13 results about "Packet injection" patented technology

The invention relates to the technical field of congestion detection and solution of a high-performance network-on-chip, and relates to a congestion detection and self-adaptive control method and system of a 2D (two-dimensional) network-on-chip (NoC). The method comprises the following steps: based on a 2D mesh NoC topology, generating a data packet injection rate of each node by using a Poisson distribution flow model, and modeling and recording a buffer load value and an arbitration waiting request number through an NoC simulator; classifying nodes according to node centrality annular distribution characteristics, and initializing a threshold register value; when congestion control is input, an input blocking index is periodically calculated, and a threshold value register is dynamically adjusted according to a threshold value; a congestion control path is output, a data packet is redirected when an output busy signal is activated, and a new request is blocked if all output ports are busy; through off-line estimation optimization and on-line detection adjustment, NoC congestion is effectively solved, and system performance and reliability are improved.

The invention discloses a performance test method and device for an Ethernet switch and a medium, and belongs to the field of Ethernet switch testing, and the method comprises the steps: constructing a dynamic random data generation model of which the test process fluctuates along with a time period; generating a test signal of which the test data volume fluctuates along with the time period; constructing a state control model of the Ethernet switch service based on a queuing theory, and adjusting the rate of injecting the test data packet into the Ethernet switch in real time through the state control model; screening an output test data packet which is most matched with each test data packet, and calculating the performance coefficient of the Ethernet switch tested by each test data packet; and calculating a performance coefficient average value of the Ethernet switch, setting a performance coefficient threshold value of the Ethernet switch, and evaluating the tested performance of the Ethernet switch. According to the method, the consistency of multi-node test data can be realized, the concurrent test of ten thousand-level ports is supported, and effective decision support is provided for configuration optimization of the Ethernet switch.

The invention relates to the technical field of data processing, and provides a hybrid CAN communication analog data processing method and system based on a TCP protocol, and the method comprises the steps: generating fault snapshot data when an operation abnormality occurs; establishing a virtual TCP communication link and a virtual CAN communication link according to the fault snapshot data, and configuring a virtual DCDC module; injecting the TCP communication data stream into the virtual TCP communication link, and reconstructing the TCP communication data stream into a CAN message sequence; capturing a response message, and converting the response message into a TCP data packet; injecting a diagnosis instruction, a monitoring behavior and a communication data flow to obtain behavior information after injection; comparing the injected behavior information with an expected behavior; recording changes of all communication data and system states in the digital mirror image test environment, analyzing the recorded communication data and system states, and positioning a fault source; and based on the root of the fault, deploying the verified diagnosis instruction to a real production system. The method has the effect of improving the real-time recovery capability of the production system.

The invention relates to a method for testing network performance based on UDP (User Datagram Protocol) bagging, which comprises the following steps: a client obtains a timestamp CT1 when the client sends a data packet, packages the timestamp CT1 into a UDP data packet and sends the UDP data packet to a server; the server side receives the UDP data packet sent by the client side and records the timestamp ST1 of the server side when the server side receives the data packet; the server side analyzes the UDP data packet, increases a timestamp ST2 when the server side sends the data packet, and packs the data packet and sends the data packet to the client side; the client receives the UDP data packet returned by the server, analyzes the content of the packet, and obtains a timestamp CT2 when the client receives the data packet; and calculating round-trip delay CST and unidirectional delay CT from the client to the server. According to the method and device for testing the network performance based on UDP bagging, the processor and the computer readable storage medium, the delay variance between the client side and the server side is obtained through packet sending and packet receiving, the packets can be received and sent in an uninterrupted pulse mode, in this way, the network state between the client side and the server side can be fed back in real time, and the user experience is improved. And technical information support is provided for long-time network operation.

The application relates to the technical field of data processing, and provides a mixed CAN communication analog data processing method and system based on a TCP protocol, which comprises the following steps: generating fault snapshot data when an operation exception occurs; establishing a virtual TCP communication link and a virtual CAN communication link according to the fault snapshot data and configuring a virtual DCDC module; injecting a TCP communication data stream into the virtual TCP communication link and reconstructing the TCP communication data stream into a CAN message sequence; capturing a response message and converting the response message into a TCP data packet; injecting a diagnosis instruction, monitoring behaviors and communication data streams, and obtaining post-injection behavior information; comparing the post-injection behavior information with expected behaviors; recording changes of all communication data and system states in a digital mirror test environment, analyzing the recorded communication data and system states, and locating a fault root; and deploying a verified diagnosis instruction to an actual production system based on the fault root. The application has the effect of improving the real-time recovery capability of a production system.

The invention discloses a multi-source data time sequence synchronization method and system for magnetic drive rotor twin control, and the method comprises the steps: multi-source heterogeneous data collection and time mark injection: collecting multi-source heterogeneous data at a data collection end of a magnetic drive rotor control system, and injecting a time mark for each data packet; data transmission and delay calculation: transmitting the data packets with the time scales to a time sequence synchronization center, and calculating and recording the transmission delay of each data packet in real time to obtain a delay historical queue of each data source; in the time sequence synchronization center, delay prediction of a data packet and self-adaptive buffer alignment of data are carried out to obtain a synchronous data sequence with a unified time scale; distributing the synchronous data sequence with the unified time scale; the system is used for implementing the method. According to the method, the problem of data time sequence misalignment caused by transmission and processing delay uncertainty in a heterogeneous network environment is effectively solved, and microsecond-level high-precision synchronization is realized.

The embodiment of the invention provides an automatic delivery system and method of an application product, which can package a customization demand input by a user into an HTTPS request data packet after receiving the customization demand on an aPaaS webpage, send the HTTPS request data packet to a transit server, and trigger a CI / CD assembly line task preset by a GitLab server, thereby realizing automatic delivery of the application product. The method comprises the following steps: receiving an HTTPS request data packet from an aPaaS webpage, so as to inject the HTTPS request data packet into a GitLab server through a transit server to generate a corresponding APP, and feeding back an execution result to the aPaaS webpage through a Webhook mechanism, so that a downloading instruction input by a user is received through the aPaaS webpage, and a corresponding application product is downloaded from an assembly line product library of the GitLab server through a GitLab API. The problem that a GitLab server is difficult to realize complex user customized logic and real-time component linkage through a graphical interface, so that the user experience is poor is solved.

This application provides an automated delivery system and method for application artifacts. After receiving customized requests from users on an aPaaS webpage, the system encapsulates these requests into HTTPS request packets and sends them to a relay server. This triggers a pre-defined CI / CD pipeline task on the GitLab server. The relay server then injects the HTTPS request packets into the GitLab server to generate the corresponding application. The execution result is fed back to the aPaaS webpage via a Webhook mechanism. The aPaaS webpage then receives download commands from users and downloads the corresponding application artifacts from the GitLab server's pipeline artifact library via the GitLab API. This addresses the problem of GitLab servers struggling to implement complex user customization logic and real-time component linkage through a graphical interface, resulting in a poor user experience.

The invention discloses a data transmission method and device, equipment, a medium and a product. The method is applied to a system comprising a plurality of processors, each processor in the plurality of processors deploys an operating system, the operating systems deployed on the plurality of processors are different, and the method comprises the following steps: writing a first data packet to be sent into a shared memory through a first operating system, sending a first inter-core interrupt signal to a data transmission module in the core of the second operating system, wherein the first inter-core interrupt signal carries the identifier of the first data packet; acquiring the first data packet from the shared memory based on the identifier of the first data packet through the data transmission module, and sending the first data packet to a network protocol stack interface module of the second operating system; and injecting the first data packet into a kernel network protocol stack of the second operating system through a network protocol stack interface module, so that the network protocol stack sends the first data packet to a physical network card. According to the method, the inter-core data transmission efficiency of multiple processors can be improved.

This invention discloses a high-efficiency task scheduling system for heterogeneous multi-core on-chip networks based on a multi-armed slot machine, relating to the field of task scheduling in heterogeneous multi-core systems. It addresses the problem of high power consumption in existing scheduling technologies by proposing this solution. During runtime, the following scheduling steps are periodically executed using time slices as the cycle: S1. Status acquisition of the underlying hardware platform and upper-layer application load; S2. Time and power consumption estimation based on on-chip network prediction; S3. Calculation of priority index and net gain; S4. Allocation of heterogeneous cores based on integer linear programming; S5. Runtime closed-loop feedback and control parameter updates. The advantage is that when the scheduler calculates the net gain, it imposes a strong mathematical penalty on applications with high packet injection rates that excessively distribute core allocation. This forces the system to perceive NoC traffic hotspots during runtime and intelligently converge applications with high communication demands to core configurations that are physically closer. This avoids problems such as virtual channel deadlock and bus congestion in NoC routers.

This invention discloses a performance testing method, device, and medium for Ethernet switches, belonging to the field of Ethernet switch testing. The method includes: constructing a dynamic random data generation model that reflects the fluctuation of the testing process over time; generating test signals that reflect the fluctuation of test data volume over time; constructing a state control model for Ethernet switch services based on queuing theory, and adjusting the rate at which test data packets are injected into the Ethernet switch in real time using the state control model; selecting the output test data packet that best matches each test data packet, and calculating the performance coefficient of the Ethernet switch tested for each test data packet; calculating the average performance coefficient of the Ethernet switch, setting a performance coefficient threshold for the Ethernet switch, and evaluating the performance of the tested Ethernet switch. This method can achieve consistency of test data across multiple nodes, supports concurrent testing of tens of thousands of ports, and provides effective decision support for the configuration optimization of Ethernet switches.