CN116489096A - Gateway load balancing algorithm optimization method - Google Patents
Gateway load balancing algorithm optimization method Download PDFInfo
- Publication number
- CN116489096A CN116489096A CN202310290101.6A CN202310290101A CN116489096A CN 116489096 A CN116489096 A CN 116489096A CN 202310290101 A CN202310290101 A CN 202310290101A CN 116489096 A CN116489096 A CN 116489096A
- Authority
- CN
- China
- Prior art keywords
- server
- weight
- value
- load
- real
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 26
- 238000005457 optimization Methods 0.000 title abstract description 8
- 238000004364 calculation method Methods 0.000 claims abstract description 22
- 230000015654 memory Effects 0.000 claims description 33
- 230000005484 gravity Effects 0.000 claims description 7
- 230000005540 biological transmission Effects 0.000 claims description 3
- 230000002159 abnormal effect Effects 0.000 abstract description 3
- 230000002093 peripheral effect Effects 0.000 description 3
- 230000003068 static effect Effects 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000002922 simulated annealing Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/12—Avoiding congestion; Recovering from congestion
- H04L47/125—Avoiding congestion; Recovering from congestion by balancing the load, e.g. traffic engineering
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/10—Active monitoring, e.g. heartbeat, ping or trace-route
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/50—Reducing energy consumption in communication networks in wire-line communication networks, e.g. low power modes or reduced link rate
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Cardiology (AREA)
- General Health & Medical Sciences (AREA)
- Debugging And Monitoring (AREA)
Abstract
The invention discloses a gateway load balancing algorithm optimization method, which comprises the following steps: s1, collecting a value of a load factor; s2, comparing the value of each load factor with a corresponding threshold value, S3, calculating the processing capacity extreme value N of the server (i) And server real-time load rate R (i) Is a value of (2); s4, each server R (i) The value of (2) and the threshold value Y corresponding to the server (i) S5, calculating the weight C by the calculation server (i) S6, calculating the real distribution weight CW between the servers (i) According to the weight polling distribution task, the invention adopts a weighted polling algorithm to process data, simultaneously reads network request data in a near-time period in an asynchronous mode, analyzes the quantity and the duration of the processed data and abnormal conditions, and gives outAdjusting the optimal weight proportion; the weight of the optimized algorithm is obtained according to the load condition of the server, the floating condition of the server can be reflected, and the server resources are utilized effectively.
Description
Technical Field
The invention relates to the technical field of weighted polling algorithms, in particular to a gateway load balancing algorithm optimization method.
Background
Polling (Polling) is a way for a CPU to decide how to provide services to peripheral devices, and is also called "Programmed I/O". The concept of the polling method is that a CPU issues a query at regular time, sequentially queries whether each peripheral device needs its service, gives the service immediately, and then asks the next peripheral after the service is finished, and then repeats.
The weighted polling algorithm is an improvement on the polling algorithm, the difference among the servers is represented by the weight, the servers with high weight are allocated more tasks, and the servers with low weight are allocated less tasks.
In the prior art, the weighted polling algorithm solves the problem of difference among servers, is relatively simple to realize, does not need to record the current connection state, and is a stateless connection. However, the weight is set manually, and the difference between servers cannot be dynamically reflected; the dynamic adjustment of the node performance according to the use conditions of the CPU, the memory, the disk IO and the network bandwidth is inconvenient, and the existing gateway load balancing algorithms are too stiff to be quickly adjusted to be suitable for the current situation.
Through retrieval, the patent with the application number of CN202111119816.2 discloses a load balancing method based on dynamic and static weighted polling, and relates to the fields of servers, load balancing, deep learning and the like. Firstly, collecting performance parameters of each node of a server cluster, and generating a node performance weight; secondly, calculating an interval threshold according to the running condition of the server; when the load balancing server receives a load request, judging that the cluster load exceeds the interval threshold, and carrying out static weighted polling on the server load according to the node performance weight; otherwise, dynamically adjusting the performance weight of the server node, and carrying out dynamic weighted polling on the server load according to the adjusted node performance weight. Compared with the traditional load balancing method, the method for calculating the cluster load threshold by the simulated annealing algorithm improves the problems of low efficiency under low load and unstable efficiency under high load.
The above scheme is not convenient for dynamically adjusting the node performance according to the use conditions of the CPU, the memory, the disk IO and the network bandwidth, and the existing gateway load balancing algorithms are too stiff to be quickly adjusted to be suitable for the current situation, so we need to propose a gateway load balancing algorithm optimization method.
Disclosure of Invention
The invention aims to provide a gateway load balancing algorithm optimization method, which adopts a weighted polling algorithm to process data, simultaneously reads network request data in a near-time period in an asynchronous mode, analyzes the quantity and the duration of the processed data and abnormal conditions, and gives out an optimal weight proportion; the weight of the optimized algorithm is obtained according to the load condition of the server, the floating condition of the server can be reflected, and the server resources are effectively utilized, so that the problems in the background technology are solved.
In order to achieve the above purpose, the present invention provides the following technical solutions: a gateway load balancing algorithm optimization method comprises the following steps:
s1, collecting a value of a load factor;
s2, comparing the value of each load factor with a corresponding threshold value, if L exists (ci) >Y (ci) 、L (mi) >Y (mi) Or L (ni) >Y (ni) The weight of the server is set to 0, the server is not allocated with tasks in the period, otherwise, the weight vector T according to the load factor (i) Computing server real-time load M (i) Is a value of (2);
s3, calculating the processing capacity extreme value N of the server (i) And server real-time load rate R (i) Is a value of (2);
s4, each server R (i) The value of (2) and the threshold value Y corresponding to the server (i) Comparing if R (i) >Y (i) The service is provided withThe weight of the device is set to 0, otherwise, the next step is carried out;
s5, calculating the weight C by the calculation server (i) And according to the weight W of the server (i) Calculating server true allocation weight CW (i) ;
S6, calculating the real distribution weight CW between the servers (i) And (3) obtaining the weight of the server, and polling and distributing tasks according to the weight.
Preferably, in S2-S6, Y (ci) 、Y (mi) And Y (ni) Threshold values respectively representing CPU, memory and network of server, Y (i) Representing the threshold of the server, N (i) Representing server processing capability extremum, M (i) Representing real-time load of server, R (i) Representing real-time load rate of server, T (i) Weight vector representing each load factor, W (i) Representing the weight of the server.
Preferably, in S1, the load factors include a CPU usage rate C (N i), a memory usage rate M (N i), a disk IO usage rate D (N i), and a network bandwidth usage rate B (N i); wherein,,
preferably, in formulas (1) - (4), Δu represents a difference in user state between two samples, Δs represents a difference in memory system between two samples, and Δj represents a difference in CUP utilization between two samples;
memotal represents memory, memfree represents free memory;
used represents Used disk space, block represents total disk space;
Δr represents the difference in reception bandwidth between the two samples, Δt represents the difference in transmission bandwidth between the two samples, Δt represents a period of time, total bw represents the port bandwidth, and the server uses the eth0 port, with a bandwidth of 100Mbps.
Preferably, in S3, the calculation formula of the server processing capability extremum is:
N (i) =P (i) *T (i) (5)
wherein P is (i) =[P (ci) ,P (mi) ,P (ni) ],P (ci) 、P (mi) And P (ni) Respectively representing the processing speed, the memory size and the network throughput of the CPU of the server;
T (i) =[T (c) ,T (m) ,T (n) ],T (c) 、T (m) and T (n) Respectively representing weights of CPU, memory and network, T (c) 、T (m) And T (n) The sum of (2) is 1.
Preferably, in S3 and S4, the calculation formula of the value of the server real-time load is:
wherein L is (i) =[L (ci) ,L (mi) ,L (ni) ],T (ci) 、T (mi) And T (ni) Respectively representing the utilization rates of a CPU, a memory and a network of the server;
the calculation formula of the real-time load rate of the server is as follows:
preferably, in S5, the calculation formula of the server calculating the weight is:
wherein,,n represents the number of servers, since R is the same and the division calculation speed is slow, i.e., equation (8) can be optimized to C (i) =R-R (i) ;
When C of two servers (i) When the values are the same, only replaceThe processing capacity of the server is not representative of the processing capacity of the server in all servers, and the weight W can be set for the server (i) The calculation formula of the server truly assigned weight is as follows:
CW (i) =C (i) *W (i) (9)。
preferably, the sum of the performance of the CPUs in the cluster is
The sum of the performances of the memories in the cluster is
The sum of the performance of disk IO in the cluster is
The sum of the performance of network bandwidths in a cluster is
Because the degree of dependence of each factor on the server performance is different, the dependence ratios of CPU, memory, disk IO and network bandwidth are assumed to be K respectively c 、K m 、K d 、K b And K is c +K m +K d +K b =1。
Preferably, the actual specific gravity of the performance of each server node in the cluster is:
wherein W is p (Ni) represents the true specific gravity of each node performance in the service cluster.
Compared with the prior art, the invention has the beneficial effects that:
the invention adopts a weighted polling algorithm to process data, simultaneously reads network request data in a near-time period in an asynchronous mode, analyzes the quantity and duration of the processed data and abnormal conditions, and gives out an optimal weight proportion; the weight of the optimized algorithm is obtained according to the load condition of the server, the floating condition of the server can be reflected, and the server resources are utilized effectively.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, the present invention provides a technical solution: a gateway load balancing algorithm optimization method comprises the following steps:
s1, collecting a value of a load factor;
s2, comparing the value of each load factor with a corresponding threshold value, if L exists (ci) >Y (ci) 、L (mi) >Y (mi) Or L (ni) >Y (ni) The weight of the server is set to 0, the server is not allocated with tasks in the period, otherwise, the weight vector T according to the load factor (i) Computing server real-time load M (i) Is a value of (2);
s3, calculating the processing capacity extreme value N of the server (i) And server real-time load rate R (i) Is a value of (2);
s4, each server R (i) The value of (2) and the threshold value Y corresponding to the server (i) Comparing if R (i) >Y (i) Setting the weight of the server to 0, otherwise, carrying out the next step;
s5, calculating the weight C by the calculation server (i) And according to the weight W of the server (i) Computing server trueReal allocation weight CW (i) ;
S6, calculating the real distribution weight CW between the servers (i) And (3) obtaining the weight of the server, and polling and distributing tasks according to the weight.
The real-time load of the server is dynamic two-dimensional data, which is obtained by multiplying the utilization rate of the load factors in the server by the corresponding weight. In view of this, the improved algorithm concept is as follows: the values of the load factors of the servers (which are dynamically changed) are collected once within a certain period of time (e.g. t=10s), and according to the collected values of the load factors, the values of the load factors are compared with the corresponding set thresholds, and when the collected values of the load factors are smaller than the corresponding set thresholds, the processing capacity extremum of each server and the real-time load value of each server are calculated, and a value is obtained by comparing the two values, which is called the load rate. If the load rate is smaller than the load rate threshold, calculating the weight of the current server, and carrying out polling processing request tasks according to the weight.
And calculating a weight ratio according to the calculated weight of the server, wherein the weight ratio represents the weight ratio of the server capable of processing tasks, and the ratio can well represent the performance of the current server. The ratio does not contain a server with the load rate greater than or equal to a threshold value (Y), and the weight of the server is determined according to the ratio, so that the normal operation of the server can be ensured.
In S2-S6, Y (ci) 、Y (mi) And Y (ni) Threshold values respectively representing CPU, memory and network of server, Y (i) Representing the threshold of the server, N (i) Representing server processing capability extremum, M (i) Representing real-time load of server, R (i) Representing real-time load rate of server, T (i) Weight vector representing each load factor, W (i) Representing the weight of the server.
In S1, the load factors include a CPU usage rate C (N i), a memory usage rate M (N i), a disk IO usage rate D (N i), and a network bandwidth usage rate B (N i); wherein,,
in the formulas (1) - (4), deltau represents the difference of user states between two sample points, deltas represents the difference of memory systems between the two sample points, and deltaj represents the difference of CUP utilization rates between the two sample points;
memotal represents memory, memfree represents free memory;
used represents Used disk space, block represents total disk space;
Δr represents the difference in reception bandwidth between the two samples, Δt represents the difference in transmission bandwidth between the two samples, Δt represents a period of time, total bw represents the port bandwidth, and the server uses the eth0 port, with a bandwidth of 100Mbps.
Because the performance of the Web server is influenced by the KPI data of the server node, the weight of the server can be dynamically modified according to the condition that the influencing factors are changed; the performance of the node is dynamically adjusted according to the CPU, the memory, the disk IO and the use condition of the network bandwidth of the node.
And because the Nginx is the number of the request connections distributed according to the weight ratio of each node in the cluster, the larger the weight is, the smaller the number of the request connections of the client is, so that the weight of the server can be dynamically modified to realize dynamic load balancing.
In S3, the calculation formula of the server processing capability extremum is:
N (i) =P (i) *T (i) (5)
wherein P is (i) =[P (ci) ,P (mi) ,P (ni) ],P (ci) 、P (mi) And P (ni) Respectively representing the processing speed, the memory size and the network throughput of the CPU of the server;
T (i) =[T (c) ,T (m) ,T (n) ],T (c) 、T (m) and T (n) Respectively representing weights of CPU, memory and network, T (c) 、T (m) And T (n) The sum of (2) is 1.
In S3 and S4, the calculation formula of the value of the server real-time load is:
M (i) =L (i) *T (i) (6)
wherein L is (i) =[L (ci) ,L (mi) ,L (ni) ],T (ci) 、T (mi) And T (ni) Respectively representing the utilization rates of a CPU, a memory and a network of the server;
the calculation formula of the real-time load rate of the server is as follows:
in S5, the calculation formula of the server calculating the weight is:
wherein,,representing the number of servers, since R is the same and the division calculation speed is slow, i.e., equation (8) can be optimized to C (i) =R-R (i) ;
When C of two servers (i) When the values are the same, the processing capacity of the server can be represented, the processing capacity of the server in all servers cannot be represented, and the weight W can be set for the servers (i) The calculation formula of the server truly assigned weight is as follows:
CW (i) =C (i) *W (i) (9)。
the sum of the performance of the CPUs in the cluster is
The sum of the performances of the memories in the cluster is
The sum of the performance of disk IO in the cluster is
The sum of the performance of network bandwidths in a cluster is
Because the degree of dependence of each factor on the server performance is different, the dependence ratios of CPU, memory, disk IO and network bandwidth are assumed to be K respectively c 、K m 、K d 、K b And K is c +K m +K d +K b =1。
The real specific gravity of the performance of each server node in the cluster is:
wherein W is p (Ni) represents the true specific gravity of each node performance in the service cluster.
Since the performance parameter indicators of each server in the service cluster are different, the performance specific gravity of each server node in the cluster is also different.
In the actual running process, the performance of each node in the cluster is continuously changed, and the weight is also dynamically changed, so that the performance ratio of each node in the cluster is also continuously changed; let W (N;) denote the weight of the i, i e {1, …, N } server nodes, then the weight ratio of each node in the cluster (i.e., the load balancing performance of the current node in the cluster) is:
wherein E is w (N i ) The weight ratio of the i-th node is represented, where i e {1, …, n }.
Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (9)
1. The gateway load balancing algorithm optimizing method is characterized by comprising the following steps:
s1, collecting a value of a load factor;
s2, comparing the value of each load factor with a corresponding threshold value, if L exists (ci) >Y (ci) 、L (mi) >Y (mi) Or L (ni) >Y (ni) The weight of the server is set to 0, the server is not allocated with tasks in the period, otherwise, the weight vector T according to the load factor (i) Computing server real-time load M (i) Is a value of (2);
s3, calculating the processing capacity extreme value N of the server (i) And server real-time load rate R (i) Is a value of (2);
s4, each server R (i) The value of (2) and the threshold value Y corresponding to the server (i) Comparing if R (i) >Y (i) Setting the weight of the server to 0, otherwise, carrying out the next step;
s5, calculating the weight C by the calculation server (i) And according to the weight W of the server (i) Calculating server true allocation weight CW (i) ;
S6, calculating the real distribution weight CW between the servers (i) And (3) obtaining the weight of the server, and polling and distributing tasks according to the weight.
2. The method for optimizing a gateway load balancing algorithm according to claim 1, wherein: in S2-S6, Y (ci) 、Y (mi) And Y (ni) Threshold values respectively representing CPU, memory and network of server, Y (i) Representing the threshold of the server, N (i) Representing server processing capability extremum, M (i) Representing real-time load of server, R (i) Representing real-time load rate of server, T (i) Weight vector representing each load factor, W (i) Representing the weight of the server.
3. The method for optimizing a gateway load balancing algorithm according to claim 1, wherein: in S1, the load factors include a CPU usage rate C (Ni), a memory usage rate M (Ni), a disk IO usage rate D (Ni), and a network bandwidth usage rate B (Ni); wherein,,
4. a gateway load balancing algorithm optimizing method according to claim 3, wherein: in the formulas (1) - (4), deltau represents the difference of user states between two sample points, deltas represents the difference of memory systems between the two sample points, and deltaj represents the difference of CUP utilization rates between the two sample points;
memotal represents memory, memfree represents free memory;
used represents Used disk space, block represents total disk space;
Δr represents the difference in reception bandwidth between the two samples, Δt represents the difference in transmission bandwidth between the two samples, Δt represents a period of time, total bw represents the port bandwidth, and the server uses the eth0 port, with a bandwidth of 100Mbps.
5. The method for optimizing a gateway load balancing algorithm according to claim 1, wherein: in S3, the calculation formula of the server processing capability extremum is:
N (i) =P (i) *T (i) (5)
wherein P is (i) =[P (ci) ,P (mi) ,P (ni) ],P (ci) 、P (mi) And P (ni) Respectively representing the processing speed, the memory size and the network throughput of the CPU of the server;
T (i) =[T (c) ,T (m) ,T (n) ],T (c) 、T (m) and T (n) Respectively representing weights of CPU, memory and network, T (c) 、T (m) And T (n) The sum of (2) is 1.
6. The method for optimizing a gateway load balancing algorithm according to claim 1, wherein: in S3 and S4, the calculation formula of the value of the server real-time load is:
M (i) =L (i) *T (i) (6)
wherein L is (i) =[L (ci) ,L (mi) ,L (ni) ],T (ci) 、T (mi) And T (ni) Respectively representing the utilization rates of a CPU, a memory and a network of the server;
the calculation formula of the real-time load rate of the server is as follows:
7. the method for optimizing a gateway load balancing algorithm according to claim 1, wherein: in S5, the calculation formula of the server calculating the weight is:
wherein,,n represents the number of servers, since R is the same and the division calculation speed is slow, i.e., equation (8) can be optimized to C (i) =R-R (i) ;
When C of two servers (i) When the values are the same, the processing capacity of the server can be represented, the processing capacity of the server in all servers cannot be represented, and the weight W can be set for the servers (i) The calculation formula of the server truly assigned weight is as follows:
CW (i) =C (i) *W (i) (9)。
8. a gateway load balancing algorithm optimizing method according to claim 3, wherein: the sum of the performances of the CPUs in the cluster is S c (
The sum of the performances of the memories in the cluster is
The sum of the performance of disk IO in the cluster is
The sum of the performance of network bandwidths in a cluster is
Because the degree of dependence of each factor on the server performance is different, the dependence ratios of CPU, memory, disk IO and network bandwidth are assumed to be K respectively c 、K m 、K d 、K b And K is c +K m +K d +K b =1。
9. The method for optimizing the gateway load balancing algorithm according to claim 8, wherein: the real specific gravity of the performance of each server node in the cluster is:
wherein W is p (Ni) represents the true specific gravity of each node performance in the service cluster.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310290101.6A CN116489096A (en) | 2023-03-23 | 2023-03-23 | Gateway load balancing algorithm optimization method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310290101.6A CN116489096A (en) | 2023-03-23 | 2023-03-23 | Gateway load balancing algorithm optimization method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116489096A true CN116489096A (en) | 2023-07-25 |
Family
ID=87216084
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310290101.6A Pending CN116489096A (en) | 2023-03-23 | 2023-03-23 | Gateway load balancing algorithm optimization method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116489096A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116680090A (en) * | 2023-08-03 | 2023-09-01 | 睿至科技集团有限公司 | Edge computing network management method and platform based on big data |
-
2023
- 2023-03-23 CN CN202310290101.6A patent/CN116489096A/en active Pending
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116680090A (en) * | 2023-08-03 | 2023-09-01 | 睿至科技集团有限公司 | Edge computing network management method and platform based on big data |
CN116680090B (en) * | 2023-08-03 | 2023-12-19 | 睿至科技集团有限公司 | Edge computing network management method and platform based on big data |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109308221B (en) | Nginx dynamic load balancing method based on WebSocket long connection | |
CN110717300B (en) | Edge calculation task allocation method for real-time online monitoring service of power internet of things | |
CN111726415B (en) | TCP long connection load balancing scheduling method and system based on negative feedback mechanism | |
CN109617826B (en) | Storm dynamic load balancing method based on cuckoo search | |
CN110233755B (en) | Computing resource and frequency spectrum resource allocation method for fog computing in Internet of things | |
CN115412497B (en) | Performance optimization method of BBR congestion control algorithm | |
CN116489096A (en) | Gateway load balancing algorithm optimization method | |
CN110445866A (en) | Task immigration and collaborative load-balancing method in a kind of mobile edge calculations environment | |
CN114465945B (en) | SDN-based identification analysis network construction method | |
Kang et al. | Application of adaptive load balancing algorithm based on minimum traffic in cloud computing architecture | |
Min et al. | Optimal super-peer selection for large-scale p2p system | |
CN115629865B (en) | Deep learning inference task scheduling method based on edge calculation | |
Zhu et al. | Improved dynamic load balancing algorithm based on Least-Connection Scheduling | |
CN115665227B (en) | Universal heterogeneous integrated computing network resource intelligent adaptation network architecture and method | |
CN117155942A (en) | Micro-service dynamic self-adaptive client load balancing method and system | |
CN112511652B (en) | Cooperative computing task allocation method under edge computing | |
CN113067863B (en) | Dynamic load balancing method based on software defined network | |
CN116302404B (en) | Resource decoupling data center-oriented server non-perception calculation scheduling method | |
CN116302578A (en) | QoS (quality of service) constraint stream application delay ensuring method and system | |
CN116166444A (en) | Collaborative reasoning method oriented to deep learning hierarchical model | |
CN110290556B (en) | Resource load balancing scheduling method based on optimal control variational method | |
CN109120480B (en) | Link optimization method and device based on data center | |
CN112148474B (en) | Loongson big data all-in-one self-adaptive task segmentation method and system for load balancing | |
CN106888237B (en) | Data scheduling method and system | |
CN114945024A (en) | Server load balancing optimization method based on long-term and short-term memory network |
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 |