76 results about "Web cluster" patented technology

The invention discloses a database all-in-one machine capable of realizing high-speed storage. The database all-in-one machine is characterized by comprising a computing cluster, an Infiniband network cluster and a Fusion-io distributed storage, wherein the computing cluster comprises a plurality of computing servers which are used for computing; the computing servers form the computing cluster; the Infiniband network cluster comprises a plurality of Infiniband interchangers which are used for data exchange; the Fusion-io distributed storage comprises a plurality of Fusion-io storage servers. A Fusion-io flash memory product is introduced to a storage system of the database all-in-one machine, the I/O (Input/Output) bottleneck problem of the database all-in-one machine is effectively solved, the exponential order promotion is brought for performance indexes such as storage IOPS (Input/Output Operations Per Second) and throughput, and the database all-in-one machine capable of realizing high-speed storage is suitable for database application scenes with higher requirements such as high concurrency and high throughput.

The invention discloses a high-availability implementation method for a large-scale Web server cluster based on SOA, and the method comprises the steps: distributing a large number of web servers in a machine room, and enabling the servers to run the same or different web services; setting needed hardware and software resources for each web service, wherein the resources comprise a server list, a service progress, a hard disk, and an IP address pool; setting a monitoring resource and a monitoring strategy for each web service; carrying out the failure transfer for the web services when the resources are monitored to be abnormal; setting a web cluster main management node election strategy, electing a main management node during cluster start, repeatedly carrying out election when the main management node loses efficacy, and preventing a dedicated server from being used as a standby server. Moreover, because a plurality of start servers can be set for a failure switching group, the service is enabled to be higher in reliability, thereby reducing the investment cost, management complexity and maintenance cost for a user on the basis of guaranteeing the high-availability services for the user.

The invention relates to a RoboBee cluster control method and system. A RoboBee can sense the external environment via a sensor module, locate the position of itself to provide convenience for group cooperation via a positioning system, execute a present task via an execution mechanism, communicate with other RoboBees via a communication modules, and form a distributed computer network after communication and networking via a coordinated controller. A user communicates with a RoboBee cluster, including instruction issuing and task distribution, via a communication module in a cluster control system, and time standards of the RoboBee cluster are made uniform via a synchronizing module. The method and system have the advantages that the whole RoboBee cluster can form a self-organizing network to implement distributed cooperative computing even the single RoboBee is not high in intelligence and is low in cost and simple in control, a high-difficulty task that cannot be completed by the single RoboBee can be completed by the cluster, manual intervention is not needed in the task execution process, and the system and method have a dynamic self-healing capability and is high in reliability.

The invention discloses a web-cluster-oriented low energy consumption scheduling system and a method. The scheduling system comprises an energy-sensible scheduling module, a hysteresis controller and a manager, wherein the energy-sensible scheduling module, the hysteresis controller and the manager are connected sequentially. By the aid of the scheduling system, the resource utilization ratio in a web cluster system is greatly improved, and the energy consumption of the system is reduced. According to the method, a load connection number in the web cluster system serves as a sensitive factor, the system can schedule cluster nodes and virtual machine resources through monitoring of the load connection number only, the method has the advantages of being simple to operate, easy to implement, capable of achieving real-time scheduling and the like, and the scheduling method can improve the sensitivity of resource scheduling of the system.

The invention discloses a session keeping system under a load balancing cluster, and belongs to the technical field of computers. The session keeping system under the load balancing cluster comprisesa load balancing module positioned in a front-end load scheduler of the cluster, session management modules positioned in back-end nodes of the cluster, and session storage modules located on servicenodes of a storage cluster in the cluster, wherein the load balancing module is connected with a user; the load balancing module is connected with the session management module; the session managementmodule is connected with the session storage module; the load balancing module comprises request classification, session pasting, load calculation, threshold triggering and node selection; and the session storage module comprises a memory storage sub-module and a durable sub-module. Efficient and accurate Web cluster session communication under a high concurrency environment is realized. The session keeping system has the characteristics of high efficiency, high availability and high adaptability.

The invention belongs to the technical field of network clustering and networked control. The invention relates to a limited load consistency hash load balancing strategy based on virtual nodes. Taking the memory size of the cache node, the total number of CPU cores and the size of a disk as input indexes of an independent information data fluctuation weighting method; calculating a performance quantized value and a performance quantized value proportion of each cache node; obtaining the number of the virtual nodes of each cache node according to the set total number of the virtual nodes; theconsistent hash ring is equally divided into m sections of arcs; and generating a corresponding number of virtual nodes for each cache node by using a Random function, mapping the virtual nodes to each section of arc in proportion, updating the load capacity of the cluster cache agent and the load upper limit of each cache node by using a load balancer, selecting the virtual nodes on the consistent hash ring, and performing service by using the cache nodes with the virtual nodes. According to the invention, the problems of non-uniform load distribution and non-support of heterogeneous clustercache agents of an original strategy are solved, and the bandwidth occupancy rate of a backbone network and the response time of the cluster cache agents are effectively reduced.

The invention relates to a network resource allocation method and a network resource allocation system applied to online office, and network equipment. The network resource allocation method comprisesthe steps of: firstly, determining a first state parameter list with each office terminal and a second state parameter list with an office network cluster; secondly, determining a list conversion logic topology between each first state parameter list and the second state parameter list; thirdly, determining a mapping parameter corresponding to each target parameter and a communication state sequence corresponding to the mapping parameters according to an inverse list conversion logic topology corresponding to each parameter feature; and finally, adjusting a network resource occupation proportion corresponding to each office terminal according to the priority of a communication state sequence corresponding to each office terminal. Thus, the network resources can be allocated according to the priorities corresponding to different office terminals, the network fluency of the office terminals with higher priorities can be ensured, the normal operation of the office terminals with lower priorities can further be ensured, and the office efficiency is further improved.

A method implemented by a first controller in a software defined networking (SDN) network to monitor a health of a second controller, where the first controller and the second controller are part of a cluster of controllers, where controllers in the cluster communicate with switches over an in-band network, and where controllers in the cluster communicate with each other over an out-of-band network. The method includes sending a first control message to a switch over the in-band network in response to a determination that the second controller cannot be reached over the out-of-band network, where the first control message includes a health check request message and an instruction for the switch to send the health check request message to the second controller, and where the switch is a designated switch that has a control channel connection to the first controller and the second controller over the in-band network.

A method for dynamically provisioning computer components using a message platform communicatively coupled to a message generator is provided. The method includes receiving a first computer message, wherein the first computer message indicates that a computer component should be provisioned for a network cluster, routing the first computer message such that a first platform that is a customer of the first queue i) receives the first computer message and ii) automatically performs a first configuration operation on the computer component based on the first computer message, receiving, at the advanced message queue exchange, a second computer message from the first platform, and routing the second computer message such that a second platform that is a customer of the second queue i) receives the second computer message and ii) automatically performs a second configuration operation on the computer component based on the second computer message.

The invention discloses a fault positioning visualization system and method based on operation and maintenance data prediction. According to the method, an existing log collection framework is used for collecting machine and application log information in a network cluster to generate operation and maintenance big data, and then an artificial intelligence method is used for processing the operation and maintenance big data, so that network faults can be predicted in advance and displayed in a visual mode. The method has the advantages that the fault can be efficiently and accurately recognized; the network fault can be predicted in advance; sufficient time is provided for operation and maintenance personnel to carry out operation and maintenance work in time; and the work efficiency of theoperation and maintenance personnel is effectively improved in an assisting mode.

The invention discloses a small logistics management platform design method based on a cloud platform. The method is characterized by using computer resources wasted by a logistics enterprise as a basic layer; using cloud computing technology to deploy the resources as a cloud platform OpenStack cloud platform to constitute the private cloud of the enterprise; using the cloud platform OpenStack tocreate a highly available Web server cluster; connecting a logistics information management platform to the Web cluster and connecting the logistics information management platform to multiple cloudservers so as to reduce a risk that the entire server is out of service due to the shutdown of a single server. The method solves the problem of high investment cost and low operational efficiency ofsystem development in the prior art.

The disclosed embodiments provide a system that performs A/B testing using ego network clusters. During operation, the system obtains a graph that includes a set of nodes representing users in an online network and a set of edges representing relationships or interactions between pairs of the nodes. Next, the system selects egos containing a subset of the users that are representative of a distribution of an attribute across the users. The system then generates clusters of users, such that each cluster contains an ego and alters selected from the ego's connections in the graph. The system further divides the clusters between a treatment group and a control group, exposes alters in the treatment group to a treatment variant of an A/B test, and exposes alters in the control group to a control variant of the A/B test. Finally, the system outputs a result of the A/B test.

The invention provides a web cluster load balance adjusting method and device. The method is applied to a configuration center of a web cluster, and comprises the following steps: computing system resource usage amount of a server in the web cluster; judging whether the system resource usage amount is greater than the preset usage amount threshold; if the system resource usage amount is greater than the preset usage amount threshold, computing the weight of the server according to the system resource usage amount, the preset usage amount threshold and a predetermined rush hour interval; and sending the weight of the server to a front-end processor of the server. Based on the method provided by the invention, the automatic configuration of the server weight is realized, and the problem of manually adjusting the weight can be solved.

The invention provides a distributed transcoding method, device and system. The method comprises the following steps: receiving a video transcoding request that is sent from a client and is distributed by a web cluster, and ensuring that the video transcoding request contains a content identifier of the video data to be transcoded; acquiring the video data to be transcoded from a cloud memory according to the content identifier; transcoding the video data to be transcoded to obtain the transcoded target video data; and sending the target video data to the web cluster, and ensuring that the web cluster sends the target video data to a client. Compared with the prior art, the process of downloading, cutting and synthesizing video files is at least saved by using the technical scheme provided by the invention, and thus not only the manual operation process (such as the process of locally downloading the video files) of the user can be omitted, but also the transcoding efficiency of the video data can be increased.

The invention provides a DECT network clustering system and clustering method. The DECT network clustering method comprises the steps that a DECT host sends a multicast instruction announcing joiningin a set multicast group to a router; DECT equipment sends a multicast detection message sent to the target multicast group to the router; the router receives the multicast instruction and the multicast detection message, and sends the multicast detection message of which the target multicast group is a set multicast group to the DECT host according to the multicast instruction; the DECT host receives the multicast detection message and sends state announcement information to the DECT equipment in a point-to-point manner according to the multicast detection message; the DECT equipment receives the state announcement information, and initiates a TCP connection request to the DECT host in a point-to-point manner according to the state announcement information; and the DECT host receives theconnection request of the DECT equipment and establishes a TCP channel with the DECT equipment. In the technical scheme of the invention, the multicast detection message is a multicast message, and the message does not need to be broadcasted based on a multicast protocol; and meanwhile, the state announcement information and the connection request are sent in a point-to-point manner, so that thebroadband requirement of cluster communication is greatly reduced.

A system, method and apparatus for routing traffic in ad-hoc networks. A routing blockchain network processes routing node information proposals received from manager nodes of network clusters. Performance metrics of one or more nodes in the system are verified using distributed ledger techniques and provided to the manager nodes as updates to each manager node's routing information. The manager nodes further determine routing paths for ad-hoc communication requests based on an authentication event that defines conditions necessary to route traffic streams in association with a particular resource.