37results about How to "Solve the deadlock problem" patented technology

The invention discloses a network equipment configuration method, a client, a server and a network system, which belong to the field of network communication. The method comprises the following steps of: receiving a configuration submission request from the client, and caching configuration information in the configuration submission request; submitting the configuration information into a candidate dataset, and running and validating the configuration information to a running configuration database by using the candidate dataset; in a configuration conflict, updating a fundament dataset based on the current running configuration dataset; and acquiring the cached configuration information, re-executing the cached configuration information on the new fundamental dataset, running and validating the re-executed configuration information to the running configuration data set; and returning an execution result to the client. By the method, the client, the server and the network system, the requirements of concurrent configuration of a plurality of users are met, the operation concurrency of the plurality of users is improved, and the problems of low user concurrency and deadlock produced in overall locking or partial locking processing in the prior art are solved.

The invention relates to an inter-process thread mutex method. The inter-process thread mutex method comprises following steps: vying for mutex information by multiple threads and initializing mutex information by a successful thread; when obtaining mutex information by the successful thread and utilizing mutex information to lock a critical zone; obtaining a lock of mutex information by the thread after locking operation and visiting shared resources; waiting by other threads at the same time; parting the critical zone after the threads run out of shard resources; and continuing to compete for mutex information by other threads and repeating above steps. The inter-process thread mutex method has following beneficial effects: combined with a semaphore mechanism, atomicity of locking and unlocking is ensured; the method has thread safety and helps to effectively solve the problem of garbage value obtained by mutex information; abnormity is avoided when garbage value is obtained by using threads; the deadlocking problem is also overcome when the self body asks for a lock during locking; and the problem of unlocking out of limits of authority is avoided.

The invention discloses a limit shell script parallel executing method in the linux and unix platform, which comprises the following steps: using script mode processor to write request into pipe when the shell script needs to be executed; reading the request through token mode processor at the other end of pipe sequently; transmitting the token to the shell script according to the sequence of request when the token is free until the token is finished; obtaining the process permitting operation of token; exiting the execution if not obtaining the process of token. The invention guides token concept to realize 'serialization' of multi-process in the critical area through linux pipe, which uses single locking detecting script to detect the locking to solve the locking problem effectively without using multi-parallel process to operate one token.

The invention relates to the technical field of robots, and provides a path scheduling method and device, a background server and a storage medium. The method includes the following steps: receiving and generating a first navigation path from the current position of a first robot to a first preset task point of the first robot according to an obstacle processing request; acquiring the current navigation path of a second robot; when there is a conflict between the current navigation path and the first navigation path, generating a second navigation path from the current position of the second robot to a second preset task point of the second robot; and when there is no conflict between the second navigation path and the first navigation path, sending the first navigation path and the secondnavigation path to the first robot and the second robot respectively. Through reasonable scheduling of the moving path of a robot encountering an obstacle and the moving path of a robot encounteringno obstacle, the 'deadlock' problem of swarm robots in the same map is solved.

The invention discloses a link state delivery method and a device of a packet transport network. The method comprises the following steps: a home-terminal service provider edge PE receives a client layer signal failure message (CSF) from a far end PE for the first time; the physical port of the home-terminal user network interface (UNI) is closed; the far-end link of a home terminal indicates a position site, wherein the far-end link indication site is used for indicating that the link state from a far-end CE to the far-end PE is disconnected. The invention discloses a link state delivery device of a packet transport network. In the invention, the far-end link indication site is additionally arranged on a UNI physical port and is used for recording that the disconnection of the UNI physical port of a CE-PE link is caused by the home terminal or the far end; and when the UNI physical port disconnection is caused by home terminal or the far end, the signal failure message is not triggered to the far end so as to avoid the phenomenon that two ends continuously send the signal failure message, solve the problem of deadlock and improve system reliability.

The invention relates to the technical field of robots, and provides a branch node type path scheduling method, an apparatus, a background server and a first robot. The method comprises the followingsteps: receiving an obstacle processing request sent by the first robot; judging whether the obstacle is a second robot or not based on the obstacle processing request; sending a waiting instruction to the first robot to enable the first robot to execute the waiting instruction if the obstacle is the second robot, and waiting for a preset time at a current position of the first robot; and sendingan avoidance instruction to the first robot to enable the first robot to execute the avoiding instruction, so as to avoid the obstacle if the obstacle is not the second robot. According to the branchnode type path scheduling method, the apparatus, the background server and the first robot, moving paths of a plurality of robots are planned into branch node type paths, obstacles are uniformly and reasonably avoided in the background, thereby avoiding the deadlock problems that the plurality of robots is blocked to each other and the paths are constantly planned.

The invention provides a dense storage container scheduling method and device and electronic equipment, and the method comprises the steps: determining a to-be-scheduled container and a destination ofthe to-be-scheduled container based on an obtained to-be-executed task; determining a first path according to the current position and the destination of the to-be-dispatched container, wherein the first path comprises the position where the robot needs to pass when the robot transports the to-be-dispatched container from the current position to the edge position of the current block in the current block to which the current position belongs; and controlling the robot to convey the to-be-dispatched container to the edge position from the current position through the first path, and then conveying the to-be-dispatched container to the destination from the edge position. By means of the mode, the operation collaboration of all the robots in intensive warehousing operation can be effectivelyimproved; conflict and disorder of multiple robots during task execution are avoided, the problem of blockage or deadlock among the robots is solved, and the problem that the robots are blocked by other containers during task execution is solved, so that the operation efficiency of intensive storage is improved.

The invention provides a deadlock unlocking method, which comprises the following steps that data in each running service process are backed up in real time; when the deadlock of the service processes occurs, all of the service processes are terminated, and all of the service processes are restarted; during the service process initialization, the backed up data are restored into each service process, each service process normally runs, and in addition, the service processing starts. The method has the advantages that the data of each service process are backed up in real time, the data are restored in real time after the restarting of each service process due to deadlock, the service is started, and the deadlock problem is simply and efficiently solved.

The invention discloses a multi-robot path planning method and device, and belongs to the technical field of robot path planning, and the method comprises the steps: constructing a grid environment according to a robot channel environment, and carrying out the environment reconstruction of a cross point graph through reserving channel cross points; through a heuristic A * algorithm, performing one-time optimal path planning on a single robot in the cross point map environment; according to the primary optimal path, planning all channels in the cross point map environment as one-way streets; for the one-way streets, through the heuristic A * algorithm, carrying out secondary optimal path planning on a single robot in the cross point map environment; and aiming at the secondary optimal path, reconstructing a grid environment to obtain an optimal path in the grid environment. The defect that deadlock is caused by robot conflicts in channels in a traditional heuristic algorithm is overcome.

The invention discloses a message storage scheduling method. The method comprises steps of obtaining to-be-scheduled messages; according to preset write-in conditions, storing the to-be-scheduled messages of the same characteristic information in the same storage array; storing the corresponding relation between the characteristic information of the to-be-scheduled messages and address information in the storage array in a cache array; obtaining current entry information of each storage array, wherein the current entry information is a valid bit information group used for indicating whether or not messages in each storage array can be scheduled to enter a processing module; when the current entry information is valid by indication, messages of the corresponding storage array can enter the processing module by scheduling; and according to preset scheduling rules, reading the to-be-scheduled messages from a valid storage array indicated by the current entry information for entering the processing module. The invention solves the problem that block and deadlock may occur when messages of the same characteristic information are processed at the same time, and is advantaged by being convenient to operate and high in reliability.

The invention relates to a dynamic application program updating method for vehicle-mounted embedded equipment, and belongs to the technical field of ECU dynamic updating. The remote server compares the program information reported by the automobile ECU with the existing latest program version of the ECU and sends an instruction that the ECU of the automobile is about to be updated to the TBOX of the automobile where the ECU needing to be updated is located, the TBOX sends the updating packet to the ECU through the CAN bus after receiving the updating packet, the ECU waits for the TBOX to send an updating instruction after receiving the updating packet, and the ECU starts to perform dynamic updating after receiving the updating instruction of the TBOX, and the updating is completed. The method has the advantages that software updating can be completed under the condition that operation of an original program is not affected, original data are protected from being lost, and the problems of function semantic ambiguity and multi-thread resource deadlock are solved.

The invention discloses a multi-robot intelligent traffic control method and device, and the method comprises the steps: carrying out the collision detection and deadlock detection when a robot applies for a next element, wherein the deadlock detection is used for judging whether the robot will form a deadlock with other robots after the robot applies for the next element; and when it is judged that the collision detection and/or the deadlock detection are/is not passed, prohibiting the application passing the robot. According to the method and device, whether deadlock is formed when the multiple robots act at the same time or not can be automatically judged, the deadlock phenomenon is completely avoided, the traffic efficiency is improved, and human intervention is reduced.

Embodiments of the present invention provide a packet sending control method, and a processing method, device and system. The method comprises: a packet sending control device obtaining available space information of a receiving channel corresponding to each packet of a first protocol type in a receiving end device, and sending an indication packet of a second protocol type carrying the available space information to a packet sending processing device, so that the packet sending processing device performs, according to the available space information, sending control on the packet of the first protocol type sent from a sending end device. With the packet sending control method provided by the embodiment of the present invention, the packet sending processing device performs, according to the available space information, sending control on the packet sent from the sending end device, and sends the packet only when the available space information of the packet is valid, thereby effectively avoiding a deadlock problem caused by non-matching virtual channels during conversion of packets between different protocols.

The invention discloses a software defined network oriented balanced domain division method, and aims to provide a balanced domain division method which is fast in convergence, high in efficiency and capable of reducing network control time delay and improving network performance. The method is realized through the following technical scheme: after a network topology, a node position, a pre-estimated traffic matrix and a maximum scale allowed by a single domain are given, executing an initial domain division stage based on a K-means algorithm framework, and traversing initialized distributed domain center nodes by using a width-first graph; repeatedly carrying out distance-first domain growth and domain center node updating; executing a quotient graph-based local tuning stage on the basis of an initial domain division result, converting local tuning performed by movement of domain boundary nodes in adjacent domains in the domain division result into a quotient graph, and repeatedly searching a negative weight ring or calculating a minimum cost path on the quotient graph obtained by conversion of the domain division result; and the target function is optimized and the absolute balance of the domain scale is ensured until no negative weight ring exists in the quotient graph and the domain division result reaches the absolute balance.