238results about How to "Realize load sharing" patented technology

The invention discloses an implementation method for an SDN Controller Pool. The resource pool is composed of a plurality of SDN controllers, wherein each controller in the pool is in the load-sharing working state, if faults happen to any controller, fault controller can be taken over by the other controllers in a lossless mode, synchronous states and dynamic data are sent to other nodes through an elected main node, and lossless taking over can be conducted when node faults occur. The implementation method mainly includes the step of resource pool construction, the step of data synchronization, the step of SDN switch connection establishment, the step of fault detection and the step of fault taking over.

The invention discloses a load sharing method of data center communication equipment, which comprises steps as follows: data center communication equipment regularly reads flow conditions of port group load balancing to acquire occupation conditions of each port; and a forwarding table and a Hash result modulo value in the message forwarding process are adjusted according to the occupation conditions. The invention also discloses a load sharing device of data center communication equipment. By using the scheme disclosed by the invention, static/dynamic combined load sharing of the data center communication equipment can be implemented, so that the ports can carry balanced loads and satisfy the requirements for linear speed.

The embodiment of the invention provides a communication method based on tunnel binding and network devices. The method includes the steps that service messages are transmitted between the first network device and the second network device through a first binding tunnel, wherein the first binding tunnel comprises a first tunnel and a second tunnel; when the first tunnel fails or transmission quality does not meet the preset conditions, service messages are transmitted between the first network device and the second network device through a second binding tunnel, wherein the second binding tunnel comprises a standby tunnel of the first tunnel and a second tunnel. Thus, when a certain tunnel of an existing binding tunnel fails, another binding tunnel is established through a standby tunnel of the failed tunnel and a fail-free tunnel, and large-bandwidth service of tunnel binding continues to be provided. Therefore, compared with the prior art, the reliability of large-bandwidth communication can be improved, and user satisfaction is improved.

The invention discloses a method and a device for realizing multicast load sharing. The method comprises the following steps: determining a corresponding relation between a router and a multicast group set, wherein the multicast group set comprises at least one multicast group; selecting the router corresponding to the set of the multicast group as a designated router; and processing service of the multicast group by the designated router. The device and the method provide a load sharing solution based on DR for routers of a receiving terminal and a multicast source terminal, and achieve balance load of service of the multicast groups among the multiple routers so as to prevent the problem that single DR is over loaded.

This invention discloses one terminal device, which comprises net card match unit, fault test unit and multiple card set unit connected to fault test unit. The invention also discloses one network load sharing system with multiple terminal device comprising multi-terminal device and multiple routers and broadcast network. The terminal device is connected to routers through network card connecting to wide network. This invention also discloses one multi-terminal device network loading method, which comprises steps of setting address information and then sending data by terminal device through multiple routers.

The invention discloses a domain split network and a method for obtaining a network topology map of the domain split network, belonging to the technical field of packet data networks. The network comprises at least two control management domains, wherein each control management domain comprises at least one router and at least one control management device; a control management channel is established between the router in each control management domain and the control management device dominating the router; and full connection and information synchronization are carried out between each two control management devices. The method comprises the following steps: the control management device of each domain respectively collects link state advertisement information sent by the router in the administering domain; synchronization is carried out on the collected link state advertisement information of the administering domain among the control management device of every domain to obtain the link state advertisement information of a whole network; and the network topology map of the whole network is calculated according to the link state advertisement information of the whole network. The invention realizes load share of each control management device and reliable guarantee after single-point failure.

The application discloses an opening and dormancy method for small cells. The method comprises an establishing process of double connections of a control plane and a user plane which are completely separated: control plane connections are established between UE and a macro base station, and data plane connections are established between the UE and small cell base stations; a current load of the macro base station and current loads of the opened small cell base stations are calculated by the macro base station and the opened small cell base stations, and, when the current load of the certain base station is greater than a set first threshold and exceeds a set first time length, the certain base station is configured with the accessed UE and performs an adjacent region finding measurement; the corresponding UE determines whether adjacent regions of a cell where the corresponding base station is located are at opening or dormancy states; and the corresponding UE determines the adjacent regions are at the dormancy states according to a find reference signal sent by the dormant adjacent region. The macro base station obtains state information of the various adjacent regions of all the UE of the corresponding base stations, a dormant small cell which unloads flow most effectively is selected and then is opened, the data plane of the UE of the corresponding small cell of an overload part is connected and transferred to the opened dormant small cell, the control panel is remained in the macro base station, and the double connection that the control plane and the user plane are separated is established. Through application of the method, load sharing can be achieved, a switching frequency is reduced at the same time, and a signaling expense is reduced.

The invention discloses a network server access method which is applied to an L2TP virtual private network (VPN) which comprises an access concentrator LAC and three or more network servers LNS. The method comprises the following steps: recording number of sessions carried by an L2TP tunnel of each LNS, and when receiving an L2TP connection establish request message sent by the LAC, according to the recorded number of sessions carried by the L2TP tunnel of each LNS, distributing LNS. According to the method, under the condition of not depending on a service provider, load sharing of L2TP LNS is simply and effectively realized.

The disclosed route selection method comprises: router takes all of the MB address, MB mask and relative next-jump router address as input, calculates with Hash algorithm, and selects the maximal result from multiple next-jump router addresses. This invention shares the loading of MB data flow.

The invention discloses an IGMP message transmitting method comprising that an IGMP agent device receives IGMP adding messages from a host; the IGMP agent device obtains the multicast source address corresponding to the IGMP adding messages, searches a uni-cast routing table of itself according to the multicast source address to obtain an outgoing interface of the next hop; if the outgoing interface is an uplink interface of IGMP, the IGMP agent device sends out the IGMP adding messages through the outgoing interface; if the outgoing interface is not an uplink interface of IGMP, the IGMP agent device selects an interface from the IGMP uplink interfaces of itself and sends out the IGMP adding messages through the selected interface. The invention also discloses an IGMP agent device. The invention avoids rounding of multicast flows and realizes load-sharing between IGMP uplink interfaces.

The invention discloses a gateway migration processing method and a gateway migration processing device based on a software defined network, wherein the gateway migration processing method is applied on a first virtual extensible local area network IP gateway (VXLAN IP GW) to be migrated in at least two VXLAN IP GWs. The gateway migration processing method comprises the steps of reducing the priority of a router which carries IP address of a VXLAN tunneling end point (VTEP) and transmitting the priority to the VTEP, reducing the priority of a static router to a virtual machine (VM) and transmitting the priority to exchange equipment of a non-virtual network; wherein before priority reduction of the router on the first VXLAN IP GW, the priorities of routers which carry the IP address of the VTEP on at least two VXLAN IP GWs are same, and the priorities of static routers from at least two VXLAN IP GWs to the VM are same; and when a controller monitors no data flow between the VM and the exchange equipment on the first VXLAN IP GW, performing a migration action.

The invention provides a method, a device and a system realizing load balancing, wherein the method for realizing the load balancing comprises the following steps: during a communication process between a Diameter client and a first Diameter server, sending partial request messages in all subsequently sent request messages to a second Diameter server according to a scheduled proportion by the Diameter client if determining that the load of the first Diameter server exceeds a first preset threshold value, and sending partial request messages in all subsequently sent request messages to a second Diameter server when the Diameter client determines that the load of the first Diameter server is oversized, thereby flexibly realizing load sharing and increasing availability of resources.

The invention provides a message processing method and system. The message processing method comprises the following steps: starting VFs (Virtual Forwarders) in an amount equal to that of cluster servers at each cluster server, and making same VFs at different cluster servers become backups for each other; configuring a same virtual IP (Internet Protocol) address for all the VFs, configuring a unique virtual MAC (Media Access Control) address for each VF, and configuring a state of each VF, wherein the state of each VF comprises an activated state or a listening state; after receiving an ARP (Address Resolution Protocol) request message by each VF in the activated state, judging whether to respond to the ARP request message by each VF; and when judging the self responses by the AFs, transmitting an ARP response message, wherein the ARP response message comprises a self virtual MAC address, so that when a receiver transmits a message to the cluster servers subsequently, a destination MAC address can be determined according to the virtual MAC address. The message processing method provided by the invention can solve the problem that the number of the servers is difficult to be expanded, and load sharing can be realized between the servers.

The invention provides a message transmission method, device and system. The method comprises the steps of sending a first message to a slave VPN (Virtual Private Network) server by employing a source address after the first message from an export device is received, thereby enabling the slave VPN server to unpack the first message; sending a second message to a management server by employing a destination address after the second message from the slave VPN server is received, thereby enabling the management server to carry out service processing on the second message; sending a third message to the slave VPN server by employing the destination address after the third message from the management server is received, thereby enabling the slave VPN server to package the third message; and sending a fourth message to the export device by employing the destination address after the fourth message from the slave VPN server is received, thereby enabling the export device to forward the fourth message to a video device. Through adoption of the technical scheme of the method, the device and the system, the processing performance of each VPN server can be improved, and the integrated performance of a video monitoring system can be improved.

The invention discloses a load sharing method and system of link aggregation, and relates to the link field of data communication technology. The system includes a control plane and a forwarding plane. Forwarding ports are assigned by the control plane for large-flow sessions, port adjustment is carried out according to average bandwidth ratio difference of the large-flow sessions of all the ports, and port forwarding weights are adjusted at the same time according to bandwidth utilization rates of all the ports of link aggregation. The received large-flow sessions are forwarded by the forwarding plane according to the assigned ports, and load balancing is carried out on remaining sessions according to the forwarding weights of all the ports. The system can avoid network congestion and load difference, improve resource utilization rates, and realize load balancing of a data center network.

The embodiment of the invention discloses a service carrying method, which may comprise the following steps: a communication terminal performs media negotiation with a communication opposite end at a first access network according to a service request; the communication terminal performs first media resource reservation at the first access network according to the media negotiation result; if the reserved resource of the first access network cannot meet a target resource, the communication terminal performs second media resource reservation at a second access network; after the communication terminal sends a message of completing first media resource reservation to the communication opposite end, the communication terminal receives a message of completing second media resource reservation sent by the communication opposite end after completing media resource reservation; the communication terminal builds service connection with the communication opposite end, and respectively carries media streams at the first access network and the second access network according to carrying modes of the media streams corresponding to target services in the message of completing first media resource reservation. The embodiment of the invention also provides a communication terminal, a control network element S-CSCF and a system, thus, network pressure can be shared effectively, quality of service of the services can be improved, and user experience is increased.

An embodiment of the invention discloses a link scheduling method and an apparatus thereof. The method comprises the following steps that a domain name system DNS agent server receives a first DNS request message carrying domain name information, and then determines a resource pool corresponding to the domain name information acquired from a load balancing LB equipment request, wherein the resource pool includes at least one link information; from the at least one link information included by the resource pool, target link information is determined and a first DNS response message carrying a target link IP address corresponding to the target link information is sent to a client. In the embodiment of the invention, load sharing of a link under the condition that the client uses a domain name system DNS agent is realized.

The invention discloses a method and a device for message transmission in a virtual private local area network service network. The method comprises the steps: when the VPLS network is arranged, each NPE in each VPN case satisfies the conditions: each NPE is reachable directly or indirectly through PW; aiming at each VPN case, each PE sends owned topological information and a bridge MAC address to other PE in the case and receives and records topological information and bridge MAC addresses sent from other PE; aiming at each VPN case, each PE generates a user MAC list in the VPN case according to the self-recorded topological information and the bridge MAC addresses of all the PE, an unicast minimum spanning tree with the PE as a root is computed, and an unicast transmission list in the VPN case is generated according to the unicast minimum spanning tree; and one of the PE is selected to serve as a root bridge, the root bridge serves as a root, a broadcast minimum spanning tree is computed, and a broadcast tress transmission list in the VPN case is generated according to the broadcast minimum spanning tree. According to the method and the device for message transmission in the VPLS network, complexity of VPLS network networking is reduced.