CN114827677B - Artificial intelligence analysis load balancing method and device - Google Patents
Artificial intelligence analysis load balancing method and device Download PDFInfo
- Publication number
- CN114827677B CN114827677B CN202210219861.3A CN202210219861A CN114827677B CN 114827677 B CN114827677 B CN 114827677B CN 202210219861 A CN202210219861 A CN 202210219861A CN 114827677 B CN114827677 B CN 114827677B
- Authority
- CN
- China
- Prior art keywords
- artificial intelligence
- intelligence analysis
- router
- gateway
- analysis task
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000004458 analytical method Methods 0.000 title claims abstract description 173
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 133
- 238000000034 method Methods 0.000 title claims abstract description 30
- 230000008707 rearrangement Effects 0.000 claims description 17
- 238000013508 migration Methods 0.000 abstract description 4
- 230000005012 migration Effects 0.000 abstract description 4
- 238000012545 processing Methods 0.000 description 8
- 238000011144 upstream manufacturing Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 3
- 238000004590 computer program Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012916 structural analysis Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/23—Processing of content or additional data; Elementary server operations; Server middleware
- H04N21/24—Monitoring of processes or resources, e.g. monitoring of server load, available bandwidth, upstream requests
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/60—Network structure or processes for video distribution between server and client or between remote clients; Control signalling between clients, server and network components; Transmission of management data between server and client, e.g. sending from server to client commands for recording incoming content stream; Communication details between server and client
- H04N21/63—Control signaling related to video distribution between client, server and network components; Network processes for video distribution between server and clients or between remote clients, e.g. transmitting basic layer and enhancement layers over different transmission paths, setting up a peer-to-peer communication via Internet between remote STB's; Communication protocols; Addressing
- H04N21/64—Addressing
- H04N21/6405—Multicasting
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/60—Network structure or processes for video distribution between server and client or between remote clients; Control signalling between clients, server and network components; Transmission of management data between server and client, e.g. sending from server to client commands for recording incoming content stream; Communication details between server and client
- H04N21/63—Control signaling related to video distribution between client, server and network components; Network processes for video distribution between server and clients or between remote clients, e.g. transmitting basic layer and enhancement layers over different transmission paths, setting up a peer-to-peer communication via Internet between remote STB's; Communication protocols; Addressing
- H04N21/647—Control signaling between network components and server or clients; Network processes for video distribution between server and clients, e.g. controlling the quality of the video stream, by dropping packets, protecting content from unauthorised alteration within the network, monitoring of network load, bridging between two different networks, e.g. between IP and wireless
- H04N21/64723—Monitoring of network processes or resources, e.g. monitoring of network load
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Computer Security & Cryptography (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses an artificial intelligence analysis load balancing method and device.A client initiates an artificial intelligence analysis request, a DR router firstly sends a scheduling message after receiving the artificial intelligence analysis request to acquire which gateway routers can bear the load, then carries out load sharing, and schedules the gateway routers to respectively bear split artificial intelligence analysis tasks. And when the gateway router is invalid, load migration is performed. According to the technical scheme, the artificial intelligent analysis task is distributed to each edge gateway router, and the service efficiency of each gateway router is improved.
Description
Technical Field
The application belongs to the technical field of artificial intelligence analysis, and particularly relates to an artificial intelligence analysis load balancing method and device.
Background
Machine vision is a branch of the rapid development of artificial intelligence. In short, machine vision is to use a machine instead of a human eye to make measurements and decisions. The machine vision system converts the shot target into an image signal through a machine vision product (namely an image shooting device, namely CMOS and CCD), and transmits the image signal to a special image processing system to obtain the form information of the shot target, and converts the form information into a digital signal according to the pixel distribution, brightness, color and other information; the image system performs various operations on the signals to extract the characteristics of the target, and then accurately identifies, effectively pushes and accurately guides the target according to the discrimination result. The most basic characteristic of the machine vision system is to improve the accuracy and automation degree of information identification. Through the identification and analysis of external characteristic information such as figures, articles and trademarks, the digital living space service of the novel social system is provided by effectively using the digital living space service in application scenes such as intelligent communities, intelligent business circles and virtual shopping in combination with LBS (location based service, location Based Services, LBS) and XR (Extended Reality) technologies.
The current artificial intelligent analysis of the machine vision is relatively centralized through centralized processing of a cloud center or collaborative processing of cloud edge ends, equipment which bears the task of the artificial intelligent analysis often needs more processing tasks, has higher performance requirements on the equipment, and is relatively expensive. Meanwhile, when the artificial intelligence analysis task is processed, the occupied time length is larger, and the real-time performance is poorer.
Disclosure of Invention
The purpose of the application is to provide an artificial intelligence analysis load balancing method and device, so as to overcome the defects caused by centralized processing in the prior art.
In order to achieve the above purpose, the technical scheme of the application is as follows:
the artificial intelligence analysis load balancing method is applied to a video stream multicast network, wherein the video stream multicast network comprises a video source, a DR router, a client and a gateway router corresponding to the client, and is characterized by comprising the following steps:
the method comprises the steps that a multicast video stream sent by a video source is distributed to a client through a multicast distribution tree, the client initiates an artificial intelligent analysis request, a destination IP address of the artificial intelligent analysis request is an IP address of the video source, and a source IP address is a multicast group IP address;
after receiving an artificial intelligent analysis request, the DR router sends an artificial intelligent analysis task scheduling message, wherein the destination IP address of the artificial intelligent analysis task scheduling message is a multicast group IP address, and the source IP address is a video source IP address;
after receiving the artificial intelligent analysis task scheduling message, any gateway router in the multicast group replies a registration message if the router can bear the artificial intelligent analysis task, wherein the destination IP address of the registration message is a video source, the source IP address is the multicast group IP address, and the registration message carries a gateway router ID;
after the DR router receives the registration message, if a plurality of gateway routers bear the same artificial intelligence analysis task, the DR router splits the artificial intelligence analysis task, sends an artificial intelligence analysis task rearrangement message and rearranges the split artificial intelligence analysis task to the plurality of gateway routers;
the gateway routers that receive the artificial intelligence analysis task rearrangement message each assume the distributed split artificial intelligence analysis task.
Further, the artificial intelligence analysis load balancing method further comprises the following steps:
the gateway routers that receive the artificial intelligence analysis task rearrangement message each undertake the arranged split artificial intelligence analysis task and then send registration messages.
Further, the artificial intelligence analysis load balancing method further comprises the following steps:
each gateway router sends the artificial intelligence analysis result to the DR router, and the DR router sends the artificial intelligence analysis result in a multicast mode.
Further, the artificial intelligence analysis load balancing method further comprises the following steps:
and when the client finds that the artificial intelligence analysis result corresponding to the artificial intelligence analysis task to be performed can be obtained from the multicast group, the client does not send the corresponding artificial intelligence analysis request.
Further, the artificial intelligence analysis load balancing method further comprises the following steps:
the DR router periodically detects whether a gateway router bearing an artificial intelligence analysis task is effective, if an ineffective gateway router is detected, an artificial intelligence analysis task arrangement message is sent, one gateway router is selected from the gateway routers replying to the registration message, an artificial intelligence analysis task rearrangement message is issued, and the selected gateway router is designated to bear the artificial intelligence analysis task of the ineffective gateway router.
The application also provides an artificial intelligence analysis load balancing device which comprises a processor and a memory storing a plurality of computer instructions, wherein the computer instructions realize the steps of the artificial intelligence analysis load balancing method when being executed by the processor.
According to the artificial intelligence analysis load balancing method and device, an artificial intelligence analysis request is initiated by a client, and a gateway router shares a load to execute artificial intelligence analysis work. After receiving the artificial intelligence analysis task, the DR router sends a scheduling message to acquire which gateway routers can bear the load, and then performs load sharing. And when the gateway router is invalid, load migration is performed. According to the technical scheme, the artificial intelligent analysis task is distributed to each edge gateway router, and the service efficiency of each gateway router is improved.
Drawings
Fig. 1 is a schematic diagram of a network structure according to an embodiment of the present application;
FIG. 2 is a flow chart of an artificial intelligence analysis load balancing method of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application will be further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application.
The artificial intelligence analysis load balancing method provided by the application can be applied to an application environment shown in figure 1. The network of the application environment includes a video source, a network connection device, and a client device that receives a video stream. The video source, i.e. the device for capturing or distributing video, such as the network cameras IPC1 and IPC2 in fig. 1, may also be a media server. The present application describes, as an example, routers among which a designated router DR, a gateway router GW connected to a client PC or IPC, and other routers connected to these DR and GW, all of which constitute the entire network. Routers in the network all support multicast, which may also be referred to as multicast routers, and have artificial intelligence analysis capabilities. The video stream transmitted by IPC is transmitted in a multicast manner, for example, (S, G). The client PC receives (S, G) the video stream in a multicast manner. The multicast network runs the common PIM SM multicast routing protocol. A multicast distribution tree from a video source to clients has been established. The analysis requirements of artificial intelligence are initiated by the client. All messages in the network contain router IDs (router identities), each router ID being unique and non-conflicting throughout the network.
In one embodiment, as shown in fig. 2, an artificial intelligence analysis load balancing method is provided and applied to a video stream multicast network, wherein the video stream multicast network comprises a video source, a DR router, a client and a gateway router corresponding to the client. The artificial intelligence analysis load balancing method comprises the following steps:
step S1, a multicast video stream sent by a video source is distributed to a client through a multicast distribution tree, the client initiates an artificial intelligent analysis request, the destination IP address of the artificial intelligent analysis request is the IP address of the video source, and the source IP address is a multicast group IP address.
The router in the network of the embodiment supports multicast and has the analysis capability of artificial intelligence. The multicast video stream sent from the video source such as IPC1 is distributed to the respective client PCs through the multicast distribution tree. The client PC initiates an artificial intelligent analysis request, the destination IP of the request message is a video source, the source IP is a multicast group G, the request message is reversely transmitted to the video source along the path of the video stream (S, G), namely, the request message is reversely transmitted to the video source along the multicast distribution tree, each multicast router sends the request message to an upstream router from the entry interface of the (S, G) table entry, and each multicast router on the multicast distribution tree can receive the request message. The forwarding line between the multicast sender and the receiver forms a multicast distribution tree, in this embodiment, the path of the video stream (S, G), where the multicast distribution tree includes a DR router and other multicast routers, and the multicast distribution tree is a mature technology in the art, which is not described herein.
The artificial intelligence analysis request destination IP address is the IP address of the video source, and the source IP address is the multicast group IP address, so that the video source can be reversely transmitted to the DR router along the multicast distribution tree. The traditional IP message only routes to the destination IP address, and the DR router in the multicast cannot be found.
For example, if the client PC1 has an artificial intelligence analysis task a, an artificial intelligence analysis request is initiated and transmitted along the path of the video stream (S, G) to the video source IPC 1.
Step S2, after receiving the artificial intelligent analysis request, the DR router sends an artificial intelligent analysis task scheduling message, wherein the destination IP address of the artificial intelligent analysis task scheduling message is a multicast group IP address, and the source IP address is a video source IP address.
The artificial intelligent analysis request is transmitted in the network and can necessarily reach the DR router, and the DR router sends an artificial intelligent analysis task scheduling message, wherein the destination IP address of the artificial intelligent analysis task scheduling message is a multicast group IP address, and the source IP address is a video source IP address. The artificial intelligence analysis task schedule message carries artificial intelligence analysis task information such as content and identification of the artificial intelligence analysis task, etc., mainly to facilitate the router receiving the message to know what artificial intelligence analysis task is.
And step S3, after receiving the artificial intelligent analysis task scheduling message, any gateway router in the multicast group replies a registration message if the router can bear the artificial intelligent analysis task, wherein the destination IP address of the registration message is a video source, the source IP address is the multicast group IP address, and the registration message carries the gateway router ID.
The artificial intelligence analysis task scheduling message is sent along the multicast distribution tree, and each gateway router in the same multicast group can receive the artificial intelligence analysis task scheduling message.
For example, the gateway router GW1 connected to the client PC1 receives the artificial intelligence analysis task arrangement message; the gateway router GW2 connected to the client PC2 also receives the artificial intelligence analysis task arrangement message. If the multicast group further includes other gateway routers, such as GW3, the artificial intelligence analysis task can also be received, which is not described herein.
The gateway router is connected with the client, and can automatically identify itself as the gateway router. In addition, when the gateway router itself is both a PIM router and an IGMP querier, it is the gateway router itself. How to determine itself to be a gateway router is a relatively mature technology in the art, and will not be described in detail herein.
The gateway router receiving the artificial intelligence analysis task scheduling message can judge whether the gateway router is suitable for executing the artificial intelligence analysis task in the scheduling message or not, for example, the gateway router can judge according to the self capacity and the load condition, and if the gateway router is suitable for executing, the gateway router replies the registration message.
For example, GW1 and GW2 reply to the registration message, indicating to the multicast router upstream of the multicast distribution tree that itself can afford to arrange the artificial intelligence analysis task specified by the message, the registration message carrying the identity of the artificial intelligence analysis task and the identity of the gateway router itself.
For artificial intelligence analysis task a, gateway router GW1 and gateway router GW2 both reply to the registration message after receiving the arrangement message. The registration message replied by the gateway router GW1 carries: artificial intelligence analysis task A, GW. The registration message replied by the gateway router GW2 carries: artificial intelligence analysis task A, GW. The destination IP address of the registration message is video source IPC1, the source IP is multicast group IP address G, and the registration message is transmitted to the video source along the path of the video stream (S, G).
It should be noted that, the client PC initiates an artificial intelligence analysis request, for example, performing structural analysis on the video image, or performing intelligent recognition on a target in the video image, which is not limited in the specific content of the artificial intelligence analysis. If the gateway router does not support the corresponding artificial intelligence analysis task, the registration message is not replied. Or the gateway router can not increase the task any more due to the self-load problem, and does not reply to the registration message.
The registration message is passed to the DR router, the steps of this embodiment make the DR router in the network aware that the artificial intelligence analysis task can be undertaken by the gateway routers GW1, GW2.
And S4, after the DR router receives the registration message, if a plurality of gateway routers bear the same artificial intelligence analysis task, splitting the artificial intelligence analysis task, sending an artificial intelligence analysis task rearrangement message, and rearranging the split artificial intelligence analysis task to the plurality of gateway routers.
When DR routers in the network receive registration messages and find that a plurality of gateway routers can bear an artificial intelligence analysis task, the DR routers split the artificial intelligence analysis task and send an artificial intelligence analysis task rearrangement message.
For example, DR router receives registration messages from gateway routers GW1 and GW2 indicating that they can assume artificial intelligence analysis task a, then split artificial intelligence analysis task a into subtasks A1 and A2, where A1 is assigned to GW1 and A2 is assigned to GW2.
The artificial intelligence analysis task rearrangement message may carry the assigned artificial intelligence task information and an identification of the gateway router that is to assume the artificial intelligence task, thereby facilitating the gateway router that receives the message to determine whether it is itself to assume the assigned task.
It is also possible that only one gateway router replies to the registration message, and the DR router directly issues the rearrangement message without task splitting, designating the network router to assume the artificial intelligence analysis task. When splitting tasks, the proportion of the split subtasks can be the same or different. For example, subtask A1 and subtask A2 each account for 50% of total task A, or A1 accounts for all of total task A, while A2 is 0. The proportion of the subtasks may also be determined by referring to the performance and load of the respective gateway router, and will not be described here.
And S5, the gateway routers receiving the artificial intelligence analysis task rearrangement message respectively bear the arranged split artificial intelligence analysis tasks.
In this embodiment, the artificial intelligence analysis task rearrangement message carries the assigned subtask and the identifier of the gateway router that bears the subtask, so that the gateway routers each bear the corresponding subtask after receiving the artificial intelligence analysis task rearrangement message. GW1 assumes subtask A1 and GW2 assumes subtask A2.
And then, the gateway router can reply a registration message, and the multicast router at the upstream of the multicast distribution tree is indicated to bear the designated artificial intelligence analysis task by itself, wherein the registration message carries the identifier of the borne artificial intelligence analysis task and the identifier of the gateway router.
All multicast routers along the multicast distribution tree, such as upstream routers of gateway router GW1, can receive information "gateway router GW1 is assuming artificial intelligence analysis task A1" up to DR router. So that each multicast router upstream knows the artificial intelligence analysis task undertaken by each gateway router.
According to the technical scheme, each gateway router averagely shares the artificial intelligence analysis task. For example, the client PC1 sends 10 artificial intelligence analysis tasks, while the client PC2 does not have an artificial intelligence analysis task, and the gateway router GW2 is idle, so that by the technical scheme of the present application, the GW1 and GW2 share the artificial intelligence analysis task.
For the same artificial intelligence analysis task A, if after the client PC1 initiates the request, the GW1 and the GW2 respectively bear the A1 and the A2, after the client PC2 initiates the same artificial intelligence analysis task request, the task A is found to be split into the A1 and the A2, the DR router does not need to send the artificial intelligence analysis task arrangement message any more, and the client PC2 directly receives the analysis result.
In a specific embodiment, the artificial intelligence analysis load balancing method further includes:
each gateway router sends the artificial intelligence analysis result to the DR router, and the DR router sends the artificial intelligence analysis result in a multicast mode.
Considering that the characteristics of the (S, G) multicast entry do not allow the (S, G) multicast packet to be received from the outbound interface, GW1 and GW2 unicast the analysis result to DR, and DR forwards the analysis result in a multicast (S, G) manner, and at this time, client PC1 receives the analysis results of task A1 and task A2. Of course, the client PC2 will also receive the analysis result, it may choose to discard, but if it needs to analyze the result of task a, it will find that the corresponding analysis result can already be received, i.e. it will not need to send an analysis request any more.
In another specific embodiment, the artificial intelligence analysis load balancing method further includes:
the DR router periodically detects whether a gateway router bearing an artificial intelligence analysis task is effective, if an ineffective gateway router is detected, an artificial intelligence analysis task arrangement message is sent, one gateway router is selected from the gateway routers replying to the registration message, an artificial intelligence analysis task rearrangement message is issued, and the selected gateway router is designated to bear the artificial intelligence analysis task of the ineffective gateway router.
Specifically, since the DR router has split the task a into A1 and A2, the DR router has the responsibility to monitor the working states of the GW1 and GW2, so that the migration of the task can be arranged when the analysis execution of one of the routers fails.
In this embodiment, the DR router periodically multicasts the task scheduling message to the gateway routers GW1 and GW2, adds router IDs of GW1 and GW2 in the message, and when the GW1 and GW2 find that there is an ID of the router itself after receiving the message, replies a registration message to the DR router. If GW2 does not reply to the registration message for 3 consecutive periods and GW2 does not send any analysis result to the DR router during this period, the DR router considers GW2 to have failed, requiring task migration.
Since the DR router does not know which routers can take on the analysis task exist, a scheduling message is sent by multicast, and the information of "analysis task A2" is included. The message will arrive at each gateway router, and if GW1 and GW3 can assume analysis task A2, the registration message is replied to the DR router, which selects one of the routers, for example GW3, according to a policy, which may be selected according to the task amount, execution time, performance index, etc. of each router stored by itself. The DR router sends a rearrangement message to gateway router 3 asking GW3 to assume task A2.
The DR router of this embodiment also continuously sends a number of scheduling cancellation messages to GW2 informing GW2 to cancel the burden on task A2. This is to prevent GW2 from possibly continuing the analysis task because the task is relaxed, resulting in a repetition being scheduled and multiple copies of repeated data being received by the client.
In another embodiment, the present application also provides an artificial intelligence analysis load balancing apparatus, including a processor and a memory storing a number of computer instructions which, when executed by the processor, implement the steps of the artificial intelligence analysis load balancing method.
For specific limitations of the artificial intelligence analysis load balancing apparatus, reference may be made to the above limitations of the artificial intelligence analysis load balancing method, and no further description is given here. The artificial intelligence analysis load balancing device described above may be implemented in whole or in part by software, hardware, and combinations thereof. May be embedded in hardware or may be independent of a processor in the computer device, or may be stored in software in a memory in the computer device, so that the processor invokes the corresponding operations.
The memory and the processor are electrically connected directly or indirectly to each other for data transmission or interaction. For example, the components may be electrically connected to each other by one or more communication buses or signal lines. The memory stores a computer program that can be executed on a processor that implements the network topology layout method in the embodiment of the present invention by executing the computer program stored in the memory.
The Memory may be, but is not limited to, random access Memory (Random Access Memory, RAM), read Only Memory (ROM), programmable Read Only Memory (Programmable Read-Only Memory, PROM), erasable Read Only Memory (Erasable Programmable Read-Only Memory, EPROM), electrically erasable Read Only Memory (Electric Erasable Programmable Read-Only Memory, EEPROM), etc. The memory is used for storing a program, and the processor executes the program after receiving an execution instruction.
The processor may be an integrated circuit chip having data processing capabilities. The processor may be a general-purpose processor including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), and the like. The methods, steps and logic blocks disclosed in the embodiments of the present invention may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The above examples merely represent a few embodiments of the present application, which are described in more detail and are not to be construed as limiting the scope of the invention. It should be noted that it would be apparent to those skilled in the art that various modifications and improvements could be made without departing from the spirit of the present application, which would be within the scope of the present application. Accordingly, the scope of protection of the present application is to be determined by the claims appended hereto.
Claims (6)
1. The artificial intelligence analysis load balancing method is applied to a video stream multicast network, wherein the video stream multicast network comprises a video source, a DR router, a client and a gateway router corresponding to the client, and is characterized by comprising the following steps:
the method comprises the steps that a multicast video stream sent by a video source is distributed to a client through a multicast distribution tree, the client initiates an artificial intelligent analysis request, a destination IP address of the artificial intelligent analysis request is an IP address of the video source, and a source IP address is a multicast group IP address;
after receiving an artificial intelligent analysis request, the DR router sends an artificial intelligent analysis task scheduling message, wherein the destination IP address of the artificial intelligent analysis task scheduling message is a multicast group IP address, and the source IP address is a video source IP address;
after receiving the artificial intelligent analysis task scheduling message, any gateway router in the multicast group replies a registration message if the router can bear the artificial intelligent analysis task, wherein the destination IP address of the registration message is a video source, the source IP address is the multicast group IP address, and the registration message carries a gateway router ID;
after the DR router receives the registration message, if a plurality of gateway routers bear the same artificial intelligence analysis task, the DR router splits the artificial intelligence analysis task, sends an artificial intelligence analysis task rearrangement message and rearranges the split artificial intelligence analysis task to the plurality of gateway routers;
the gateway routers that receive the artificial intelligence analysis task rearrangement message each assume the distributed split artificial intelligence analysis task.
2. The artificial intelligence analysis load balancing method of claim 1, further comprising:
the gateway routers that receive the artificial intelligence analysis task rearrangement message each undertake the arranged split artificial intelligence analysis task before sending a registration message.
3. The artificial intelligence analysis load balancing method of claim 1, further comprising:
each gateway router sends the artificial intelligence analysis result to the DR router, and the DR router sends the artificial intelligence analysis result in a multicast mode.
4. The artificial intelligence analysis load balancing method of claim 1, further comprising:
and when the client finds that the artificial intelligence analysis result corresponding to the artificial intelligence analysis task to be performed can be obtained from the multicast group, the client does not send the corresponding artificial intelligence analysis request.
5. The artificial intelligence analysis load balancing method of claim 1, further comprising:
the DR router periodically detects whether a gateway router bearing an artificial intelligence analysis task is effective, if an ineffective gateway router is detected, an artificial intelligence analysis task arrangement message is sent, one gateway router is selected from the gateway routers replying to the registration message, an artificial intelligence analysis task rearrangement message is issued, and the selected gateway router is designated to bear the artificial intelligence analysis task of the ineffective gateway router.
6. An artificial intelligence analysis load balancing device comprising a processor and a memory storing a number of computer instructions which when executed by the processor implement the steps of the method of any one of claims 1 to 5.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210219861.3A CN114827677B (en) | 2022-03-08 | 2022-03-08 | Artificial intelligence analysis load balancing method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210219861.3A CN114827677B (en) | 2022-03-08 | 2022-03-08 | Artificial intelligence analysis load balancing method and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114827677A CN114827677A (en) | 2022-07-29 |
CN114827677B true CN114827677B (en) | 2024-02-09 |
Family
ID=82528756
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210219861.3A Active CN114827677B (en) | 2022-03-08 | 2022-03-08 | Artificial intelligence analysis load balancing method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114827677B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115426221A (en) * | 2022-10-14 | 2022-12-02 | 湖南省邮电规划设计院有限公司 | Gateway device of Internet of things |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101651609A (en) * | 2008-08-14 | 2010-02-17 | 华为技术有限公司 | Method and device for realizing multicast load sharing |
CN102238092A (en) * | 2011-08-02 | 2011-11-09 | 杭州华三通信技术有限公司 | Method for performing load sharing on encoder and encoder |
CN113239792A (en) * | 2021-05-11 | 2021-08-10 | 深圳市安软科技股份有限公司 | Big data analysis processing system and method |
CN113485842A (en) * | 2021-07-30 | 2021-10-08 | 浙江大华技术股份有限公司 | Method and device for analyzing data based on device cluster |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10567248B2 (en) * | 2016-11-29 | 2020-02-18 | Intel Corporation | Distributed assignment of video analytics tasks in cloud computing environments to reduce bandwidth utilization |
-
2022
- 2022-03-08 CN CN202210219861.3A patent/CN114827677B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101651609A (en) * | 2008-08-14 | 2010-02-17 | 华为技术有限公司 | Method and device for realizing multicast load sharing |
CN102238092A (en) * | 2011-08-02 | 2011-11-09 | 杭州华三通信技术有限公司 | Method for performing load sharing on encoder and encoder |
CN113239792A (en) * | 2021-05-11 | 2021-08-10 | 深圳市安软科技股份有限公司 | Big data analysis processing system and method |
CN113485842A (en) * | 2021-07-30 | 2021-10-08 | 浙江大华技术股份有限公司 | Method and device for analyzing data based on device cluster |
Also Published As
Publication number | Publication date |
---|---|
CN114827677A (en) | 2022-07-29 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US9258243B2 (en) | Symmetric service chain binding | |
CN108293009B (en) | Software defined data center and scheduling method of service cluster in software defined data center | |
US7580368B2 (en) | Packet distribution control method | |
CN100433730C (en) | Method and system of multicast and video-on-demand | |
US20020143951A1 (en) | Method and system for multicast to unicast bridging | |
CN105897444B (en) | Multicast group management method and device | |
US20140122704A1 (en) | Remote port mirroring | |
US20060146821A1 (en) | Virtual multicast routing for a cluster having state synchronization | |
CN101483655B (en) | Packet transmission method and proxy device for Internet group management protocol | |
CN105099898B (en) | A kind of PPPOE message forwarding methods and BRAS servers | |
CN102377640B (en) | Message processing apparatus, message processing method and preprocessor | |
KR102685010B1 (en) | Account linking methods and devices, storage media and electronic devices | |
CN103117935A (en) | Multicast data forwarding method and multicast data forwarding device applied to multi-homing networking | |
CN101605103A (en) | A kind of method of static forwarding multicast data and device | |
CN104468371B (en) | multicast service message processing method and device | |
CN114827677B (en) | Artificial intelligence analysis load balancing method and device | |
CN105743781B (en) | A kind of VRRP load-balancing method and device | |
CN115037785B (en) | Instant communication system and method | |
CN110995609A (en) | Message sending method and device, electronic equipment and storage medium | |
CN107948273B (en) | SDN-based load sharing and secure access method and system | |
US20120191873A1 (en) | Relay apparatus, communication network system, and load distribution method | |
CN114286127B (en) | Distributed artificial intelligence analysis method and device | |
CN105684386A (en) | Device and method for managing subscriptions to point-to-multipoint transmissions | |
US11025444B2 (en) | Multicast traffic optimization in multihomed edge network elements | |
CN108512671A (en) | A kind of outer layer multicast ip address distribution method and device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |