A load balancing method based on power terminal CPU kernel sharing

Abstract

The application relates to the field of power terminal chips, in particular to a load balancing method based on power terminal CPU core sharing, which is provided with a main server, a user quantity detector is fixedly connected to an output end interface of the main server, the main server is signal-interconnected with the user quantity detector, a flow distributor is fixedly connected to an output end interface of the user quantity detector, the flow distributor is signal-interconnected with the user quantity detector, and a communicating vessel is fixedly connected to an output end interface of the flow distributor. When the main server detects that a single user is used, all the flow is distributed to the single user through the flow distributor, the speed of the single user is ensured, when the user quantity detector detects that multiple users are simultaneously used, all the flow is evenly distributed through the flow distributor, the flow used by each user is ensured to be the same, and the use speed of the users is maximally ensured.

Description

Technical Field

[0001] This invention relates to the field of power terminals, specifically a load balancing method based on shared CPU cores in power terminals. Background Technology

[0002] The CPU core is the central chip of the CPU, made of monocrystalline silicon. It is used to perform all calculations, receive / store commands, process data, etc., and is the core of digital processing.

[0003] As power systems evolve towards informatization and intelligentization, applications targeting power systems place higher demands on the performance of power terminal CPUs. Various CPU cores have fixed logical structures, with logical units such as L1 cache, L2 cache, execution units, instruction-level units, and bus interfaces arranged in a scientific manner. Given the significant increase in the volume of power control information processing, servers cannot distribute traffic evenly among multiple users. This can lead to users queuing for internet access, or some users having fast network speeds while others experience no network at all, resulting in a poor user experience. Summary of the Invention

[0004] The purpose of this invention is to provide a load balancing method based on shared CPU cores in power terminals to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A load balancing method based on shared CPU cores in power terminals includes a master server. A user count detector is fixedly connected to the output interface of the master server, and the master server and the user count detector are interconnected. A flow distributor is fixedly connected to the output interface of the user count detector, and the flow distributor is interconnected with the user count detector. A communication device is fixedly connected to the output interface of the flow distributor, and the communication device is interconnected with the flow distributor. Multiple individual users are fixedly connected to the output interface of the communication device. An expert module is provided between the communication device and the flow distributor, and the expert module is interconnected with an external controller.

[0007] Through the above technical solution, when the main server detects a single user using the service, the traffic distributor distributes all traffic to that single user to ensure their speed. When a user count detector detects multiple users using the service simultaneously, the traffic distributor distributes all traffic evenly to ensure that each user uses the same amount of traffic, maximizing user speed. If a problem is detected in the traffic detection, a signal is immediately sent to the expert module, which is an external independent host. When a fault occurs and traffic distribution problems lead to network outages, the expert module immediately reports the problem, allowing external staff to manually resolve it and ensure network connectivity.

[0008] Preferably, each individual user corresponds to a single CPU. The instruction register inside the CPU decodes the instructions issued by the user and transmits the signals. Through the communication device, the instructions are sent to the flow distributor for recognition and corresponding response.

[0009] The above technical solution effectively transmits user commands to the traffic distributor via a single CPU, facilitating the distribution of signals from the traffic distributor to the main server. The main server then distributes traffic evenly to the single CPU through the connector.

[0010] Preferably, the master server puts the signal into the instruction register, decodes the instruction, and then sends the number of decoded instructions to the user count detector for detection.

[0011] The above technical solution involves a main server compiling and decoding the code, followed by a user count detector to count the number of decoded codes, making it easy to quickly detect the number of users.

[0012] Preferably, after the instruction is decoded by the individual CPU, the instruction decoder splits and interprets the retrieved instruction according to a predetermined instruction format, identifies and distinguishes different instruction categories to obtain operands, and sends the operand information to the user count detection. The user count detection then feeds back the obtained results to the traffic distributor.

[0013] Through the above technical solution, the CPU collects user data and feeds it back to the user count detector. This facilitates on-demand allocation by the communication device and avoids uneven allocation. The CPU allows the encoding of the instruction set and executes instructions sequentially. It fetches one instruction at a time and waits for it to finish executing before starting the next one.

[0014] Preferably, the master server provides a virtual IP address to the outside world, and different individual CPUs in the cluster use different IP addresses. After receiving a request, the connector returns the request to the master server via IP address according to different load balancing algorithms.

[0015] Preferably, the communicator performs round-robin DNS polling, and multiple DNS A records are configured so that requests can be distributed to different master servers.

[0016] Preferably, the individual CPU uses NAT to translate its private address into a legitimate IP address and maps it to the connector.

[0017] Compared with the prior art, the beneficial effects of the present invention are:

[0018] 1. This load balancing method based on shared CPU cores of power terminals, when the master server detects a single user using the system, distributes all traffic to that single user through a traffic distributor to ensure the speed of that single user. When a user count detector detects multiple users using the system simultaneously, the traffic distributor distributes all traffic evenly to ensure that each user uses the same amount of traffic, maximizing the speed of user access.

[0019] 2. This load balancing method based on the shared CPU core of the power terminal collects user data and feeds it back to the user count detector. This facilitates on-demand allocation by the interconnecting device and avoids uneven distribution. The CPU allows the encoding of the instruction set and executes the instructions sequentially. It fetches one instruction at a time and waits for it to finish before starting the next one.

[0020] 3. This load balancing method based on shared CPU cores of power terminals uses a master server to compile and decode, and then a user count detector to detect the number of decoded components, which facilitates rapid detection of the number of users. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the overall structure of the present invention.

[0022] In the diagram: 1. Master server; 2. User count detector; 3. Traffic distributor; 4. Connector. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Please see Figure 1 As shown, the present invention provides a technical solution:

[0025] A load balancing method based on shared CPU cores in power terminals includes a master server 1. A user count detector 2 is fixedly connected to the output interface of the master server 1, and the master server 1 and the user count detector 2 are interconnected. A flow distributor 3 is fixedly connected to the output interface of the user count detector 2, and the flow distributor 3 is interconnected with the user count detector 2. A communication device 4 is fixedly connected to the output interface of the flow distributor 3, and the communication device 4 is interconnected with the flow distributor 3. Multiple individual users are fixedly connected to the output interface of the communication device 4. An expert module is provided between the communication device 4 and the flow distributor 3, and the expert module is connected to an external controller.

[0026] Through the above technical solution, when the main server 1 detects that a single user is using the service, the traffic distributor 3 distributes all the traffic to the single user to ensure the speed of the single user. When the user count detector 2 detects that multiple users are using the service at the same time, the traffic distributor 3 distributes all the traffic evenly to ensure that each user uses the same amount of traffic, thus maximizing the user's speed.

[0027] When a problem is detected in traffic detection, the signal is immediately sent to the expert module, which is an independent external host. The expert module is divided into several departments. When a fault occurs and traffic allocation is problematic, resulting in a network outage, the expert module immediately reports the problem. Based on the user's feedback, the problem is passed to the corresponding department, and external staff manually resolve it to ensure network connectivity.

[0028] Preferably, each individual user corresponds to a single CPU 5. The instruction register inside the CPU decodes the instructions issued by the user and transmits the signals. Through the communication device 4, the instructions are sent to the flow distributor 3 for recognition and corresponding response.

[0029] Through the above technical solution, the user's instructions can be effectively sent to the traffic distributor 3 by the function of a single CPU 5, so that the traffic distributor 3 can send signals to the main server 1. Through the function of the main server 1, the traffic is evenly distributed to the single CPU 5 through the connector 4.

[0030] Preferably, the master server 1 puts the signal into the instruction register, decodes the instruction, and then sends the number of decoded instructions to the user count detector 2 for detection.

[0031] The above technical solution uses a main server 1 to compile and decode, and a user count detector 2 to detect the number of decoded data, which facilitates the rapid detection of the number of users.

[0032] Preferably, after the instruction is decoded by the single CPU 5, the instruction decoder splits and interprets the retrieved instruction according to a predetermined instruction format, identifies and distinguishes different instruction categories to obtain operands, and sends the operand information to the user count detection. The user count detection then feeds back the obtained results to the traffic distributor 3.

[0033] Through the above technical solution, the CPU collects user data and feeds it back to the user count detector 2. This allows the communication device 4 to allocate data as needed, avoiding uneven distribution. The CPU allows the encoding of the instruction set and executes instructions sequentially. It fetches one instruction at a time and waits for it to finish executing before starting the next one.

[0034] Preferably, the master server 1 provides a virtual IP address to the outside world, and different individual CPUs 5 in the cluster use different IP addresses. After receiving a request, the connector 4 returns the request to the master server 1 via IP address according to different load balancing algorithms.

[0035] Preferably, the communication device 4 performs round-robin DNS polling, and multiple DNS A records are configured so that requests can be distributed to different master servers 1.

[0036] Preferably, the single CPU 5 uses NAT to translate its private address into a legal IP address and maps it to the connector 4.

[0037] In this embodiment, a load balancing method based on shared CPU cores in power terminals is used. When the master server 1 detects a single user using the system, the traffic distributor 3 distributes all traffic to that single user to ensure their speed. When the user count detector 2 detects multiple users using the system simultaneously, the traffic distributor 3 distributes all traffic evenly to ensure that each user uses the same amount of traffic, maximizing user speed. The single CPU 5 effectively sends user instructions to the traffic distributor 3, which then sends a signal to the master server 1. The master server 1 then distributes traffic evenly to the single CPU 5 via the connector 4. The master server 1 compiles and decodes the data, and the user count detector 2 detects the number of decoded data to quickly determine the number of users. The CPU collects user data and feeds it back to the user count detector 2, allowing the connector 4 to allocate traffic as needed and avoid uneven distribution. The CPU encodes the allowed instruction set and executes instructions sequentially, fetching one instruction at a time and waiting for it to finish before starting the next one.

[0038] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A load balancing method based on shared CPU cores in power terminals, comprising a master server (1), characterized in that: A user count detector (2) is fixedly connected to the output interface of the main server (1). The main server (1) and the user count detector (2) are interconnected by signal. A flow distributor (3) is fixedly connected to the output interface of the user count detector (2). The flow distributor (3) is interconnected by signal. A communication device (4) is fixedly connected to the output interface of the flow distributor (3). The communication device (4) is interconnected by signal. Multiple individual users are fixedly connected to the output interface of the communication device (4). An expert module is provided between the communication device (4) and the flow distributor (3). The expert module is connected by signal to an external controller. Each user corresponds to a single CPU (5). The instruction register inside the CPU decodes the instruction issued by the user and transmits the signal. Through the communication device (4), the instruction is sent to the flow distributor (3) for recognition and corresponding response. The master server (1) puts the signal into the instruction register, decodes the instruction, and then sends the number of decoded instructions to the user count detector (2) for detection. After the instruction decoding stage, the single CPU (5) retrieves the instruction. The instruction decoder then splits and interprets the retrieved instruction according to the predetermined instruction format, identifies and distinguishes different instruction categories to obtain operands, and sends the operand information to the user count detector (2). The user count detector (2) then feeds back the obtained results to the traffic distributor (3).

2. The load balancing method based on shared CPU cores in power terminals according to claim 1, characterized in that: The master server (1) provides a virtual IP address to the outside world. Different individual CPUs (5) in the cluster use different IP addresses. After receiving the request, the connector (4) returns the request to the master server (1) through the IP address according to different load balancing algorithms.

3. The load balancing method based on shared CPU cores in power terminals according to claim 1, characterized in that: The connector (4) performs a round-robin DNS polling and configures multiple DNS A records so that requests can be distributed to different master servers (1).

4. The load balancing method based on shared CPU cores in power terminals according to claim 1, characterized in that: The single CPU (5) uses NAT to convert the private address into a legal IP address and maps it to the connector (4).

Images