191 results about "Daemon" patented technology

The embodiment of the invention provides a self-adaptive resource allocation method and device, and the method comprises the steps: collecting calculation node metadata and corresponding physical equipment information through deploying a daemon process at each node, associating the calculation node metadata with the corresponding physical equipment information, and constructing an allocable equipment pool; the real-time state of the node is sent to a cloud native scheduler, the cloud native scheduler generates a corresponding scheduling decision for the delay exceeding node obtained through judgment, the scheduling decision comprises an isolation decision and a migration decision, and in the migration decision, migration resource allocation is constrained based on an equipment pool, the scheduling decision of the cloud native scheduler is received, and node resources are scheduled. The method can improve the certainty of the hard real-time task and the overall efficiency of the system.

The invention discloses a node distribution method and system for a GPU (Graphics Processing Unit) use process in an NUMA (Non Uniform Memory Access) environment, and the method comprises the following steps: in a kernel mode, capturing an event that the process executes a new program and collecting meta-information of the process through an eBPF program mounted at a system call entry tracking point, and judging whether the process is a target GPU process or not based on the collected meta-information of the process; if the process is the target GPU process, transmitting meta-information of the process to a user mode, and if the process is not the target GPU process, ending and exiting; and for the meta-information transmitted to each target GPU process of the user mode, reading the meta-information of the target GPU process through the user mode daemon process, evaluating each node in the NUMA environment based on a preset node allocation rule, selecting an optimal NUMA node, and binding the target GPU process to a CPU of the optimal NUMA node for execution. The GPU computing efficiency and the resource utilization rate can be improved.

The invention discloses an intelligent cabin fault analysis method, device and equipment and a storage medium, and the method comprises the steps: starting a first thread to monitor a system fault in real time, and feeding back a daemon process to capture a field log when the system fault is monitored; after the daemon process finishes capturing the field log, a second thread is started to analyze a fault code in the field log, and a corresponding fault analysis function is matched; and the field log is analyzed again through the fault analysis function, and a fault generation reason and a fault coping strategy are obtained and broadcasted. According to the technical scheme provided by the embodiment of the invention, the fault analysis can be quickly completed, the fault generation reason can be fed back in time, and a fault coping strategy is provided for a user.

The invention discloses a virtual network card management method based on a KVM (Keyboard Video Mouse), which is applied to a KVM virtualization environment and comprises the following steps: a, starting a kvm-br-sync daemon process, b, acquiring an actual connection state of a virtual network bridge and a virtual network card from a virtual network bridge module; c, verifying the consistency of the mapping relation and the actual connection state; d, detecting whether the virtual machine is not operated or whether the virtual network card is invalid; e, detecting whether the connection between the virtual network bridge and the virtual network card is lost for no reason; and f, after the network is restarted or the fault of the virtual network bridge / hardware network card is repaired, reading the mapping relation in the database, and detecting whether the current state of the virtual machine is matched with the state of the virtual network card or not. Effective mapping is stored through a database, the core pain point that automatic recovery cannot be achieved after fault recovery is solved, and the reachability of a virtual machine network is guaranteed.

The invention provides a distributed lock batch processing method and system based on atomic scripts and a medium. The method comprises the steps that batch atomic locking is conducted, a client side initiates a single atomic request to a distributed storage server side, a first atomic script is called, the script atomically locks multiple resources, and a failure automatic rollback mechanism is built in; after batch atomic renewal and locking are successful, starting an asynchronous daemon thread, periodically initiating a single atomic request, and calling a second atomic script to realize batch renewal of a plurality of locks; and batch atomic unlocking: after the service is completed, initiating a single atomic request, calling a third atomic script, and realizing batch unlocking based on a reentrant counting mechanism. According to the method, locking, renewal and unlocking are all optimized into a single network request, the whole process and high performance are achieved, strong atomicity and consistency are guaranteed through an atomic script and a rollback mechanism, and the robustness of the system is enhanced.

The invention provides a method and system for remotely managing and controlling application programs, the method is applied to a daemon process of a Linux system, the daemon process externally provides a remote calling interface based on a remote calling protocol, and the method comprises the following steps: receiving a management and control request sent by a client through the remote calling interface, the control request is used for triggering the daemon process to control the application program; and managing and controlling the application program based on whether the sub-process corresponding to the application program exists or not. According to the method, a remote calling interface based on a remote calling protocol is externally provided by setting a daemon process, so that a client issues a management and control request to thousands of Linux servers (or virtual machine resources) only through the remote calling interface, and efficient and unified management of large-scale equipment is realized. After the daemon process receives the management and control request, the application program can be accurately managed and controlled by judging whether the sub-process of the application program exists or not.

The invention relates to the field of artificial intelligence, in particular to a downtime recovery method of a management node, electronic equipment and a storage medium. The method comprises the following steps: when a target management node is in a downtime state, configuring a new management node according to configuration information of the target management node; setting stains for the new management node through a preset container arrangement platform component, removing a fault node in a target cluster to which the target management node belongs, and restarting a distributed key value storage system service of the target cluster; and setting the priority of the high-availability daemon process service of the new management node as the lowest level in the target cluster, restarting the high-availability daemon process service of the new management node, and removing stains to complete the recovery action of the target management node. Therefore, the problems that in the prior art, the recovery process is complex and time-consuming, and the availability of running services and platform pages is affected are solved, and rapid and stable recovery of the management node is achieved.

This invention discloses a cache adjustment method, apparatus, device, and computer-readable storage medium, comprising: collecting access characteristic data from the client layer, object storage daemon layer, and device layer respectively; determining the object access mode of the system based on the access characteristic data; and adjusting the cache priority of each access task, the cache capacity ratio between the client layer and the object storage daemon layer, and the data residence time in the cache layer according to the object access mode; wherein the cache layer includes the client layer and the object storage daemon layer. This invention improves cache hit rate and overall read / write throughput performance, and reduces access latency.

The embodiment of the invention discloses a unified management method, device and system for cross-platform intelligent equipment and a medium, and the method comprises the steps: constructing a platform adaptation layer which is integrated with communication protocol modules of at least two kinds of platform equipment so as to interact with the platform equipment corresponding to the communication protocol modules; periodically scanning all connected devices based on the platform adaptation layer through an independent daemon thread; analyzing the scanned original equipment information of the different platform equipment, and converting the analyzed original equipment information into a preset unified equipment information data structure; analyzing according to the scanned equipment information to obtain an equipment state change event; transmitting an equipment state change event based on a publish-subscribe mechanism; the published equipment state change event carries equipment information of a preset unified equipment information data structure. Therefore, automatic discovery, identification, monitoring and management of cross-platform intelligent equipment can be realized, and the framework is flexible and easy to expand.

This electronic device may comprise: a memory storing instructions and including one or more pieces of storage media; and at least one processor including processing circuitry. When executed individually or collectively by the at least one processor, the instructions can control such that the electronic device executes a kernel thread in a kernel space by using the processor, checks whether a daemon is terminated in a user space after a designated time elapses since the kernel thread was executed, increases the frequency of counting on the basis that the daemon in the user space is not terminated and waits for termination of at least one thread, determines, on the basis that the frequency of counting exceeds a designated level, that the daemon is in a non-responding state, stops all requests in progress through a file system, and determines, on the basis that all threads of the daemon are terminated, that the non-responding state of the daemon is released.

The embodiment of the invention provides a control method and device for a parallel driving cabin, the parallel driving cabin and a storage medium, and relates to the technical field of parallel driving. The method comprises the following steps: constructing a three-level process framework of parallel driving cabins; wherein the three-stage process architecture comprises a host process, a daemon process and a recovery process; the running state of the host process is monitored in real time through the daemon process, and the sub-health degree is calculated based on the running state parameters; and controlling the host process, the daemon process and / or the recovery process to execute corresponding level adjustment operation according to the sub-health degree. By constructing a three-level process architecture of parallel driving cabins, modular separation of system functions is realized, so that the daemon process can monitor the running state of the daemon process in real time independent of the host process. The corresponding level adjusting operation is executed by quantifying the health state of the system, so that the system can perform self-adaptive adjustment in time when abnormality is detected, resource exhaustion or complete function failure is avoided, and the requirement for continuous and stable operation of the parallel driving cabin is met.

This invention discloses a method, device, terminal, and storage medium for sensing operating system fault information, belonging to the field of operating systems. The method includes: deploying a resident lightweight resource detection tool to detect external manifestations of anomalies; when anomalies appear, activating an information collection daemon to periodically collect system resource information using a small amount of resources, closing the resident lightweight resource detection tool, and generating system resource trend information; using the system resource trend information to determine if there is an abnormal trend deterioration; when there is an abnormal trend deterioration, selecting and loading a corresponding medium-weight resource information collection plugin based on the system resource information corresponding to the abnormal trend; using the medium-weight resource information collection plugin to collect system information and recording the system information. This method can simultaneously ensure resource effectiveness and information integrity.

The invention discloses a virtual machine peripheral secure mounting method based on deep fingerprints and dynamic risk control, which is applied to a host machine virtual machine environment. After the peripherals are accessed, a daemon process is triggered to obtain VID / PID, descriptors, class specific function information, instruction response time sequences and other characteristics, deep fingerprints are constructed, a strategy library is inquired, and the peripherals are divided into white lists, black lists or unknown devices; automatically mounting the white list, stopping mounting the black list and recording an event; and calculating a comprehensive risk value R for unknown equipment, comparing the comprehensive risk value R with a threshold value, and executing automatic mounting, secondary authentication or mounting prevention. And after mounting, continuously monitoring I / O behaviors, extracting throughput, an access mode, a burst time sequence and file semantic features to carry out anomaly detection, and carrying out transmission pause, alarm and disconnection treatment when the anomaly is detected, so that full-life-cycle safety management and control are realized.

A device may include a processor that includes a named data network (NDN) forwarding daemon (NFD) module. The NFD module may be configured to identify an accelerator on which to configure an offloaded NFD module based on a configuration table indicating a current configuration of the accelerator. The NFD module may be configured to configure the offloaded NFD module on the identified accelerator. The NFD module may be configured to receive an interest packet that includes a workload request. The interest packet may be configured according to an NDN protocol. The NFD module may be configured to determine that the offloaded NFD module is configured to perform the workload request using the identified accelerator based on the configuration table. The NFD module may be configured to offload, via an application programming interface, the workload request to the offloaded NFD module to perform the workload request using the identified accelerator.

This invention provides a self-healing scheduling method for RPA clusters based on microtask state reconstruction, relating to the field of mobile terminal RPA scheduling technology. The method includes deploying a daemon process on the mobile terminal side and establishing a direct communication link with the scheduling center; the scheduling center detects the running status of the mobile terminal application using command probes and image probes, calculates the soft fault fusion probability based on weighted fusion, and triggers self-healing when the probability exceeds a threshold; the business process is decomposed into microtasks with dependencies and a dependency graph is constructed; during execution, the execution state and business context are synchronously stored in the central storage; after a soft fault occurs, the breakpoint is determined and a backup mobile terminal is selected from the session pool to take over; the backup mobile terminal obtains the breakpoint task data and reconstructs the execution environment; before injecting business data, an environment consistency check is performed; if inconsistency occurs, a compensation operation is performed before continuing to execute subsequent microtasks, thereby improving the success rate and execution stability in cluster scenarios.

The invention discloses a Ceph performance tuning method, and relates to the technical field of cloud computing distributed storage. Comprising the following steps of: 1, dividing Ceph-osd threads into a service processing thread, a network message transceiving thread, a message distribution thread, a heartbeat thread, a command interaction thread and other threads by analyzing a Ceph-osd system source code, a key service thread model and a monitoring condition during operation; according to the method, starting and exiting conditions of an osd process of a storage node are monitored in real time, a thread number of a Ceph-osd thread is obtained through an htop tool, the Ceph-osd thread is bound to a specified CPU set through a taskset tool according to configuration of a configuration item, and resources are prevented from being preempted by the thread.

The invention provides a flexible license management system and method. The system comprises a user operation module, a control management module and a storage module, the user operation module is used for providing a Web visual interface and / or a command line interface and interacting with the control management module; the control management module is used for automatically constructing a license server FlexLM mirror image and a license file license.dat in a container according to an instruction of the user operation module; when the container runs, a license management daemon process lmgrd and a vendor daemon process provided by each manufacturer are deployed in a running environment; the storage module is used for storing an lmgrd log in the running process of the container, and the lmgrd log comprises a license check log. By means of the scheme, the FlexLM deployment operation of the license server can be simplified, and environmental dependence is eliminated.

The application relates to the technical field of data storage, and discloses a distributed storage system capacity balancing triggering method, device and equipment, which comprises the following steps: obtaining a notification message sent by a process manager daemon process, obtaining a latest object storage device data set according to the notification message, determining a current storage device failure ratio according to the number of storage device failures in the latest object storage device data set, determining whether to change the attribute value of an attribute parameter corresponding to a capacity balancing operation into a first attribute value according to the current storage device failure ratio and / or the comparison result between the latest object storage device data set and a previously obtained object storage device data set, so that the capacity balancing operation is executed according to the first attribute value in the future. The application can respond to the change of a cluster state in real time, timely trigger the capacity balancing operation, optimize the condition for triggering the capacity balancing operation execution, realize automatic triggering of the capacity balancing, does not need to depend on management software, and reduces the system maintenance cost.

The present invention relates to a real-time container engine implementation method based on the Ruihua process version operating system. In view of the resource limitations, system complexity and security problems faced by the deployment and operation of existing container technologies in embedded systems, a real-time container engine architecture and implementation are constructed. The real-time container engine architecture includes a container runtime and a container engine daemon process. The container runtime includes real-time container configuration and container operation process and state transition. The container engine daemon process includes image management, state monitoring, log information, container and image services. It can support the deployment of OCI containers in a low-resource and high-real-time requirement environment and is compatible with traditional container management programs. It can be taken over by upper-level orchestration programs for scheduling and operation. It also optimizes the startup and memory consumption of traditional container engines and runs the container engine in a smaller space. It can meet the real-time requirements of user programs on the basis of ensuring the normal operation of the container.

A system for processing database requests includes an operator module operable to receive a request requiring a long running operation, and assign an operation identification (ID) to the long running operation; and a database module for receiving the request and the operation ID, the database module including a database daemon for launching the long running operation, storing a state of the long running operation, and reporting a status of the long running operation.

The invention provides a method for docking and adapting a working node host of a cloud computing cluster and a DPU (Data Processing Unit), a DPU chip, a DPU intelligent network card and cloud computing equipment. The method comprises the following steps: calling a single-root I / O (Input / Output) virtualized network equipment plug-in to allocate a first virtual function; creating a first container group, and distributing the first virtual function to the first container group; calling the multi-interface container network element plug-in, and creating a first interface device corresponding to the first virtual function; the vector data packet processing framework agent calls a DPU virtual resource management daemon process to load and manage the first interface device; and the multi-interface container network element plug-in loads a virtualized network driver for the first virtual function, and transfers the first interface device to a network space of the first container group. According to the technical scheme provided by the invention, complex network layer functions and large-scale network topology processing can be supported, and seamless support and efficient network policy management on the cloud computing cluster are realized.

A computer-implemented scheduling method for accelerator hardware includes using a computational graph that splits jobs to be executed by the accelerator hardware into atomic compute tasks, using a scheduler to individually schedule and / or migrate each task for execution to different machines and / or accelerators at job runtime, and using a computer daemon to allocate memory and execute kernels for each task on the respective machines and / or accelerators.

The invention provides a short message receiving control method and system, electronic equipment and a storage medium, the control method is applied to an intelligent terminal, and a short message daemon is adopted to scan a first actual state of a short message receiving program in a first monitoring time period; and in response to the first actual state representing that the short message receiving program is in a non-survival state, adopting the short message daemon to generate a first awakening instruction to awaken the short message receiving program. According to the method and the device, the two-way supervision and wake-up mechanism is constructed, the background process is set to send the heartbeat request at regular time to monitor the survivability of the program, and meanwhile, the strategy of uploading failure retry and breakpoint resume is adopted, so that the problems of unstable running, service interruption and the like caused by system killing or self abnormality of the program are solved; according to the method, the short message receiving program can be ensured to run stably without missing the transmission of the short message, meanwhile, the quick discovery and notification of the service abnormality are realized, and the continuity, the reliability and the operation and maintenance of the short message receiving and uploading service are improved.

The invention relates to a development machine packaging method and system in a K8S cluster, which is characterized in that development machine resources are defined based on CRD and a controller, configuration information is described, and automatic life cycle management is realized in combination with operators; cRD is deployed in a host cluster, and is synchronized to K8S space of each tenant through an exclusive synchronizer; an SSH access path is opened by means of node port type service resources and an exclusive layer 4 forwarding entry; a public network IP path is automatically allocated by using the load balancer type service resources in combination with a metal load balancer component; remote access and safety control of the Zhu service are realized through a gateway, a signature verification mechanism and a domain name system analysis function of list monitoring; and deploying an exclusive daemon process set at each host node.

A network device identifies, using a daemon of a userspace of the network device, a packet, and processes, using the daemon, the packet to determine metadata associated with the packet. The network device stores, using the daemon, the metadata and a key value in a data structure. The network device provides, using the daemon, the packet to a kernel of the network device via a virtual interface between the userspace and the kernel, which allows the network device, when using the kernel, to process the packet and to provide the packet to an application of the userspace via a socket. Storing the metadata and the key value in the data structure allows the network device, when using the application of the userspace, to retrieve the metadata from the data structure based on receiving the packet from the kernel via the socket.

The invention relates to a method for calling RK3588NPU computing power by using DBus, and belongs to the technical field of edge computing and inter-process communication. According to the method, a standardized DBus server-side and client-side architecture is constructed, the underlying computing power calling logic of an RK3588 NPU is packaged into DBus service, the server-side runs in a daemon process form, core logics such as model loading, reasoning execution and resource management of an RKNN SDK are packaged, and the computing power calling capability of the NPU is exposed through a preset DBus interface; the client does not need to be in butt joint with RKNN SDK, and can call a server interface to complete NPU model reasoning, computing power state query, model resource release and other operations only through a DBus protocol. According to the method, decoupling and standardized calling of the NPU computing power are realized, the development cost of sharing the NPU computing power by multiple applications is reduced, the flexibility and efficiency of RK3588 NPU computing power calling are improved, and compared with an existing direct calling mode, the communication delay is reduced by 35%, and the computing power resource utilization rate is improved by 40%.

The application discloses a service calling method and device, a storage medium and an electronic equipment. It relates to the field of financial technology. The method comprises the following steps: generating a service calling request through a software development kit deployed on the target application side; sending the service calling request to a target daemon process on the target application side through a standard input and output channel, so that the target daemon process sends the service calling request to a server; receiving response data returned by the server based on the service calling request through the target daemon process, and returning the response data to the target application. Through the application, the technical problem that the data security is relatively low because the calling interface needs to be exposed when the service is called in the related art is solved.

The invention discloses a network space surveying and mapping intelligent question-answering system based on a large language model, and belongs to the technical field of network space surveying and mapping. When the system is in a working state, natural language query of a user is converted into a standardized surveying and mapping platform API statement in real time by fine tuning a Qwen-7B model, and a historical record library is established; a surveying and mapping API is periodically called to obtain related data, semantic vectorization storage is performed by using a BGE-M3 model, and a double-storage mechanism based on a relational database and a vector database is constructed; in the query stage, query statement semantics are subjected to vectorization processing, and a retrieval result is optimized by adopting semantic and keyword two-stage matching; real-time data, historical context and a local knowledge base are integrated through a dynamic Prompt engine, and an intelligent analysis conclusion is output through a Qwen-32B model; meanwhile, a daemon process is deployed to regularly detect the data freshness, and the timeliness of the knowledge base is maintained.

A software service such as a daemon monitors traffic at virtual com ports of virtual machines to facilitate debugging of virtual functions in real time as the virtual functions take turns using a hardware resource in round-robin fashion. If the daemon detects that a kernel debugger has initiated a kernel debugging session for a particular virtual function via the com port, the monitor halts world switching between virtual functions allocated to the hardware resource. The daemon makes the virtual function indicated by the kernel debugger the active partition while the kernel debugger reads one or more registers of the hardware resource allocated to the virtual function. When the kernel debugger has completed reading the register(s), the daemon signals the host to resume world switching between the virtual functions.

The invention relates to the technical field of artificial intelligence, is applied to smart medical and financial scenes, and discloses a cluster security monitoring method, device, equipment and medium, and the method comprises the steps: creating a daemon set and associating the daemon set to an HIDS client container, and creating a copy of a corresponding Pod; creating a first deployment and associating the first deployment to a container of a callback server, creating a plurality of copies of Pod for running the same callback server process, creating a callback service and associating the callback service to the callback server; establishing and verifying dynamic admission controller configuration, and appointing a receiving end of a forwarding request and a resource type and an operation type of the request needing to be forwarded; the HTTPS service is started, the report service is created, the forwarded request is detected through the callback server, the request is forwarded to the HIDS server side through the HIDS client side, and the request is detected; and creating a second deployment and associating the second deployment to the probe container, sending an HTTPS request to the callback service and the report service according to a preset period, and judging whether the callback service and the report service are available or not.