CN116016533A - Automatic weighting load balancing method and system, electronic equipment and storage medium - Google Patents
Automatic weighting load balancing method and system, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN116016533A CN116016533A CN202211655277.9A CN202211655277A CN116016533A CN 116016533 A CN116016533 A CN 116016533A CN 202211655277 A CN202211655277 A CN 202211655277A CN 116016533 A CN116016533 A CN 116016533A
- Authority
- CN
- China
- Prior art keywords
- cloud server
- weight
- server
- load balancing
- representing
- 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 61
- 238000003860 storage Methods 0.000 title claims abstract description 24
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 54
- 239000002071 nanotube Substances 0.000 claims abstract description 24
- 238000004590 computer program Methods 0.000 claims description 9
- 230000008569 process Effects 0.000 abstract description 17
- 230000004044 response Effects 0.000 abstract description 15
- 238000004364 calculation method Methods 0.000 description 12
- 230000002068 genetic effect Effects 0.000 description 12
- 230000006870 function Effects 0.000 description 11
- 238000012544 monitoring process Methods 0.000 description 11
- 238000012423 maintenance Methods 0.000 description 10
- 238000012545 processing Methods 0.000 description 10
- 230000003068 static effect Effects 0.000 description 10
- 238000011156 evaluation Methods 0.000 description 7
- 230000003287 optical effect Effects 0.000 description 5
- 238000005457 optimization Methods 0.000 description 5
- 210000000349 chromosome Anatomy 0.000 description 4
- 238000010606 normalization Methods 0.000 description 4
- 238000006243 chemical reaction Methods 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 239000013307 optical fiber Substances 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000003190 augmentative effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 239000011664 nicotinic acid Substances 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 239000000758 substrate Substances 0.000 description 1
Images
Landscapes
- Debugging And Monitoring (AREA)
Abstract
The invention provides an automatic weighting load balancing method and system, electronic equipment and storage medium, wherein the method comprises the following steps: acquiring various performance index parameters of a cloud server; according to various performance index parameters and a preset model algorithm, determining the weight of the cloud server; judging whether the weight of the cloud server needs to be adjusted according to the weight of the cloud server and a preset judging rule; if yes, taking the weight of the cloud server as an adjustment basis, and calling an interface adjustment weight for adjusting the weight of the load balancing back-end server through a server nano-tube system; therefore, all index parameters in the running process of the cloud server are quantitatively checked, the weight of the cloud server is calculated, a dynamic and automatic weight setting target is achieved, the overall average response time of a server cluster is reduced, the throughput is improved, and load balancing is further achieved.
Description
Technical Field
The invention belongs to the technical field of load balancing, and particularly relates to an automatic weighting load balancing method and system, electronic equipment and a storage medium.
Background
Under the cloud network environment, the most remarkable effect of cloud load balancing is to forward a request of a client to a real server of which the rear end is deployed in a cluster form through a specific balancing algorithm according to the set static server weight, solve the problem of single-point failure, and expand the service capacity of an application system to the outside. The load balancing algorithm of the current cloud network environment only comprises a weighted polling algorithm, a weighted minimum connection number and a source address hash scheduling algorithm, and the centers of gravity of the three algorithms are all used for distributing traffic according to a specific algorithm matched with weights after the static server weights are set. However, in the current daily operation and maintenance work, a default weight value such as 100 is generally set for all servers at the back end of the CLB (Cloud LoadBalancer, load balancing), and when the machine isolates the weight from faults by adjusting the weight to 0, or an empirical weight is set by relying on experience of manually evaluating the performance of the servers, and a quantization index is lacking.
In fact, even if the CLBs distribute traffic according to the specified weight and the load balancing algorithm, in the actual running process of the backend server, because of reasons such as the packet quantity, the packet size, the traffic, the request duration and the like of each request, the traffic cannot be completely and evenly distributed to each server over time, that is, the load condition of each server is high or low, in extreme cases, a certain server computing resource in the cluster may be exhausted, meanwhile, certain server computing resources are quite abundant, so that the overall average response time of the server cluster is higher, the throughput is reduced, and the server weight is static and cannot be dynamically adjusted along with the actual load condition.
Disclosure of Invention
In view of the above, the present invention aims to provide an automatic weighted load balancing method and system, an electronic device, and a storage medium, which are used for realizing a dynamic and automatic weight setting goal, reducing the overall average response time of a server cluster, improving the throughput goal, and further realizing load balancing.
The first aspect of the application discloses an automatic weighting load balancing method, which comprises the following steps:
acquiring various performance index parameters of a cloud server;
according to various performance index parameters and a preset model algorithm, determining the weight of the cloud server;
judging whether the weight of the cloud server needs to be adjusted according to the weight of the cloud server and a preset judging rule;
if yes, taking the weight of the cloud server as an adjustment basis, and calling the interface adjustment weight for adjusting the weight of the load balancing back-end server through a server nano-tube system.
Optionally, in the automatic weighted load balancing method, the determining the weight of the cloud server according to each performance index parameter and a preset model algorithm includes:
according to various performance index parameters, evaluating the node performance of the cloud server;
According to various performance index parameters, evaluating the comprehensive load of the cloud server;
and determining the weight of the cloud server according to the node performance and the comprehensive load of the cloud server.
Optionally, in the automatic weighted load balancing method, a formula used for evaluating the node performance of the cloud server according to each performance index parameter is:
C(S i )＝k 1 *n cpu *C cpu (S i )+k 2 *C mem (S i )+k 3 *C io (S i )+k 4 *C net (S i )+k 5 *C pps (S i )+k 6 *C con (S i )+k 7 *C dly (S i )
wherein ,
k i weight parameters representing the indexes of the ith cloud server; c (C) cpu (S i ) Representing the CPU frequency of the ith cloud server; c (C) mem (S i ) Representing the memory capacity of an ith cloud server; c (C) io (S i ) Representing the disk I/O rate of an ith cloud server; c (C) net (S i ) Representing the actual bandwidth of the ith cloud server; c (C) pps (S i ) Representing the actual package quantity of the ith cloud server, C con (S i ) Representing the actual connection number of the ith cloud server; c (C) dly (S i ) Representing the actual time delay; c (S) i ) And representing the node performance of the ith cloud server.
Optionally, in the automatic weighted load balancing method, estimating the comprehensive load of the cloud server according to each performance index parameter includes:
L(S i )＝σ 1 *L cpu (S i )+σ 2 *L mem (S i )+σ 3 *L io (S i )+σ 4 *L net (S i )+σ 5 *L pps (S i )+σ 6
*L con (S i )+σ 7 *L dly (S i )
wherein ,
σ i weight parameters representing all load indexes of the ith cloud server reflect the influence degree of all load indexes; l (L) cpu (S i ) Representing CPU utilization rate, L of ith cloud server mem (S i ) Representing memory utilization rate, L of ith cloud server io (S i ) Representing disk I/O occupancy rate, L of ith cloud server net (S i ) Representing network bandwidth occupancy rate, L, of ith cloud server pps (S i ) Representing packet occupancy, L, of an ith cloud server con (S i ) Representing connection number occupancy rate, L of ith cloud server dly (S i ) Representing the delay rate of the ith cloud server; l (S) i ) Representing the aggregate load of the ith cloud server.
Optionally, in the automatic weighted load balancing method, according to the node performance and the comprehensive load of the cloud server, the weight of the cloud server is determined, and an adopted formula is as follows:
wherein ,W(Si ) Representing the weight of the ith cloud server; n represents the total number of cloud servers; c (S) i Representing node performance of an ith cloud server; l (S) i ) Representing the aggregate load of the ith cloud server.
Optionally, in the automatic weighted load balancing method, determining whether the cloud server needs to adjust the weight according to the weight of the cloud server and a preset determination rule includes:
determining an index of cluster load balancing degree according to the comprehensive load of the cloud server and the average comprehensive load of the server clusters;
Judging whether the index of the cluster load balancing degree meets a threshold range or not;
if the index of the cluster load balancing degree meets a threshold range, the cloud server does not need to adjust the weight;
if the index of the cluster load balancing degree does not meet the threshold range, the cloud server needs to adjust the weight.
Optionally, in the automatic weighted load balancing method, a formula adopted for determining an index of the cluster load balancing degree is:
wherein F is an index of cluster load balancing degree; l (S) i ) The comprehensive load of the ith cloud server; l (L) arg An average comprehensive load for the server cluster; n is the total number of cloud servers in the server cluster.
The second aspect of the application discloses an automatic weighted load balancing system comprising:
the acquisition module is used for acquiring various performance index parameters of the cloud server;
the weight determining module is used for determining the weight of the cloud server according to various performance index parameters and a preset model algorithm;
the adjusting module is used for judging whether the cloud server needs to adjust the weight according to the weight of the cloud server and a preset judging rule; if yes, taking the weight of the cloud server as an adjustment basis, and calling the interface adjustment weight for adjusting the weight of the load balancing back-end server through a server nano-tube system.
A third aspect of the present application discloses an electronic device, comprising:
one or more processors;
a storage device having one or more programs stored thereon;
the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the automatic weighted load balancing method as described in any of the first aspects of the present application.
A fourth aspect of the present application discloses a storage medium having stored thereon a computer program, wherein the computer program, when executed by a processor, implements the automatic weighted load balancing method according to any of the first aspects of the present application.
As can be seen from the above technical solution, the automatic weighting load balancing method provided by the present invention includes: acquiring various performance index parameters of a cloud server; according to various performance index parameters and a preset model algorithm, determining the weight of the cloud server; judging whether the weight of the cloud server needs to be adjusted according to the weight of the cloud server and a preset judging rule; if yes, taking the weight of the cloud server as an adjustment basis, and calling an interface adjustment weight for adjusting the weight of the load balancing back-end server through a server nano-tube system; therefore, all index parameters in the running process of the cloud server are quantitatively checked, the weight of the cloud server is calculated, a dynamic and automatic weight setting target is achieved, the overall average response time of a server cluster is reduced, the throughput is improved, and load balancing is further achieved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of an automatic weighted load balancing method provided by an embodiment of the present invention;
FIG. 2 is a flow chart of another method of automatic weighted load balancing provided by an embodiment of the present invention;
FIG. 3 is a flow chart of another method of automatic weighted load balancing provided by an embodiment of the present invention;
FIG. 4 is a schematic diagram of an automatic weighted load balancing system according to an embodiment of the present invention;
fig. 5 is a schematic diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. 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.
In this application, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
Term interpretation:
(1) Cloud load balancing (Cloud LoadBalancer, abbreviated as CLB) virtualizes a plurality of cloud servers (abbreviated as CVM) located in the same private network (abbreviated as VPC) into a high-performance and high-availability application service pool by setting a virtual service address (abbreviated as VIP); and distributing the network request from the client into the cloud server pool according to a specified balancing algorithm. The CLB can check the health state of the cloud servers in the cloud server pool, automatically isolate the instance of the abnormal state, and solve the single-point problem of the cloud servers, thereby improving the fault tolerance of the application program; and meanwhile, the user uniformly forwards the high-concurrency application request to a plurality of application servers at the back end through the CLB, so that the service capacity of the application system outside is expanded, and the smooth running of the service is realized.
(2) Server weight: different weights are set according to the performance of the back-end server, the higher the performance is, the larger the weight is set, and the smaller the weight is otherwise; when the back-end server fails or needs to be upgraded, the weight can be adjusted to 0, and the isolation function is achieved.
(3) A Weighted Round Robin algorithm (Weighted Round Robin Scheduling) sets different weights for each CVM, and is suitable for short link services, such as http services, to support session maintenance according to the weight poll allocation request. The weighted polling scheduling algorithm can solve the problem of different performances among servers, and the weighted polling scheduling algorithm uses corresponding weights to represent the processing performances of the servers and distributes requests to the servers according to the height of the weights and a polling mode.
(4) The Weighted minimum connection number (Weighted Least-Connection Scheduling) obtains the number of active connections of the cloud hosts, and besides the weight value is added with the weight according to the processed weight value distribution request, an important index of the number of active connections is introduced, through the Weighted processing of the number of active connections, the request distribution of each cloud host is ensured to be associated with the load of the cloud host, the situation that some hosts are in a nearly idle state is avoided, and other hosts bear higher pressure is avoided.
(5) The source address hash scheduling algorithm (IPhash) performs hash operation on a source address IP of a request, uses a hash key (HashKey) to find a corresponding server from a statically allocated hash table, and distributes the request to a certain matched server in combination with a set back-end server weight, so that the request of the same client IP is always distributed to a certain specific server, and session maintenance is realized in another mode.
It should be noted that, when the operation and maintenance personnel basically configures the load balancing rule in the actual operation and maintenance working process, a static weight is set for the back-end server according to experience, and in daily operation and maintenance, more attention is paid to whether the application process runs normally, and whether the resource indexes such as the CPU, the application memory, the I/O and the like, the flow, the packet quantity, the connection number and the time delay are in a reasonable range or not, but the weight is never adjusted according to the performance index, so that the load of the server is reduced, and the performance index value returns to the normal range. The more important purpose of the weighting operation is to isolate the abnormal server and the server to be upgraded from the cluster, or to join the normal server and the server that completes the upgrade to the cluster.
In the traditional network and cloud network, a static server weight adjustment scheme is adopted at present, the static weight adjustment scheme is based on the prior empirical evaluation of the server performance, and the corresponding relation between the weight and the server performance is roughly obtained through the empirical evaluation. After the weight-server performance corresponding relation is obtained, the corresponding weight is configured for the cloud server to be put into use, and then the actual load balancing work such as a weighted minimum connection algorithm, a weighted polling algorithm and the like is carried out by combining with other load balancing scheduling algorithms needing to use the server weight.
The most important technical defects of the static server weight adjustment scheme are that the authority setting is not accurate enough, the weight cannot be adjusted in a self-adaptive and dynamic mode, and the real load balancing is not realized by the servers which are clustered and deployed at the back end of the load balancing. For example, when the same application service is deployed for the servers with the same model, the same weight and the same equalization strategy, such as a weighted polling strategy, are set, and when some requests on a certain server have longer session time and larger packet quantity and flow of the requests in a period of time, the indexes such as CPU, memory, I/O, time delay and the like are increased to higher level, namely, the load is too high, which leads to the increase of response time and the decrease of throughput of the processing request of the server; meanwhile, the requests on other servers have shorter session duration, and the packet quantity and flow of the requests are smaller, so that indexes such as CPU, memory, I/O, time delay and the like are all at lower level, namely the load is lower. However, since each server has the same weight and distributes the requests according to the weighted polling algorithm, when a new request comes, the requests are not distributed too little because some servers are loaded too much, and the requests are distributed too much because some servers are loaded too little, and the polling is performed step by step according to the same weight, so that the load balancing is not really achieved. This results in a high overall average response time-slice for the server cluster, and a low throughput, which is perceived by the client as "this service response is too slow".
Based on the above, the application provides an automatic weighting load balancing method, which is used for solving the problems that in the prior art, the overall average response time of a server cluster is higher, the throughput is reduced, the server weight is static, and dynamic adjustment cannot be carried out along with the actual load condition.
Referring to fig. 1, the automatic weighted load balancing method includes:
s101, acquiring various performance index parameters of a cloud server.
The performance index parameter may be actual usage data of indexes such as CPU, memory, I/0, bandwidth, packet size, connection number, delay, etc.
The performance index parameters may further include other parameters, which are not described in detail herein, and may be determined according to actual situations, and are all within the protection scope of the present application.
Specifically, the actual use data of indexes such as CPU, memory, I/0, bandwidth, packet quantity, connection number, time delay and the like of the cloud server can be monitored in real time through the capability of the existing cloud monitoring alarm platform, and the actual use data is uploaded to the weight calculation process.
It should be noted that, when automatic weight adjustment occurs each time, the cloud monitoring alarm platform may also issue an alarm to remind the operation and maintenance personnel.
S102, determining the weight of the cloud server according to each performance index parameter and a preset model algorithm.
That is, according to actual usage data of the indexes such as the CPU, the memory, the I/0, the bandwidth, the packet quantity, the connection number, the time delay and the like of the cloud server, the weight of the cloud server is determined according to a preset model algorithm.
The model algorithm can be selected in various modes as long as the model algorithm can determine the weight of the cloud server, and the model algorithm is within the protection scope of the application.
And S103, judging whether the weight of the cloud server needs to be adjusted according to the weight of the cloud server and a preset judgment rule.
The weight adjustment of the cloud server is adjusted according to the load condition of the cloud server, however, frequent adjustment of the node weight of the whole server cluster tends to consume a large amount of computing resources. Therefore, the timing of selecting the weight adjustment is particularly important.
Therefore, the scheme determines whether to adjust the weight of the cloud server by defining a determination rule.
The judging rule can be the contents of weight adjustment range, load balancing index and the like, the specific contents of which are not described in detail here, and the judging rule can be determined according to actual conditions and are all within the protection range of the application.
If the cloud server needs to adjust the weight, step S104 is performed.
S104, taking the weight of the cloud server as an adjustment basis, and calling the interface adjustment weight for adjusting the weight of the load balancing back-end server through a server nano-tube system.
That is, the weight calculation and management component: and accurately calculating the reasonable weight of the server according to the set model algorithm from the performance index data grasped by the monitoring alarm platform, making a decision whether the weight needs to be adjusted according to the set judgment rule, outputting a weight value if the weight needs to be adjusted, and calling the interface adjustment weight for adjusting the weight of the server at the back end of load balancing through a server nano tube system.
Cloud server nanotube system: and the cloud server nano-tube system is used for carrying out unified nano-tube on the cloud servers, so that the capability of pushing executable commands to a plurality of cloud servers and collecting execution results on one platform is realized.
In the embodiment, various performance index parameters of a cloud server are acquired; according to various performance index parameters and a preset model algorithm, determining the weight of the cloud server; judging whether the weight of the cloud server needs to be adjusted according to the weight of the cloud server and a preset judging rule; if yes, taking the weight of the cloud server as an adjustment basis, and calling an interface adjustment weight for adjusting the weight of the load balancing back-end server through a server nano-tube system; therefore, all index parameters in the running process of the cloud server are quantitatively checked, the weight of the cloud server is calculated, a dynamic and automatic weight setting target is achieved, the overall average response time of a server cluster is reduced, the throughput is improved, and load balancing is further achieved.
In practical application, referring to fig. 2, step S102, according to each performance index parameter, determines the weight of the cloud server according to a preset model algorithm, including:
and S201, evaluating the node performance of the cloud server according to various performance index parameters.
Note that the server cluster is composed of n server nodes, denoted as s= (S) 1 ,S 2 ,…,S n ). Server node S i The performance of (C) is defined as C (S i ) The processing capacity of the node is represented, and the processing capacity is evaluated through indexes such as CPU frequency, memory capacity, disk I/O rate, bandwidth, packet quantity, connection number, time delay and the like.
In practical application, according to various performance index parameters, a formula adopted for evaluating the node performance of the cloud server is as follows:
C(S i )＝k 1 *n cpu *C cpu (S i )+k 2 *C mem (S i )+k 3 *C io (S i )+k 4 *C net (S i )+k 5 *C pps (S i )+k 6
*C con (S i )+k 7 *C dly (S i )
wherein ,
k i weight parameters representing the indexes of the ith cloud server; c (C) cpu (S i ) Representing the CPU frequency of the ith cloud server; c (C) mem (S i ) Representing the memory capacity of an ith cloud server; c (C) io (S i ) Representing the disk I/O rate of an ith cloud server; c (C) net (S i ) Representing the actual bandwidth of the ith cloud server; c (C) pps (S i ) Representing the actual package quantity of the ith cloud server, C con (S i ) Representing the actual connection number of the ith cloud server; c (C) dly (S i ) Representing the actual time delay; c (S) i ) And representing the node performance of the ith cloud server.
Since the indexes such as CPU frequency, memory capacity, disk I/O rate, bandwidth, packet quantity, connection number, and time delay of the server node can be regarded as variables, the weight parameters thereof can be pre-allocated according to the actual situation of the server node.
The determining process of the weight parameters of the indexes comprises the following steps:
(1) And (5) parameterizing the performance of the cloud server.
According to experience in actual production, the performance parameter vector of the cloud server is evaluated by using 7 dimensions of indexes such as CPU utilization rate, memory capacity, disk I/O rate, bandwidth, packet quantity, connection number and time delay of the cloud server.
The server cluster consists of n server nodes, denoted s= (S) 1 ,S 2 ,…,S n ). The performance of the server node s_i is defined as C (S i ) The processing capacity of the node is represented, and the processing capacity is evaluated through indexes such as CPU frequency, memory capacity, disk I/O rate, bandwidth, packet quantity, connection number, time delay and the like.
C(S i )＝k 1 *n cpu *C cpu (S i )+k 2 *C mem (S i )+k 3 *C io (S i )+k 4 *C net (S i )+k 5
*C pps (S i )+k 6 *C con (S i )+k 7 *C dly (S i )
wherein ,
k i weight parameters representing the respective indices. n is n cpu Indicating the number of CPU cores, C cpu Indicating CPU utilization, C mem Representing the memory usage capacity, C io Representing disk I/O rate, C net Representing the actual bandwidth, C pps Representing the actual packet quantity, C con Representing the actual number of connections, C dly Representing the actual time delay.
And in actual calculation, normalizing the indexes according to the highest value in the server node. Normalization is a way of simplifying computation, converting different parameters with dimensions into parameters without dimensions through variation. Through normalization processing, parameters of different dimensions can be reasonably calculated in one expression.
The normalization formula is as follows:
where x is the value before conversion and y is the value after conversion. X is x min Is the minimum value of x, x max Is the maximum value of x.
Because the units of the CPU, the memory, the I/O, the bandwidth, the packet quantity, the connection number and the actual time delay are different, the unit of the indexes is processed by a normalization formula, and then the C (Si) is calculated by an evaluation formula, namely the comprehensive index C (Si) calculated by the normalized data is reasonable.
(2) And setting server weights.
Calculating to obtain the performance value of the serverC(S i ) The weights of the server nodes are then set as follows:
W(S i )＝C(S i )*M
wherein, M is a weight setting threshold, and its value is set by the cloud platform developer, typically 100.
(3) And (5) determining performance weight parameters.
In the node performance formula of the present invention,
k i the determination of the weight parameters for each performance indicator is one of the important challenges of the present solution.
In order to determine the performance index weight parameters, genetic algorithm is adopted for parameter optimization. The genetic algorithm is an algorithm for performing self-adaptive global optimization probability search by simulating natural biological genetic rules. The genetic algorithm has stronger robustness to the problem to be solved, and is widely applied to the fields of function optimization, production scheduling, machine learning and the like.
The algorithm idea is as follows: the genetic algorithm is an intelligent bionic algorithm designed according to a Darwin biological evolution mechanism. In the actual execution process of the genetic algorithm, the corresponding chromosome is firstly decoded in the actual problem, and the genetic algorithm is not directly used for processing the problem solution, so that the encoded parameter set is also used for expanding. And randomly forming an initial population, performing unfolding evaluation on the fitness of each chromosome according to a fitness function, performing operations such as selection, crossover, mutation and the like through the calculated fitness value, ensuring that the chromosomes are continuously optimized, and finally obtaining an optimal solution of the problem.
The genetic algorithm comprises the following steps:
1. individual coding, the coding process is to code the real problem into corresponding chromosome.
2. Constructing a fitness function:
the choice of fitness function directly affects the speed of the genetic algorithm iteration and eventually whether the optimal solution can be found, the smaller the fitness, the easier the individual is kept. The scheme defines the fitness function as follows:
where RT represents system response time and TPS represents system throughput. The smaller the system response time, the greater the system throughput and the smaller the fitness function value.
3. An initial population is generated and the population size is set.
4. And calculating the fitness evaluation value of the individuals in the population according to the fitness function.
5. Selecting, crossing and mutating. Setting a selection function, namely setting a crossing rate and a variation rate.
6. Setting an algorithm termination condition: and if the change of the weighted average value of the continuous N generations of fitness functions is smaller than 1e-6, stopping the algorithm.
7. And outputting a parameter optimization result. The parameter optimization result is the optimal performance weight parameter K opt ＝[k 1 ,k 2 ,k 3 ,k 4 ,k 5 ,k 6 ,k 7 ]。
That is, server node performance assessment uses the static weight adjustment scheme designed by the present scheme to determine performance index parameters. I.e. obtaining the optimal K by genetic algorithm opt ＝[k 1 ,k 2 ,k 3 ,k 4 ,k 5 ,k 6 ,k 7 ]。
S202, evaluating the comprehensive load of the cloud server according to various performance index parameters.
Note that, the cloud server S i Is defined as L (S) i ) Load indexes such as CPU utilization rate, memory utilization rate, disk I/O occupancy rate, network bandwidth occupancy rate, packet volume occupancy rate, connection number occupancy rate, delay rate and the like are selected as evaluation parameters.
In practical application, according to each performance index parameter, the comprehensive load of the cloud server is estimated, including:
L(S i )＝σ 1 *L cpu (S i )+σ 2 *L mem (S i )+σ 3 *L io (S i )+σ 4 *L net (S i )+σ 5 *L pps (S i )+σ 6 *L con (S i )+σ 7 *L dly (S i )
wherein ,
σ i weight parameters representing all load indexes of the ith cloud server reflect the influence degree of all load indexes; l (L) cpu (S i ) Representing CPU utilization rate, L of ith cloud server mem (S i ) Representing memory utilization rate, L of ith cloud server io (S i ) Representing disk I/O occupancy rate, L of ith cloud server net (S i ) Representing network bandwidth occupancy rate, L, of ith cloud server pps (S i ) Representing packet occupancy, L, of an ith cloud server con (S i ) Representing connection number occupancy rate, L of ith cloud server dly (S i ) Representing the delay rate of the ith cloud server; l (S) i ) Representing the aggregate load of the ith cloud server.
It should be noted that the number of the substrates,
σ i the calculation method of (2) is as follows:
wherein index= { cpu, mem, io, net, pps, con, dly }; l'. index (S i ) Is the reciprocal of the corresponding index; l (L) full-index Is the maximum value of the corresponding index; l (L) index (S i ) The utilization rate sigma of the corresponding index index (S i ) The weight value is the weight value of the corresponding index; l'. cpu (S i ) L 'for CPU utilization' index (S i ) The conversion value of the formula is also a calculation mode, and no actual name exists。
S203, determining the weight of the cloud server according to the node performance and the comprehensive load of the cloud server.
In practical application, according to node performance and comprehensive load of a cloud server, weight of the cloud server is determined, and an adopted formula is as follows:
wherein ,W(Si ) Representing the weight of the ith cloud server; n represents the total number of cloud servers; c (S) i Representing node performance of an ith cloud server; l (S) i ) Representing the aggregate load of the ith cloud server.
In practical application, referring to fig. 3, step S103, determining whether the cloud server needs to adjust the weight according to the weight of the cloud server and a preset determination rule includes:
s301, determining an index of cluster load balancing degree according to the comprehensive load of the cloud server and the average comprehensive load of the server clusters.
The weight adjustment of the server nodes is adjusted according to the load condition of the server nodes, however, frequent adjustment of the node weights of the whole server cluster tends to consume a large amount of computing resources. Therefore, the timing of selecting the weight adjustment is particularly important.
Therefore, the scheme quantitatively evaluates the load balance degree of the whole server cluster by defining a load balance index, and determines the time for adjusting the weight by setting a plurality of threshold ranges, namely, when the load balance index is in a certain threshold range, the weight is kept unchanged; when the load balancing index is beyond (below or above) the threshold range, the server weight is adjusted.
The load balance index quantifies and evaluates the load balance degree of the whole server cluster, and the time for adjusting the weight is determined by setting a plurality of threshold ranges, namely, when the load balance index is in a certain threshold range, the weight is kept unchanged; when the load balancing index is beyond (below or above) the threshold range, the server weight is adjusted.
S302, judging whether an index of the cluster load balancing degree meets a threshold range.
If the index of the cluster load balancing degree meets the threshold range, the cloud server does not need to adjust the weight.
If the index of the cluster load balancing degree does not meet the threshold range, the cloud server needs to adjust the weight.
In practical application, the formula adopted for determining the index of the cluster load balancing degree is as follows:
wherein F is an index of cluster load balancing degree; l (S) i ) The comprehensive load of the ith cloud server; l (L) arg An average comprehensive load for the server cluster; n is the total number of cloud servers in the server cluster.
The weight calculation and management component decides whether to adjust the weight of the cloud server through the evaluation result of the load balancing degree of the server cluster, sends the decision and the execution command to the cloud server nano-tube system, and the cloud server nano-tube system performs and manages the work of adjusting the weight of the cloud server in a unified way by dropping the interface.
In the embodiment, on the basis of realizing dynamic and automatic setting of the server weight, the selection strategy of the server is dynamically optimized by combining with the existing load balancing weighted scheduling algorithm, and the real load balancing is realized as much as possible, so that the overall average response time of the server cluster is reduced, the throughput is improved, and the overall service performance of the application is improved; that is, not only the actual performance index of the single cloud server is used as the basis for adjusting the weight, but also the load balancing degree of the whole cloud server cluster is used as the final decision basis.
It should be noted that, compared with the static server weight adjustment scheme, the invention has the advantages that the weight of the cloud server can be dynamically adjusted according to the actual dynamic performance index in the running process of the server, so that the load degree of each cloud server in the cloud server cluster is similar as much as possible, and the actual load balancing is implemented.
For example, when the same application service is deployed for the servers with the same model, the same weight and the same equalization strategy, such as a weighted polling strategy, are set, and when some requests on a certain server have longer session time and larger packet quantity and flow of the requests, the indexes such as CPU, memory, I/O and the like are increased to higher level, namely, the load is too high, and if the weight scheme is statically adjusted, the response time of the server for processing the requests is increased and the throughput is reduced; meanwhile, the requests on other servers have shorter session duration, and the packet quantity and flow of the requests are smaller, so that the indexes of CPU, memory, I/O and the like are all at lower level, namely the load is lower. However, since each server has the same weight and distributes the requests according to the weighted polling algorithm, when a new request comes, the requests are not distributed too little because some servers are too high in load, too many requests are distributed because some servers are too low in load, and the polling is still performed step by step according to the same weight, so that the load balancing is not really achieved, and the overall average response time slice of the server cluster is high and the throughput is low.
If a dynamic weight adjustment scheme is adopted, under the same configuration and scene, when the load of some cloud servers in the cluster is too high, and the load of other cloud servers is too low, the load of the cloud servers with too high load is automatically and dynamically reduced, and the load of the cloud servers with too low load is increased, so that the load of the cloud servers is gradually reduced and the load of the cloud servers is improved along with the time, and the dynamic adjustment is repeatedly performed, so that the real load balance of all cloud servers in the cluster is finally realized, and the overall average response time, throughput and application overall service performance of the server cluster are finally reduced.
Another embodiment of the present application provides an automatic weighted load balancing system, see fig. 4, comprising:
the acquisition module 101 is configured to acquire various performance index parameters of the cloud server.
The weight determining module 102 is configured to determine the weight of the cloud server according to each performance index parameter and a preset model algorithm.
The adjusting module 103 is configured to determine whether the cloud server needs to adjust the weight according to the weight of the cloud server and a preset determining rule; if yes, taking the weight of the cloud server as an adjustment basis, and calling the interface adjustment weight for adjusting the weight of the load balancing back-end server through a server nano-tube system.
The specific working process and principle of each module are described in detail in the automatic weighting load balancing method provided in the above embodiment, which is not described in detail herein, and the details are required to be within the protection scope of the present application according to the actual situation.
In this embodiment, the acquisition module 101 acquires various performance index parameters of the cloud server; the weight determining module 102 determines the weight of the cloud server according to various performance index parameters and a preset model algorithm; the adjusting module 103 judges whether the weight of the cloud server needs to be adjusted according to the weight of the cloud server and a preset judging rule; if yes, taking the weight of the cloud server as an adjustment basis, and calling the interface adjustment weight for adjusting the weight of the load balancing back-end server through a server nano-tube system;
it should be noted that, in the scheme, the monitoring platform, the weight calculation and management component and the cloud server nanotube platform are combined, dynamic server resources and performance indexes are creatively used as factors for dynamically adjusting the weights of the servers, modeling is performed according to actual operation and maintenance experience, and meanwhile, the optimal performance index parameters are calculated by utilizing a genetic algorithm.
The self-adaptive dynamic weight adjustment scheme comprises the following components: the system comprises a cloud monitoring alarm platform, a cloud server nano-tube system and a weight calculation and management component (container).
All the modules are arranged in the weight calculating and managing component. That is, the weight calculation and management component realizes dynamic weight adjustment by calling the cloud monitoring alarm platform and the cloud server nano-tube system.
Cloud monitoring alarm platform: the method comprises the steps of using the capability of an existing cloud monitoring alarm platform, monitoring actual use data of indexes such as CPU, memory, I/0, bandwidth, packet quantity, connection number, time delay and the like of cloud service in real time, uploading the actual use data to a weight calculation process, and sending an alarm to remind operation and maintenance personnel when automatic weight adjustment occurs each time.
Weight calculation and management component: and accurately calculating the reasonable weight of the server according to the set model algorithm from the performance index data captured by the monitoring alarm platform, making a decision whether to adjust the weight according to the set judgment rule, outputting a weight value if the weight is adjusted, and calling the interface adjustment weight for adjusting the weight of the server at the back end of load balancing through a server nano-tube system.
Cloud server nanotube system: and the cloud server nano-tube system is used for carrying out unified nano-tube on the cloud servers, so that the capability of pushing executable commands to a plurality of cloud servers and collecting execution results on one platform is realized.
The method combines a monitoring platform, a weight calculation and management component and a cloud server nano-tube platform, creatively takes dynamic server resources and performance indexes as factors for dynamically adjusting the weight of the server, models according to actual operation and maintenance experience, and calculates optimal performance index parameters by utilizing a genetic algorithm
Another embodiment of the present application provides a storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the automatic weighted load balancing method according to any of the above embodiments.
In the context of this disclosure, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. The machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
It should be noted that the computer readable medium described in the present disclosure may be a computer readable signal medium or a computer readable storage medium, or any combination of the two. The computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples of the computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this disclosure, a computer-readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In the present disclosure, however, the computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, with the computer-readable program code embodied therein. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: electrical wires, fiber optic cables, RF (radio frequency), and the like, or any suitable combination of the foregoing.
The computer readable medium may be contained in the electronic device; or may exist alone without being incorporated into the electronic device.
Another embodiment of the present invention provides an electronic device, as shown in fig. 5, including:
one or more processors 201.
A storage device 202 having one or more programs stored thereon.
The one or more programs, when executed by the one or more processors 201, cause the one or more processors 201 to implement the automatic weighted load balancing method as in any of the embodiments described above.
In particular, according to embodiments of the present disclosure, the processes described above with reference to flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a non-transitory computer readable medium, the computer program comprising program code for performing the method shown in the flow chart.
Although the subject matter has been described in language specific to structural features and/or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are example forms of implementing the claims.
While several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of the disclosure. Certain features that are described in the context of separate embodiments can also be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment can also be implemented in multiple embodiments separately or in any suitable subcombination.
The foregoing description is only of the preferred embodiments of the present disclosure and description of the principles of the technology being employed. It will be appreciated by persons skilled in the art that the scope of the disclosure referred to in this disclosure is not limited to the specific combinations of features described above, but also covers other embodiments which may be formed by any combination of features described above or equivalents thereof without departing from the spirit of the disclosure. Such as those described above, are mutually substituted with the technical features having similar functions disclosed in the present disclosure (but not limited thereto).
The automatic weighting load balancing method and system, the electronic equipment and the storage medium provided by the invention can be used in the artificial intelligence field, the blockchain field, the distributed field, the cloud computing field, the big data field, the Internet of things field, the mobile interconnection field, the network security field, the chip field, the virtual reality field, the augmented reality field, the holographic technical field, the quantum computing field, the quantum communication field, the quantum measurement field, the digital twin field or the financial field. The foregoing is merely an example, and the application fields of the automatic weighted load balancing method and system, the electronic device and the storage medium provided by the present invention are not limited.
Features described in the embodiments in this specification may be replaced or combined, and identical and similar parts of the embodiments may be referred to each other, where each embodiment focuses on differences from other embodiments. In particular, for a system or system embodiment, since it is substantially similar to a method embodiment, the description is relatively simple, with reference to the description of the method embodiment being made in part. The systems and system embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
Those of skill would further appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative elements and steps are described above generally in terms of functionality in order to clearly illustrate the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. An automatic weighted load balancing method, comprising:
acquiring various performance index parameters of a cloud server;
according to various performance index parameters and a preset model algorithm, determining the weight of the cloud server;
judging whether the weight of the cloud server needs to be adjusted according to the weight of the cloud server and a preset judging rule;
if yes, taking the weight of the cloud server as an adjustment basis, and calling the interface adjustment weight for adjusting the weight of the load balancing back-end server through a server nano-tube system.
2. The automatic weighted load balancing method according to claim 1, wherein the determining the weight of the cloud server according to the performance index parameters and a preset model algorithm includes:
According to various performance index parameters, evaluating the node performance of the cloud server;
according to various performance index parameters, evaluating the comprehensive load of the cloud server;
and determining the weight of the cloud server according to the node performance and the comprehensive load of the cloud server.
3. The automatic weighted load balancing method according to claim 2, wherein the formula used for evaluating the node performance of the cloud server according to each performance index parameter is:
C(S i )＝k 1 *n cpu *C cpu (S i )+k 2 *C mem (S i )+k 3 *C io (S i )+k 4 *C net (S i )+k 5 *C pps (S i )+k 6 *C con (S i )+k 7 *C dly (S i )
wherein ,
ki represents weight parameters of each index of the ith cloud server; c (C) cpu (S i ) Representing the CPU frequency of the ith cloud server; c (C) mem (S i ) Representing the memory capacity of an ith cloud server; c (C) io (S i ) Representing the disk I/O rate of an ith cloud server; c (C) net (S i ) Representing the actual bandwidth of the ith cloud server; c (C) pps (S i ) Representing the actual package quantity of the ith cloud server, C con (S i ) Representing the actual connection number of the ith cloud server; c (C) dly (S i ) Representing the actual time delay; c (S) i ) And representing the node performance of the ith cloud server.
4. The method for automatic weighted load balancing according to claim 2, wherein the evaluating the integrated load of the cloud server according to each performance index parameter comprises:
L(S i )＝σ 1 *L cpu (S i )+σ 2 *L mem (S i )+σ 3 *L io (S i )+σ 4 *L net (S i )+σ 5 *L pps (S i )+σ 6 *L con (S i )+σ 7 *L dly (S i )
wherein ,
σ i weight parameters representing all load indexes of the ith cloud server reflect the influence degree of all load indexes; l (L) cpu (S i ) Representing CPU utilization rate, L of ith cloud server mem (S i ) Representing memory utilization rate, L of ith cloud server io (S i ) Representing disk I/O occupancy rate, L of ith cloud server net (S i ) Representing network bandwidth occupancy rate, L, of ith cloud server pps (S i ) Representing packet occupancy, L, of an ith cloud server con (S i ) Representing connection number occupancy rate, L of ith cloud server dly (S i ) Representing the delay rate of the ith cloud server; l (S) i ) Representing the aggregate load of the ith cloud server.
5. The automatic weighted load balancing method according to claim 2, wherein the formula adopted for determining the weight of the cloud server according to the node performance and the comprehensive load of the cloud server is:
wherein ,W(Si ) Representing the weight of the ith cloud server; n represents the total number of cloud servers; c (S) i ) Representing node performance of an ith cloud server; l (S) i ) Representing the aggregate load of the ith cloud server.
6. The automatic weighted load balancing method according to claim 1, wherein determining whether the cloud server needs to adjust the weight according to the weight of the cloud server and a preset determination rule comprises:
Determining an index of cluster load balancing degree according to the comprehensive load of the cloud server and the average comprehensive load of the server clusters;
judging whether the index of the cluster load balancing degree meets a threshold range or not;
if the index of the cluster load balancing degree meets a threshold range, the cloud server does not need to adjust the weight;
if the index of the cluster load balancing degree does not meet the threshold range, the cloud server needs to adjust the weight.
7. The method of automatic weighted load balancing according to claim 6, wherein the formula used to determine the index of the cluster load balancing level is:
wherein F is an index of cluster load balancing degree; l (S) i ) The comprehensive load of the ith cloud server; l (L) arg An average comprehensive load for the server cluster; n is the total number of cloud servers in the server cluster.
8. An automatic weighted load balancing system, comprising:
the acquisition module is used for acquiring various performance index parameters of the cloud server;
the weight determining module is used for determining the weight of the cloud server according to various performance index parameters and a preset model algorithm;
the adjusting module is used for judging whether the cloud server needs to adjust the weight according to the weight of the cloud server and a preset judging rule; if yes, taking the weight of the cloud server as an adjustment basis, and calling the interface adjustment weight for adjusting the weight of the load balancing back-end server through a server nano-tube system.
9. An electronic device, comprising:
one or more processors;
a storage device having one or more programs stored thereon;
the method of automatic weighted load balancing of any of claims 1-7, when the one or more programs are executed by the one or more processors, causing the one or more processors to implement the method.
10. A storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the automatic weighted load balancing method according to any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211655277.9A CN116016533A (en) | 2022-12-22 | 2022-12-22 | Automatic weighting load balancing method and system, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211655277.9A CN116016533A (en) | 2022-12-22 | 2022-12-22 | Automatic weighting load balancing method and system, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116016533A true CN116016533A (en) | 2023-04-25 |
Family
ID=86026077
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211655277.9A Pending CN116016533A (en) | 2022-12-22 | 2022-12-22 | Automatic weighting load balancing method and system, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116016533A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116991561A (en) * | 2023-09-27 | 2023-11-03 | 国网北京市电力公司 | Model conversion scheduling method, device, equipment and medium |
| CN117149099A (en) * | 2023-10-31 | 2023-12-01 | 江苏华鲲振宇智能科技有限责任公司 | Calculation and storage split server system and control method |
-
2022
- 2022-12-22 CN CN202211655277.9A patent/CN116016533A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116991561A (en) * | 2023-09-27 | 2023-11-03 | 国网北京市电力公司 | Model conversion scheduling method, device, equipment and medium |
| CN117149099A (en) * | 2023-10-31 | 2023-12-01 | 江苏华鲲振宇智能科技有限责任公司 | Calculation and storage split server system and control method |
| CN117149099B (en) * | 2023-10-31 | 2024-03-12 | 江苏华鲲振宇智能科技有限责任公司 | Calculation and storage split server system and control method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Hussein et al. | Efficient task offloading for IoT-based applications in fog computing using ant colony optimization | |
| CN107196869B (en) | The adaptive load balancing method, apparatus and system of Intrusion Detection based on host actual loading | |
| CN113950103B (en) | Multi-server complete computing unloading method and system under mobile edge environment | |
| CN116016533A (en) | Automatic weighting load balancing method and system, electronic equipment and storage medium | |
| CN109656702B (en) | Cross-data center network task scheduling method based on reinforcement learning | |
| CN108667878A (en) | Server load balancing method and device, storage medium, electronic equipment | |
| CN106657379A (en) | Implementation method and system for NGINX server load balancing | |
| Liu et al. | Task scheduling in fog enabled Internet of Things for smart cities | |
| TW201424305A (en) | CDN load balancing in the cloud | |
| CN111381971A (en) | Nginx-based dynamic weight load balancing method | |
| CN102035737A (en) | Adaptive load balancing method and device based on cognitive network | |
| CN112737823A (en) | Resource slice allocation method and device and computer equipment | |
| CN106161552A (en) | Load-balancing method and system under a kind of mass data environment | |
| CN113141317A (en) | Streaming media server load balancing method, system, computer equipment and terminal | |
| CN108200156A (en) | The dynamic load balancing method of distributed file system under a kind of cloud environment | |
| CN113011678A (en) | Virtual operation platform operation control method based on edge calculation | |
| CN106790381A (en) | Dynamic feedback of load equalization methods based on weighting Smallest connection | |
| CN115629865B (en) | Deep learning inference task scheduling method based on edge calculation | |
| CN110351376A (en) | A kind of edge calculations node selecting method based on negative feedback mechanism | |
| Wu et al. | HiTDL: High-throughput deep learning inference at the hybrid mobile edge | |
| CN115580882A (en) | Dynamic network slice resource allocation method and device, storage medium and electronic equipment | |
| CN116467082A (en) | Big data-based resource allocation method and system | |
| CN117155942A (en) | Micro-service dynamic self-adaptive client load balancing method and system | |
| Zhou et al. | Load balancing prediction method of cloud storage based on analytic hierarchy process and hybrid hierarchical genetic algorithm | |
| CN113543160A (en) | 5G slice resource allocation method and device, computing equipment and computer storage medium |
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 |