62results about How to "Uninterrupted service" patented technology

The invention relates to a system level virtualization technology and an energy-saving technology in the field of the structure of a computer system, and discloses an energy-saving method of a cloud data center based on virtual machine migration and load perception integration. The method comprises the steps: dynamically completing the migration and the reintegration of the load of a virtual machine in the cloud data center by monitoring the resource utilization rate of a physical machine and the virtual machine and the resource use condition of current each physical server under the uniform coordination and control of an optimized integration strategy management module of the load perception and an on-line migration control module of the virtual machine, and tuning off the physical servers which run without the load, so that the total use ratio of the server resource is improved, and the aim of energy saving is achieved. The energy-saving method of the cloud data center based on the virtual machine on-line migration and the load perception integration technology is effectively realized, the amount of the physical servers which are actually demanded by the cloud data center is reduced, and the green energy saving is realized.

The invention discloses a method and system for allocating IP addresses to containers and relates to the technical field of computers. According to one specific embodiment of the invention, the methodcomprises the following steps: establishing a one-to-one corresponding relation between a part of IP addresses in a total address pool and corresponding containers according to a first allocation requirement; when a request for allocating an IP address is received, judging whether a container whose IP address is to be allocated has a binding relation with a certain IP address; if so, allocatingthe IP address bound with the container whose IP address is to be allocated to the container; and otherwise, allocating an unbound IP address in the total address pool to the container whose IP address is to be allocated. According to the embodiment of the invention, the IP address and the container can be bound such that it can be guaranteed that the IP address is not changed when the container is restarted or migrated, and it is guaranteed that the service is not interrupted.

The invention discloses a cloud computing platform fault detection method, a cloud computing platform fault detection method, a solving method and a solving device, and aims to achieve the purpose of automatically discovering and resolving a fault of a cloud computing platform. According to the methods and the devices, the cloud computing platform is divided into various detection layers, whether the operating state of the initial detection layer is normal is detected, the abnormality of an operating state of the initial detection layer is taken as an entrance for each stage of the detection layer, and the operating state of the current detection layer is obtained by detecting the next stage of the detection layer as the current detection layer stage by stage, so that the lowest stage detection layer of which the operating state is abnormal is searched and the detection layer on which the final fault is positioned is determined; therefore, the purpose of automatically discovering the fault is achieved; in addition, on the aspect of resolving the fault, different solving mechanisms are made to solve the problems of each stage of the detection layer to reduce the fault time as possible by judging the detection layer on which the fault is positioned, so that the purpose of recovering service quickly is achieved.

The invention provides a service capability control method of a server cluster. The service capability control method comprises the steps of monitoring service state information of servers in a cluster, and judging whether the service state information exceeds a preset threshold or not, if so, determining the quantity of the servers issued by subsequent service requests according to the service state information, wherein the quantity of the servers issued by the subsequent service requests is smaller than that of the servers issued by the service requests when the service state information does not exceed the preset threshold; and determining corresponding quantity of target servers in the cluster according to the preset rule and according to the quantity of the servers issued by the subsequent service requests, and issuing the received service requests to the target servers. The burst traffic exceeding the service capability can be quickly coped with by the method provided by the invention, and service without interruption is also ensured.

A software upgrade method for uninterrupted network services involving the network communication industry. First, the service program is ready for initialization; the socket reuse mechanism is used to establish a listening socket with the same conditions as the service port receiving the service before the upgrade, and the operating system Adopt the more appropriate matching principle and the principle of post-priority, and send the client's request to establish a new service directly to the upgraded service instead of the previous service; the upgraded service provides services to the client according to the normal process; the upgraded service provides services to the previous The service sends an "exit" message; after the previous service receives this message, it closes the service listening port and no longer receives new service requests, and waits for the service that has been accepted to complete, and then actively exits. The upgrade software completely takes over the service under the condition of uninterrupted network terminal service.

The invention discloses a method, a system and a device for recovering control and provisioning for wireless access point (CAPWAP) tunnel. The method comprises the steps that: when the CAPWAP tunnel between an access point (AP) and an access controller (AC) fails, the AP and AC set states as operating information locking states respectively and save the session identifier (ID) of the failed CAPWAP tunnel; the AP locks the saved operating information, aborts the detection for a station and does not respond to the connection request or offline request of the station; the AP interacts with the AC for the tunnel recovery message, the AP and the AC recover states as operating sates respectively to recover the failed CAPWAP tunnel; the message comprises the information about the AP and the AC being in the operation information locked state and the session ID of the failed CAPWAP tunnel; the AP starts the detection for the station to update the operating information. According to the invention, the AP still can keep the service for the station uninterrupted during the recovery of the CAPWAP tunnel.

Embodiments of the invention provide a method, a device, and a system for arbitration of a double-active data center, and relates to the technical field of computers. The arbitration method comprises: when communication between a first data center and a second data center is determined to be disconnected, and an arbitration device cannot arbitrate, the first data center checking a preset arbitration strategy; when according to the arbitration strategy, the first data center is determined to be a priority data center, continuing to provide service; when according to the arbitration strategy, the first data center is determined not to be the priority data center, stopping to provide service; wherein the first data center and the second data center are double-active data centers. Through the above scheme, when the communication between the first data center and the second data center is determined to be disconnected, and the arbitration device cannot arbitrate, just the priority data center continues to provide service, so that a problem of service interrupt caused by faults of the arbitration device is solved, and uninterrupted service is ensured to be provided for users.

The invention relates to a distributed system-based serial number generation method and apparatus, and an electronic device. A distributed system comprises multiple servers; each server comprises at least one serial number service unit; and each serial number service unit comprises at least one serial number generator. The method for generating a serial number by each serial number service unit comprises the steps of receiving a serial number application request; generating an internal serial number through the serial number generator; and obtaining a corresponding instance number from a localend, and according to the instance number and the internal serial number, generating the serial number. According to the technical scheme, the serial numbers are generated in parallel through the serial number service units, so that excessively high database pressure caused by dependence on a database can be avoided and continuous services can be provided when the database has a fault.

The invention relates to a system for achieving a safety audit. The system comprises a plurality of clients, a plurality of servers and a safety audit device. The invention further relates to a safety audit method. The method includes the steps that service request messages are sent to the safety audit device by the clients; the safety audit device forwards the service request messages to the servers corresponding to the service request messages, and a service request message mirror image is copied so that a corresponding mirror image file can be obtained; the safety audit device analyzes the mirror image file so that corresponding service operation instructions can be obtained; the servers respond to the service request messages and send corresponding service operation logs to the safety audit device; the safety audit device compares and analyzes the service operation instructions and the service operation logs. The system for achieving the safety audit has the functions of double-machine hot standby and cluster deployment, is easy to popularize, stable and reliable, supports dynamic expansion and has a wider application range.

The invention discloses a method for realizing network flow load balancing by utilizing a consistency Hash strategy. A DNS analyzes an ip address corresponding to a set of service devices; the set of devices externally provides services at the same time; the flow is distributed to the service devices corresponding to the set of service devices by utilization of the consistency Hash strategy of a switch according to a consistency Hash algorithm; a new load balancing device is not increased; the load balancing function is realized completely with the help of the routing function of the switch; when a fault occurs, because a device cluster bears the flow at the same time, services can be provided continuously; when the fault of a certain device is recovered, the device is re-added into the consistency Hash list of the switch; therefore, the flow is smoothly switched; and jitter is avoided.

The invention provides a dynamic loading method used for a distributed container. The dynamic loading method comprises steps that a message aiming at predetermined service is received, and comprises an operation code field, which is used to indicate that the message is a service updating message or a service requesting message; the message is forwarded to a logic container according to the predetermined service, and the logic container comprises a logic component used for the predetermined service; when the message is the service updating message according to the operation code, the logic container is used to execute logic updating operation, and when the message is the update requesting message according to the operation code, the logic container is used to execute the loading of the logic component used for the predetermined service. According to the technical scheme, by focusing on logic separation after logic changing, and the logic after the change is packaged by a message mechanism, and therefore the smooth changing of the business logic is completed, the dynamic loading is realized, effects such as uninterrupted service and multi-version parallel operation are achieved, and then a user can use the service provided by a product smoothly.

The invention discloses a disaster tolerance processing method of a 5G network slice. A network slice set is reconstructed and a disaster recovery slice set is established. Under the condition that the 5G network slice is abnormal, uninterrupted service is ensured through unperceived switching of abnormal slice users, the influence of abnormal slices on the service is avoided, the service qualityof the network slice is effectively improved, disaster recovery resources are convenient to share, and the utilization rate of the disaster recovery resources is effectively improved.

A hot backup method based on the TFTP comprises the steps as follows: one backup device is used for backuping more than one main device, and a backup process is completed on the basis of the TFTP; the main device writes to-be-backuped data into a file, then the file is transferred to the backup device, and the backup device stores received data in a temporary file; whether transmission is successful and whether the verification file is received correctly are judged; and if the transmission is successful and the file is received correctly, the original backup file is replaced with the temporary file, otherwise, error information is sent to the main device. According to the method, data backup is completed by using files; and the backup device can be guaranteed to upload effective backup data at any time, influences of backup on normal operation are reduced, the high efficiency and the reliability of the backup are effectively improved, and the purpose of un-interruption of device service is achieved.

The embodiment of the invention provides a data migration method, a data migration device and equipment. The method comprises the following steps: migrating a full-volume data index in an old storagesystem to a new storage system, and marking a time period for migrating a first data index as a full-volume migration time period; obtaining an incremental data index generated in the old storage system in the full migration time period; if the quantity of the incremental data indexes is smaller than a migration data threshold value, switching a storage system providing service from an old storagesystem to a backup storage system; migrating the incremental data index to a new storage system; switching the storage system providing the service from the backup storage system to a new storage system; and migrating the data in the old storage system to the new storage system according to the data index in the new storage system. By means of the method, data migration in the storage system canbe achieved under the condition that services provided to the outside are not affected, so that the service request of a user is guaranteed. The user experience is improved.

The invention provides a method for implementing fusion of network middleware, which comprises the following steps of: allocating a unique service identifier for each resource by applying the Hash algorithm so as to uniquely identify the resources; selecting a name routing based on the Chord routing algorithm; performing name routing analysis; and performing network middleware fusion according to the network middleware strategy. The method for implementing the fusion of the network middleware guarantees network openness and transparency for new application, and can successfully fuse the network middleware into the Internet architecture. Therefore, the problems of Internet business and safety are effectively solved on the premise of not damaging the hierarchical structure of the Internet.

The invention discloses a disaster recovery switching method and device based on a virtual machine. The disaster recovery switching method comprises steps of transmitting a detection order to a switch positioned on a service plane side according to a preset time interval in order to detect whether connection of a gateway of the service plane is normal or not, if the gateway connection of the service plane is detected to be not normal according to the detection order, reporting to an operation system and sending alarming information, obtaining a control order transmitted by the operation system when the operation system determines the switch has faults according to the alarming information, and switching a host machine to a memory plane according to the control order. The embodiment of the disaster recovery switching method and device can transfer a flow of the service plane to the memory plane on the basis of that the current hardware facility is not changed, multiplex a current storage channel to bear service transmission during a fault period, reduce resource waste, guarantee that the service that the virtual machine on the host machine provides to the external is not interrupted and dramatically improve reliability of the disaster recovery capability and service of the virtual machine during the external service of the virtual machine.

The invention discloses a method for detecting invalidation of group members of an internal server and network address translation (NAT) equipment. The method comprises the following steps of: if the NAT equipment receives an ICMP (Internet Control Message Protocol) target non-accessible message sent by internal network equipment, determining a non-accessible server according to the message, switching the state of the server to detection, detecting whether the server is accessible on three layers, and if not, switching the state of the server to invalid; otherwise, switching the state of the server to stale, marking a session corresponding to service flow distributed to the server with an S, switching the state of the server to delay, setting a first timer for the server, switching the state of the server to valid if the session with the mark S is switched from open state to ready state or three times of handshake of a TCP (Transmission Control Protocol) specific to the session with the mark S are successful; and otherwise, switching the state of the server to invalid. Due to the adoption of the method and the equipment, the invalidation of group members of the internal server can be detected.

The invention discloses a method for realizing high availability and load balancing of Influxdb-proxy on the basis of zookeeper, and the method comprises the following steps that S1, the Influxdb-proxy is started, service information is registered to a zookeeper father node, and a temporary child node is generated; S2, the zookeeper monitors the operation state of each influxdb-proxy service in real time through a watcher, after a certain service is abnormally offline, the zookeeper can actively delete the temporary node of the service, and meanwhile, the zookeeper informs a client of node change; and S3, the client obtains the available influxdb-proxy service address from the zookeeper again. The method for realizing high availability and load balancing of the Influxdb-proxy based on thezookeeper has the advantages that the availability is high; the operation and maintenance cost can be reduced, the stability of the influxdb cluster is improved; the influxdb cluster can be guaranteedto continuously provide the service for the outside, loss caused by hardware, software and human error proneness is reduced, concurrency is improved, the zookeeper can be configured with multiple influxdb-proxy, load balance is realized, and influxdb cluster performance is improved.

The invention discloses an HDFS storage system, comprising: a plurality of metadata management nodes, a distributed system high-availability component connected with each metadata management node, a metadata storage pool. The metadata management nodes are used for receiving and processing a storage request of data to be stored. The distributed system high-availability component is used for transferring a storage request sent to the current metadata management node to another metadata management node when the current metadata management node corresponding to the current distributed system high-availability component is down. The metadata storage pool is used for storing the data to be stored, and the plurality of metadata management nodes establish communication links with the metadata storage pool. The HDFS storage system provided by the invention guarantees the data consistency in the service switching process and will not lose the data to be stored in the switching process. The invention also provides a data storage method, which has the beneficial effects.

The invention relates to the technical field of artificial intelligence, and discloses a Rocket MQ distributed deployment method and device, equipment and a medium, and the method comprises the steps: enabling to-be-deployed server group data to comprise at least two to-be-deployed server groups, wherein the to-be-deployed server groups are in one-to-one correspondence with to-be-deployed machine rooms; respectively deploying an addressing service node and at least two working nodes in each to-be-deployed server group in the to-be-deployed server group data by adopting the Rocket MQ node file; locating at least two working nodes of each Broker component in the to-be-registered Broker component set in at least two to-be-deployed machine rooms, wherein all the working nodes in the Broker components are configured as a main working node and at least one standby working node; and determining a target Rocket MQ cluster according to all the addressing service nodes and the to-be-registered Broker component set. Message loss is avoided when the machine room breaks down, and service is not interrupted.

The invention relates to a lots of real time data storage distribution and query information processing method. The method comprises the following steps that: 1). A data query server is started, after the starting, a corresponding table of data gateway corresponding to each terminal is loaded into a memory and a query service is provided externally; 2). A data storage server is started, is used to store data distributed by a data distribution server and simultaneously response the distributed query of the data query server; 3). The data distribution server carries out data distribution; 4). The data query server carries out the query. Compared to the prior art, by using the method of the invention, a plurality of the data storage servers can be configured which means more numbers of terminals can be supported.

The invention relates to an asymmetric single-double current hybrid indoor coverage system. The system comprises an LTE base station remote radio unit, near-end equipment and an active antenna which are in two-way transmission connection. The LTE base station remote radio unit is provided with a double current downlink signal outputting function and a single-double current uplink signal receiving function at the same time, downlink signals are output through the active antenna, and the uplink signals are received through the active antenna. The near-end equipment comprises frequency conversion equipment and a combiner. Output ends of the LTE base station remote radio unit are connected with input ends of the frequency conversion equipment and the combiner, and an output end of the frequency conversion equipment is connected with an input end of the combiner. The active antenna is provided with frequency conversion and power detection functions. An output end of the combiner is connected with an input end of the active antenna. The system has the advantages that the used active antenna is capable of carrying out power detection and calculating the monitoring the loss condition of each feeder cable in real time, so that the problem that a service provider fails to monitor indoor feeder cables in real time is solved.

The invention discloses a virtual machine data service system and method. The system provides primary and backup sites, a host at the primary site and a host at the backup site share a storage interface, the system supports a storage active-active function, and the host at the primary site creates a virtual machine based on virtual machine data on a storage system volume and provides virtual machine services to the outside. When the host or a storage end at the primary site has failures, the host or the storage end at the primary site can be switched to the host or a storage end at the backupsite by using the shared storage interface and the storage active-active function, and the virtual machine services can be continuously provided to the outside. Thereby, even if the failures occur, the virtual machine services can also be provided to the outside, and the phenomena that the virtual machine is not crashed and the virtual machine services are uninterrupted can be ensured.

The invention provides an enterprise instant communication method, which comprises the following steps: distributing all on-line enterprises to different instant communication servers according to a predefined regular, further distributing newly-added enterprises to an instant communication server according to the regular, controlling the server to check whether all instant communication servers are normal or not at regular, if an instant communication server is dead or can not normally work, notifying other instant communication servers to take charge enterprises which are served by the failure instant communication server. And then, the instant communication system provides instant communication servers for many enterprises through an Internet on-line mode, and assures the services of on-line enterprises without interrupting.

The invention provides a virtual machine switching system. The virtual machine switching system is used for switching virtual machines when one virtual machine is abnormal, and comprises a base image setting module, a virtual machine generating module and an automatic switching module. The base image setting module is used for correspondingly setting and storing at least one virtual machine base image according to at least one virtual machine, and the virtual machine generating module is used for generating and starting the corresponding new virtual machine according to virtual machine information and virtual machine image information corresponding to the virtual machine when the virtual machine is abnormal. The automatic switching module is used for achieving switching between the new virtual machine and the corresponding abnormal virtual machine so that the new virtual machine can replace the abnormal virtual machine to work normally. The switching system achieves automatic switching between the new generated virtual machine and the abnormal virtual machine when the virtual machine works abnormally so that the new virtual machine can replace the abnormal virtual machine to work normally, operation is convenient, and continuity of services provided by the virtual machine can be guaranteed. The invention further provides an automatic virtual machine switching method.

The invention relates to a system level virtualization technology and an energy-saving technology in the field of the structure of a computer system, and discloses an energy-saving method of a cloud data center based on virtual machine migration and load perception integration. The method comprises the steps: dynamically completing the migration and the reintegration of the load of a virtual machine in the cloud data center by monitoring the resource utilization rate of a physical machine and the virtual machine and the resource use condition of current each physical server under the uniform coordination and control of an optimized integration strategy management module of the load perception and an on-line migration control module of the virtual machine, and tuning off the physical servers which run without the load, so that the total use ratio of the server resource is improved, and the aim of energy saving is achieved. The energy-saving method of the cloud data center based on the virtual machine on-line migration and the load perception integration technology is effectively realized, the amount of the physical servers which are actually demanded by the cloud data center is reduced, and the green energy saving is realized.

The invention relates to the technical field of third-party equipment data processing, and discloses an access method and equipment of third-party equipment, electronic equipment and a readable storage medium. The method comprises the steps: responding to the configuration operation of model parameters of at least one third-party equipment; collecting a channel program of the at least one third-party equipment based on the equipment type of the at least one third-party equipment; analyzing the channel program, and determining an application program interface corresponding to the configurationcommand of the at least one third-party equipment; determining debugging data of the at least one third-party equipment based on the configuration attribute of the at least one third-party equipment and the application program interface; and determining an access state of the at least one third-party equipment according to the debugging data of the at least one third-party equipment so as to debugthe channel program. By implementing the invention, the code burden is reduced, the real-time return of debugging data and the online debugging of the third-party equipment are realized, and the access state of the third-party equipment is determined in time.

The invention discloses a low-delay and high-concurrency media stream distribution method based on a multi-layer structure. A core server layer, an exchange server layer and a user server layer are involved. The method includes the steps that after a user machine requests for a media stream, a user server determines whether the media stream which the user machine requests for has been in a service list of the user server or not, the media stream is directly sent to the user machine if the media stream which the user machine requests for has been in the service list of the user server, the request is sent to an exchange server if the media stream which the user machine requests for is not in the service list of the user server, the exchange server determines whether the media stream which the user machine requests for has been in a service list of the exchange server or not, the media stream is sent to the user server and then transmitted downwards to the user machine if the media stream which the user machine requests for has been in the service list of the exchange server, and the exchange server requests a core server for the media stream if the media stream which the user machine requests for is not in the service list of the exchange server. By means of the method, the number of concurrency users of classroom stream media is greatly increased; meanwhile, the utilization rate of all stages of servers is increased, and the bandwidth between the servers is reduced.

The present invention discloses a user identification method of mobile communication terminal in wireless local area network which belongs to network using technology field of mobile communication terminal, aiming at dual-mode terminal which can use mobile communication network and wireless local network, when mobile communication terminal moves into wireless local network, to implement explicit or implicit identification operation according to application programme in use selectively. Including steps as follows: the step to identify the presetting identification information of the application programme when implementing application programme through user, the step to implement identification operation through the confirmed identification information. This invention has the following effect, implementing uninterrupted service from mobile to wireless local network, at the same time, preventing from needless cost.

The embodiment of the invention discloses a network traffic transfer method. The method comprises the following steps of recording a connection state of network traffics of at least one host and a hardware address of the host by a connection state recorder; receiving a transfer request by the connection state recorder; transmitting network traffics under a connected state to a source host by the connection state recorder according to a connection state of the network traffic of the source host in the transfer request; forwarding network traffics newly initiated by the network to a destination host of the transfer request by the connection state recorder. The network traffic transfer method provided by the embodiment of the invention can achieve that the service is not interrupted during IP (Internet Protocol) address switching or host switching, guarantee the seamless network traffic transfer and improve the user experience.