单元中除所述主节点之外的 节点的数据平面发生故障时, 所述主节点接收并保存发生故障的节点上报的 变化后的拓朴信息, 重新根据所述拓朴类型对所述主节点本地保存的所有拓 朴信息进行拓朴校验。
9、 根据权利要求 1所述的方法, 其特征在于, 还包括:
在所述拓朴校验通过之后, 当所述虚拟网络单元中的内部链路发生故障 时, 所述主节点接收并保存发生故障链路上的节点通过备用路径上报的变化 后的拓朴信息, 以及接收并保存与所述发生故障链路上的节点相连的节点上 报的变化后的拓朴信息, 并重新根据所述拓朴类型对所述主节点本地保存的 所有拓 4卜信息进行拓 4卜校验。
10、 根据权利要求 1-9任意一项所述的方法, 其特征在于, 所述拓朴信 息包括所述虚拟网络单元中的节点的标识符、所述节点所在设备的基本信息、 外连接口信息和内连接口信息。
11、 根据权利要求 10所述的方法, 其特征在于, 所述主节点根据所述虚 拟网络单元的拓朴类型对所述主节点本地保存的所有拓朴信息进行拓朴校验 包括:
当所述拓朴类型为双归树形拓朴时, 所述主节点根据所述主节点本地保 存的所有拓朴信息确定所述双归树形拓朴的校验通过条件是否已满足;
所述拓朴校验通过包括: 所述双归树形拓朴的校验通过条件已满足; 所述双归树形拓朴的校验通过条件包括: 所述虚拟网络单元内部所有节 点均至少有两个状态可用的内连接口; 并且所述主节点与备用主节点通过内 连接口直连,所述主节点与所述备用主节点之间的内连接口和链路状态正常; 并且所述虚拟网络单元中所有从节点有且只有两个内连接口, 分别连接到所 述主节点和所述备用主节点上; 并且所述虚拟网络单元中已上报自身拓朴信 息的从节点的个数与预先规划的从节点的个数相同。
12、 根据权利要求 10所述的方法, 其特征在于, 所述主节点根据所述虚 拟网络单元的拓朴类型对所述主节点本地保存的所有拓朴信息进行拓朴校验 包括:
当所述拓朴类型为环形拓朴时, 所述主节点根据所述主节点本地保存的 所有拓朴信息确定所述环形拓朴的校验通过条件是否已满足;
所述拓朴校验通过包括: 所述环形拓朴的校验通过条件已满足; 所述环形拓朴的校验通过条件包括: 所述虚拟网络单元中的每个节点都 有两个状态正常的内连接口; 并且所述主节点与备用主节点通过内连接口直 连; 并且所述虚拟网络单元中的节点两两相连, 从所述主节点的一个内连接 口开始, 通过所述虚拟网络单元中除所述主节点之外的各节点能够到达所述 主节点的另一个内连接口; 并且从所述主节点的两个内连接口出发形成的两 条环状路径上, 经过节点的顺序正好相反。
13、 根据权利要求 10所述的方法, 其特征在于, 所述主节点根据所述虚 拟网络单元的拓朴类型对所述主节点本地保存的所有拓朴信息进行拓朴校验 包括:
当所述拓朴类型为半环形拓朴时, 所述主节点根据所述主节点本地保存 的所有拓朴信息确定所述环形拓朴的校验通过条件是否已满足;
所述拓朴校验通过包括: 所述半环形拓朴的校验通过条件已满足; 所述半环形拓朴的校验通过条件包括: 所述虚拟网络单元中除所述主节 点和备用主节点之外的每个节点均有至少两个状态正常的内连接口; 并且, 所述主节点和所述备用主节点均有至少一个内连接口 , 且所述主节点与所述 备用主节点之间没有直连的内部链路; 并且, 所述虚拟网络单元中除所述主 节点与所述备用主节点之外的节点两两相连, 从所述主节点的一个内连接口 开始, 通过所述虚拟网络单元中除所述主节点与所述备用主节点之外的各节 点, 能够到达所述备用主节点的一个状态正常的内连接口; 并且, 从所述主 节点的一个内连接口出发形成的半环状路径通过了所述虚拟网络单元中除所 述主节点之外的所有节点, 所述半环状路径的最后一个节点为所述备用主节 点。
14、 根据权利要求 10所述的方法, 其特征在于, 所述主节点根据所述拓 朴类型和所述拓朴信息计算所述虚拟网络单元的内部路径和所述内部路径的 权重值包括:
当所述拓朴类型为双归树形拓朴时, 所述主节点计算所述虚拟网络单元 中每两个从节点之间的两条路径, 以及为所述虚拟网络单元中的每个从节点 计算所述从节点到所述主节点, 以及所述从节点到所述备用主节点的路径; 所述主节点确定所述两条路径中通过所述主节点的路径的权重值为第一 从节点与所述主节点之间路径的权重值, 与所述主节点与第二从节点之间路 径的权重值之和; 确定所述两条路径中通过所述备用主节点的路径的权重值 为所述第一从节点与所述备用主节点之间路径的权重值, 与所述备用主节点 与所述第二从节点之间路径的权重值之和; 以及确定所述两条路径中具有较 'J、权重值的路径为主路径, 具有较大权重值的路径为次路径;
所述主节点与所述第一从节点或所述第二从节点之间路径的权重值是预 先配置的, 或者所述主节点根据所述第一从节点或所述第二从节点的编号分 配的; 所述备用主节点与所述第一从节点或第二从节点之间路径的权重值是 预先配置的, 或者所述主节点根据所述第一从节点或所述第二从节点的编号 分配的。
15、 根据权利要求 14所述的方法, 其特征在于, 还包括:
当所述两条路径的权重值相同时 , 所述主节点比较所述第一从节点与所 述主节点之间路径的权重值与所述第一从节点与所述备用主节点之间路径的 权重值, 当所述第一从节点与所述主节点之间路径的权重值较小时, 所述主 节点以所述通过所述主节点的路径为主路径, 以所述通过所述备用主节点的 路径为次路径; 当所述第一从节点与所述备用主节点之间路径的权重值较小 时, 所述主节点以所述通过所述备用主节点的路径为主路径, 以所述通过所 述主节点的路径为次路径。
16、 根据权利要求 15所述的方法, 其特征在于, 所述主节点根据所述内 部路径和所述内部路径的权重值生成并保存内部路由条目包括:
所述主节点根据所述主路径、 所述主路径的权重值、 所述次路径和所述 次路径的权重值生成并保存内部路由条目。
17、 根据权利要求 10所述的方法, 其特征在于, 所述主节点根据所述拓 朴类型和所述拓朴信息计算所述虚拟网络单元的内部路径和所述内部路径的 权重值包括:
当所述拓朴类型为环形拓朴时 , 所述主节点为所述虚拟网络单元中的每 个节点, 计算所述节点到所述虚拟网络单元中除所述节点之外的其他所有节 点的两条路径; 所述主节点确定所述两条路径中经过节点较少的路径为主路径, 经过节 点较多的路径为次路径, 并为所述主路径分配较小的权重值, 为所述次路径 分配较大的权重值; 或者, 所述主节点确定所述两条路径的权重值为所述两 条路径上相邻两个节点之间路径上预先配置的权重值之和, 并确定具有较小 权重值的路径为主路径, 具有较大权重值的路径为次路径;
所述主节点根据所述内部路径和所述内部路径的权重值生成并保存内部 路由条目包括:
所述主节点根据所述主路径、 所述主路径的权重值、 所述次路径和所述 次路径的权重值生成并保存内部路由条目。
18、 根据权利要求 10所述的方法, 其特征在于, 所述主节点根据所述拓 朴类型和所述拓朴信息计算所述虚拟网络单元的内部路径和所述内部路径的 权重值包括:
当所述拓朴类型为半环形拓朴时, 所述主节点为所述虚拟网络单元中的 每个从节点计算所述从节点到所述主节点 , 以及所述从节点到所述备用主节 点的路径;
所述主节点比较所述从节点到所述主节点的路径与所述从节点到所述备 用主节点的路径所经过的节点数, 并为经过节点数较多的路径设置较大的权 重值, 为经过节点数较少的路径设置较小的权重值; 或者, 所述主节点确定 所述从节点到所述主节点的路径的权重值为所述从节点到所述主节点的路径 上相邻两个节点之间路径上预先配置的权重值之和, 确定所述从节点到所述 备用主节点的路径的权重值为所述从节点到所述备用主节点的路径上相邻两 个节点之间路径上预先配置的权重值之和;
所述主节点确定具有较小权重值的路径为主路径, 具有较大权重值的路 径为次路径;
所述主节点根据所述内部路径和所述内部路径的权重值生成并保存内部 路由条目包括: 所述主节点根据所述主路径、 所述主路径的权重值、 所述次路径和所述 次路径的权重值生成并保存内部路由条目。
19、 一种虚拟网络单元中的主节点设备, 其特征在于, 包括:
接收模块, 用于接收虚拟网络单元中除所述主节点设备之外的节点设备 上报的拓朴信息;
保存模块, 用于保存所述接收模块接收的拓朴信息;
校验模块, 用于根据所述虚拟网络单元的拓朴类型对所述保存模块保存 的所有拓朴信息进行拓朴校验, 所述保存模块保存的所有拓朴信息包括所述 主节点设备的拓朴信息和所述接收模块接收的拓朴信息;
计算模块, 用于在所述拓朴校验通过之后, 根据所述拓朴类型和所述保 存模块保存的拓朴信息, 计算所述虚拟网络单元的内部路径和所述内部路径 的权重值;
生成模块, 用于根据所述计算模块计算的所述内部路径和所述内部路径 的权重值生成并保存内部路由条目;
发送模块, 用于将所述生成模块生成的内部路由条目发送给所述虚拟网 络单元中除所述主节点设备之外的节点设备。
20、 根据权利要求 19所述的设备, 其特征在于,
所述发送模块, 还用于以广播方式发送虚拟网络单元的拓朴收集请求消 息, 所述拓朴收集请求消息携带所述主节点设备的信息、 所述虚拟网络单元 的标识、 名称和拓朴类型;
所述接收模块, 具体用于接收所述虚拟网络单元中除所述主节点设备之 外的节点设备发送的拓朴收集响应消息 , 所述拓朴收集响应消息携带所述拓 朴信息; 所述拓朴收集响应消息是所述虚拟网络单元中除所述主节点设备之 外的节点设备接收到所述拓朴收集请求消息之后, 确定所述拓朴收集请求消 息中携带的虚拟网络单元的标识对应的虚拟网络单元为自身归属的虚拟网络 单元之后发送给所述主节点设备的; 所述保存模块, 具体用于保存所述拓朴信息, 并将所述虚拟网络单元中 的所有节点设备纳入所述虚拟网络单元集群管理中。
21、 根据权利要求 19所述的设备, 其特征在于,
所述接收模块, 具体用于接收所述虚拟网络单元中除所述主节点设备之 外的节点设备在获知所述主节点设备的位置后上报的拓朴信息;
所述保存模块, 具体用于保存所述拓朴信息, 并将所述虚拟网络单元中 的所有节点设备纳入所述虚拟网络单元集群管理中。
22、 根据权利要求 19所述的设备, 其特征在于,
所述发送模块, 还用于在所述拓朴校验通过之后, 在所述虚拟网络单元 中定时或周期性发送拓朴报文, 所述拓朴报文携带所述虚拟网络单元中各节 点设备的拓朴信息; 以便接收到所述拓朴报文的节点设备确定自身的拓朴信 息与所述拓朴报文中携带的拓朴信息不同之后, 将自身的拓朴信息发送给所 述主节点设备。
23、 根据权利要求 19-22任意一项所述的设备, 其特征在于,
所述校验模块, 具体用于根据所述保存模块保存的拓朴信息确定所述拓 朴类型对应的校验通过条件是否已满足。
24、 根据权利要求 19-22任意一项所述的设备, 其特征在于, 所述计算 模块包括:
第一计算子模块, 用于当所述拓朴类型为双归树形拓朴时, 计算所述虚 拟网络单元中每两个从节点设备之间的两条路径, 以及为所述虚拟网络单元 中的每个从节点设备计算所述从节点设备到所述主节点设备, 以及所述从节 点设备到所述备用主节点设备的路径;
第一确定子模块, 用于确定所述两条路径中通过所述主节点设备的路径 的权重值为第一从节点设备与所述主节点设备之间路径的权重值, 与所述主 节点设备与第二从节点设备之间路径的权重值之和; 确定所述两条路径中通 过所述备用主节点设备的路径的权重值为所述第一从节点设备与所述备用主 节点设备之间路径的权重值, 与所述备用主节点设备与所述第二从节点设备 之间路径的权重值之和; 以及确定所述两条路径中具有较小权重值的路径为 主路径, 具有较大权重值的路径为次路径;
其中, 所述主节点设备与所述第一从节点设备或所述第二从节点设备之 间路径的权重值是预先配置的, 或者所述主节点设备根据所述第一从节点设 备或所述第二从节点设备的编号分配的; 所述备用主节点设备与所述第一从 节点设备或第二从节点设备之间路径的权重值是预先配置的, 或者所述主节 点设备根据所述第一从节点设备或所述第二从节点设备的编号分配的。
25、 根据权利要求 24所述的设备, 其特征在于,
所述第一确定子模块, 还用于当所述两条路径的权重值相同时, 比较所 述第一从节点设备与所述主节点设备之间路径的权重值与所述第一从节点设 备与所述备用主节点设备之间路径的权重值, 当所述第一从节点设备与所述 主节点设备之间路径的权重值较小时, 以所述通过所述主节点设备的路径为 主路径, 以所述通过所述备用主节点设备的路径为次路径; 当所述第一从节 点设备与所述备用主节点设备之间路径的权重值较小时, 以所述通过所述备 用主节点设备的路径为主路径,以所述通过所述主节点设备的路径为次路径。
26、 根据权利要求 25所述的设备, 其特征在于,
所述生成模块, 具体用于根据所述主路径、 所述主路径的权重值、 所述 次路径和所述次路径的权重值生成内部路由条目;
所述保存模块, 具体用于保存所述生成模块生成的内部路由条目。
27、 根据权利要求 19-22任意一项所述的设备, 其特征在于, 所述计算 模块包括:
第二计算子模块, 用于当所述拓朴类型为环形拓朴时, 为所述虚拟网络 单元中的每个节点设备, 计算所述节点设备到所述虚拟网络单元中除所述节 点设备之外的其他所有节点设备的两条路径;
第二确定子模块, 用于确定所述两条路径中经过节点设备较少的路径为 主路径, 经过节点设备较多的路径为次路径, 并为所述主路径分配较小的权 重值, 为所述次路径分配较大的权重值; 或者, 确定所述两条路径的权重值 为所述两条路径上相邻两个节点设备之间路径上预先配置的权重值之和, 并 确定具有较小权重值的路径为主路径, 具有较大权重值的路径为次路径; 所述生成模块, 具体用于根据所述主路径、 所述主路径的权重值、 所述 次路径和所述次路径的权重值生成内部路由条目。
28、 根据权利要求 19-22任意一项所述的设备, 其特征在于, 所述计算 模块包括:
第三计算子模块, 用于当所述拓朴类型为半环形拓朴时, 为所述虚拟网 络单元中的每个从节点设备计算所述从节点设备到所述主节点设备, 以及所 述从节点设备到所述备用主节点设备的路径;
第三确定子模块, 用于比较所述从节点设备到所述主节点设备的路径与 所述从节点设备到所述备用主节点设备的路径所经过的节点设备数, 并为经 过节点设备数较多的路径设置较大的权重值, 为经过节点设备数较少的路径 设置较小的权重值; 或者, 第三确定子模块, 用于确定所述从节点设备到所 述主节点设备的路径的权重值为所述从节点设备到所述主节点设备的路径上 相邻两个节点设备之间路径上预先配置的权重值之和, 确定所述从节点设备 到所述备用主节点设备的路径的权重值为所述从节点设备到所述备用主节点 设备的路径上相邻两个节点设备之间路径上预先配置的权重值之和;
第三确定子模块, 还用于确定具有较小权重值的路径为主路径, 具有较 大权重值的路径为次路径;
所述生成模块, 具体用于根据所述主路径、 所述主路径的权重值、 所 述次路径和所述次路径的权重值生成内部路由条目。
PCT/CN2011/074664 2011-05-25 2011-05-25 虚拟网络单元中的路由计算方法和主节点设备 WO2011144082A2 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
PCT/CN2011/074664 WO2011144082A2 (zh) 2011-05-25 2011-05-25 虚拟网络单元中的路由计算方法和主节点设备
EP20110782999 EP2621130B1 (en) 2011-05-25 2011-05-25 Route calculation method and master node device in virtual network element
CN2011800006322A CN102204188B (zh) 2011-05-25 2011-05-25 虚拟网络单元中的路由计算方法和主节点设备
US13/930,856 US9077608B2 (en) 2011-05-25 2013-06-28 Route calculation method and master node device in virtual network element

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2011/074664 WO2011144082A2 (zh) 2011-05-25 2011-05-25 虚拟网络单元中的路由计算方法和主节点设备

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US13/930,856 Continuation US9077608B2 (en) 2011-05-25 2013-06-28 Route calculation method and master node device in virtual network element

Publications (2)

Publication Number Publication Date
WO2011144082A2 true WO2011144082A2 (zh) 2011-11-24
WO2011144082A3 WO2011144082A3 (zh) 2012-04-26

Family

ID=44662810

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2011/074664 WO2011144082A2 (zh) 2011-05-25 2011-05-25 虚拟网络单元中的路由计算方法和主节点设备

Country Status (4)

Country Link
US (1) US9077608B2 (zh)
EP (1) EP2621130B1 (zh)
CN (1) CN102204188B (zh)
WO (1) WO2011144082A2 (zh)

Families Citing this family (31)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9425970B2 (en) * 2012-05-21 2016-08-23 Nec Corporation Multicast communication method, communication node device and program
US9178807B1 (en) 2012-09-20 2015-11-03 Wiretap Ventures, LLC Controller for software defined networks
CN103129750B (zh) * 2013-01-25 2015-11-25 航天东方红卫星有限公司 基于状态感知的分离模块航天器系统和自组网方法
WO2014166065A1 (zh) * 2013-04-09 2014-10-16 华为技术有限公司 保护业务可靠性的方法、设备及网络虚拟化系统
CN103281747B (zh) * 2013-04-18 2015-09-23 上海桑锐电子科技有限公司 一种无线中继路径优选方法
CN104348732B (zh) * 2013-07-25 2018-09-07 华为技术有限公司 拓扑结构发现方法及装置
CN103441936B (zh) * 2013-08-22 2017-02-22 华为技术有限公司 转发邻接链路的发布方法
CN104579966B (zh) * 2013-10-14 2018-08-17 华为技术有限公司 转发表项生成的方法、转发节点和控制器
US9491102B1 (en) 2014-02-10 2016-11-08 Google Inc. Traffic load balancing in a multi-connect topology
WO2015121925A1 (ja) * 2014-02-12 2015-08-20 株式会社日立製作所 システム管理方法
CN103888333B (zh) * 2014-03-14 2017-06-20 中国科学院电子学研究所 合成孔径雷达控制接口网络
CN105391632B (zh) * 2014-09-04 2019-11-05 中兴通讯股份有限公司 路由选择方法及装置
CN106331047A (zh) * 2015-06-30 2017-01-11 中兴通讯股份有限公司 一种集群设备性能同步统计方法及系统
CN106470142B (zh) * 2015-08-14 2019-09-06 中国电信股份有限公司 虚拟网络拓扑自助生成方法和系统
US9898378B2 (en) 2015-11-10 2018-02-20 International Business Machines Corporation Smart selection of a storage module to be excluded
US10404571B2 (en) * 2015-12-29 2019-09-03 Futurewei Technologies, Inc. Communication among network controllers
US9967053B2 (en) * 2016-05-31 2018-05-08 Fujitsu Limited Shortest minimum regeneration path search in networks
TWI633773B (zh) * 2016-10-19 2018-08-21 神雲科技股份有限公司 電腦叢集系統
TW201822579A (zh) * 2016-11-30 2018-06-16 蔡奇雄 一種偵測空間物體與溫度的無線裝置與方法
CN108881014B (zh) * 2017-05-12 2021-10-01 北京猎户星空科技有限公司 一种数据发送系统及方法
CN109104348B (zh) * 2017-06-21 2020-09-15 比亚迪股份有限公司 基于CANopen协议的列车网络数据传输方法、系统及其装置
CN107342942B (zh) * 2017-07-05 2020-09-18 苏州浪潮智能科技有限公司 一种多节点控制器集群的信息传输方法及装置
US10367654B2 (en) * 2017-07-14 2019-07-30 Fujitsu Limited Network design method for ethernet ring protection switching
WO2019111154A1 (en) * 2017-12-04 2019-06-13 Telefonaktiebolaget Lm Ericsson (Publ) Network link verification
WO2021148015A1 (zh) 2020-01-23 2021-07-29 飞诺门阵(北京)科技有限公司 一种计算机系统
CN110932920B (zh) * 2020-01-23 2021-02-02 飞诺门阵(北京)科技有限公司 一种网络拓扑结构
CN111327504A (zh) * 2020-02-26 2020-06-23 深信服科技股份有限公司 虚拟局域网环路检测方法、装置、设备及可读存储介质
CN111404779B (zh) * 2020-03-12 2022-02-25 深圳市吉祥腾达科技有限公司 一种路由器节点角色切换的测试方法、装置及系统
US11595292B2 (en) * 2021-02-24 2023-02-28 Nokia Solutions And Networks Oy Dynamic node cluster discovery in an unknown topology graph
CN113115133B (zh) * 2021-03-10 2022-08-30 烽火通信科技股份有限公司 一种扩展子框拓扑发现方法及系统
CN117687800B (zh) * 2024-02-02 2024-05-03 山东海量信息技术研究院 一种跨域分布式计算方法、系统、存储介质和电子设备

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7502315B2 (en) * 1998-12-21 2009-03-10 At&T Intellectual Property Ii, L.P. Method for hybrid network planning
US6473408B1 (en) * 1999-05-19 2002-10-29 3Com Corporation Building a hierarchy in an asynchronous transfer mode PNNI network utilizing proxy SVCC-based RCC entities
US6973023B1 (en) * 2000-12-30 2005-12-06 Cisco Technology, Inc. Method for routing information over a network employing centralized control
CN100459534C (zh) * 2002-10-07 2009-02-04 日本电信电话株式会社 分层网络节点及通过该节点构成的网络、节点和分层网络
US7792987B1 (en) * 2003-04-21 2010-09-07 Juniper Networks, Inc. Supporting virtual private networks using a first network topology for forwarding and a subset of the first network topology or a smaller topology for signaling
US8200789B2 (en) * 2004-10-12 2012-06-12 International Business Machines Corporation Method, system and program product for automated topology formation in dynamic distributed environments
CN100417115C (zh) * 2005-06-21 2008-09-03 中兴通讯股份有限公司 一种实现以太网业务的自动配置方法
KR20080084163A (ko) * 2007-03-15 2008-09-19 삼성전자주식회사 시스템 이중화된 네트워크상에서 Nonstop 패킷포워딩 방법 및 라우터
CN101110832A (zh) * 2007-08-20 2008-01-23 华为技术有限公司 一种发送路由的方法和路由设备
CN101621747B (zh) * 2009-08-12 2012-07-11 广州海格通信集团股份有限公司 一种无线mesh网络的可靠组播传输方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
None

Also Published As

Publication number Publication date
EP2621130A2 (en) 2013-07-31
WO2011144082A3 (zh) 2012-04-26
US20130286893A1 (en) 2013-10-31
US9077608B2 (en) 2015-07-07
EP2621130A4 (en) 2013-12-04
EP2621130B1 (en) 2015-02-18
CN102204188A (zh) 2011-09-28
CN102204188B (zh) 2013-03-20

Similar Documents

Publication Publication Date Title
WO2011144082A2 (zh) 虚拟网络单元中的路由计算方法和主节点设备
JP6369698B2 (ja) トラフィック切り替え方法、デバイス、およびシステム
US9807017B2 (en) Multicast traffic load balancing over virtual link aggregation
US10462049B2 (en) Spanning tree in fabric switches
US10924409B2 (en) Method for implementing load balancing, apparatus, and network system
JP5488979B2 (ja) コンピュータシステム、コントローラ、スイッチ、及び通信方法
WO2015000362A1 (zh) 一种业务节点配置方法、业务节点池注册器及系统
US20170078150A1 (en) Ip-based interconnection of switches with a logical chassis
JP2015515809A (ja) 仮想ファブリックリンク障害復旧のためのシステムおよび方法
EP3883286A1 (en) Data processing method, controller and forwarding device
US20220174026A1 (en) Efficient convergence in network events
WO2020114017A1 (zh) 数据中心流量互通方法、装置、设备及存储介质
WO2014059867A1 (zh) 链路聚合组中流量路径的协商方法及装置
US20230010837A1 (en) Fault diagnosis method and apparatus thereof
WO2020135339A1 (zh) 一种网络路径收敛的方法以及相关设备
JP6109954B2 (ja) 仮想シャーシシステムにおけるパススルーモードのためのシステムおよび方法
CN113872868A (zh) 通知消息传输方法、装置及系统、存储介质
JP2022044029A (ja) データ伝送方法、装置、およびネットワークデバイス
CN116074236A (zh) 报文转发方法及装置
CN116094987A (zh) 转发路径的确定方法及装置
EP4395246A1 (en) Data forwarding method and system, electronic device, and storage medium
WO2012155351A1 (zh) 一种生成隧道转发表项的方法和网络设备
US20230336484A1 (en) Load sharing method, root node device, leaf node device, and system
WO2021143213A1 (zh) 跨设备链路聚合方法及设备
WO2019061520A1 (zh) 切换路径的方法及装置

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 201180000632.2

Country of ref document: CN

WWE Wipo information: entry into national phase

Ref document number: 2011782999

Country of ref document: EP

NENP Non-entry into the national phase

Ref country code: DE