CN114785793A - Elastic expansion load balancing method and system - Google Patents
Elastic expansion load balancing method and system Download PDFInfo
- Publication number
- CN114785793A CN114785793A CN202210253038.4A CN202210253038A CN114785793A CN 114785793 A CN114785793 A CN 114785793A CN 202210253038 A CN202210253038 A CN 202210253038A CN 114785793 A CN114785793 A CN 114785793A
- Authority
- CN
- China
- Prior art keywords
- load balancing
- end server
- virtual machine
- timing task
- utilization rate
- 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 31
- 238000012544 monitoring process Methods 0.000 claims abstract description 93
- 230000009467 reduction Effects 0.000 claims description 26
- 230000000694 effects Effects 0.000 claims description 12
- 230000008602 contraction Effects 0.000 abstract description 7
- 230000008569 process Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 238000011144 upstream manufacturing Methods 0.000 description 2
- 239000002023 wood Substances 0.000 description 2
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 235000012054 meals Nutrition 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
Images
Classifications
-
- 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
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
-
- 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
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
- H04L67/1004—Server selection for load balancing
- H04L67/1008—Server selection for load balancing based on parameters of servers, e.g. available memory or workload
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Computer And Data Communications (AREA)
Abstract
The invention discloses a method and a system for balancing elastic telescopic load, belongs to the technical field of load balancing, and aims to solve the technical problem of how to realize dynamic elastic expansion and contraction capacity of load balancing. The method comprises the following steps: configuring a load balance and a back-end server, and configuring a plurality of timing tasks; regularly pulling the resource utilization rate of the monitoring index of the rear-end server side through a first timing task of the load balancing side, if the resource utilization rate of the monitoring index of the rear-end server side is higher than a threshold value, creating and adding a virtual machine of the rear-end server through Openstack to achieve load balancing, and adding 1 to the elastic expansion level of the virtual machine of the rear-end server; and regularly pulling the resource utilization rate of the monitoring index of the load balancing side through a second timing task at the rear-end server side, and if the resource utilization rate of the monitoring index of the load balancing side is not enough to support concurrency, creating and adding a load balancing virtual machine to the load balancing cluster through Openstack.
Description
Technical Field
The invention relates to the technical field of load balancing, in particular to a method and a system for balancing elastic telescopic loads.
Background
The rapid development of computer communication network and other technologies and the diversity of user use scenes make the network overload and overload become common meals. In order to solve the problem of network load, a load balancing technology is required to be used for providing more flexible, convenient, rapid and efficient cloud computing service capability. Load balancing builds on existing network architectures and provides an inexpensive, efficient, and transparent way to extend the bandwidth of network devices and servers, increase throughput, enhance the data processing capabilities of the network, and increase the flexibility and availability of the network.
Definition of load balancing in Wiki encyclopedia: in computing, load balancing improves the distribution of workload among multiple computing resources (e.g., computers, clusters of computers, network links, central processing units, or disk drives). Load balancing aims at optimizing resource usage, maximizing throughput, minimizing response time, and avoiding overloading any individual resource. Using multiple components with load balancing rather than a single component may improve reliability and availability through redundancy. Load balancing typically involves dedicated software or hardware, such as a multi-tier switch or domain name system server process.
Load balancing is divided into hardware load balancing and software load balancing, and the hardware load balancing is realized by software load balancing in most service scenes due to high cost and poor flexibility. The software load balancing can be flexibly changed according to the actual requirements of users. The 4, 7-layer load balancing provided by the commonly used software load balancing LVS, NGINX also has strong effect.
Openstack is widely applied to the fields of public cloud and private cloud as an open-source cloud computing operating system, and multi-active load balancing based on Openstack is an urgent need in the industry at present. By utilizing Openstack, a server or a soft load balancing virtual machine can be quickly and conveniently created according to a service image.
The elastic expansion is to automatically adjust the management service of the elastic computing resources according to the business requirements and strategies to achieve the service capability of optimizing the resource combination. The computing capacity is increased when the traffic volume rises, and the computing capacity is reduced when the traffic volume falls, so that the stability and high availability of a business system are guaranteed, and meanwhile, the cost of computing resources is saved. For a cloud service provider, the elastic expansion meets the requirements of resource supply as required and dynamic management, and can reasonably allocate material resources and virtual resources, thereby avoiding service interruption caused by insufficient resource supply, and avoiding idle resource idling and system utilization reduction caused by excessive resource supply; for the user, the use cost of the user on computing resources and storage resources is reduced by elastic expansion, the user pays accurately and flexibly according to the used resources and time, and the maintenance cost is reduced and the service efficiency is improved by customized services such as dynamic migration of virtual services and allocation of virtual resources as required;
cloud service providers are various, and the implementation architecture of load balancing is also various. The difference of the architectures can affect the actual load balancing load effect, but the load balancing of a single instance has the maximum load capacity, which exceeds the maximum load, and the load balancing can even become the bottleneck affecting the concurrency of the service. The large number of request cards causes traffic disruption on the load balancing side. Therefore, for flexibility of services, load balancing must implement dynamic elastic capacity expansion and reduction to meet the requirements.
How to realize the dynamic elastic expansion and contraction of load balance is a technical problem to be solved.
Disclosure of Invention
The technical task of the invention is to provide a method and a system for balancing elastic telescopic load to solve the technical problem of how to realize dynamic elastic expansion and contraction of load balancing.
In a first aspect, the method for balancing elastic stretching load of the present invention comprises the following steps:
configuring a load balance and a back-end server, and configuring a plurality of timing tasks;
regularly pulling the resource utilization rate of the monitoring index of the rear-end server side through a first timing task of the load balancing side, if the resource utilization rate of the monitoring index of the rear-end server side is higher than a threshold value, creating and adding a virtual machine of the rear-end server through Openstack to achieve load balancing, and adding 1 to the elastic expansion level of the virtual machine of the rear-end server;
and pulling the resource utilization rate of the monitoring index of the load balancing side at regular time through a second timing task of the rear-end server side, and if the resource utilization rate of the monitoring index of the load balancing side is not enough to support concurrency, creating and adding a load balancing virtual machine to the load balancing cluster through Openstack.
Preferably, configuring the load balancing and backend server includes:
installing load balancing software on the virtual machine mirror image with balanced load, and configuring the load balancing;
configuring a software environment and a hardware environment required by creating a virtual machine through Openstack;
configuring a timing task and timing time of the timing task, wherein the timing task can access a back-end server and a load balancing side;
configuring a threshold value of a monitoring index, wherein the threshold value comprises an upper threshold value and a lower threshold value;
and creating a load balancing virtual machine and a back-end server virtual machine as standby through Openstack.
Preferably, when the load balancing and the back-end server are configured, all the configuration information is stored in the same configuration file or the same database.
Preferably, the monitoring indexes of the back-end server side comprise memory occupancy rate, CPU occupancy rate and culture descriptor utilization rate;
the load balancing side monitoring indexes comprise memory occupancy rates, CPU occupancy rates, culture descriptor utilization rates and the current connection number of load balancing.
Preferably, when the resource utilization rate of the monitoring index of the rear-end server side is pulled regularly through a first timing task at the load balancing side, the pulled resource utilization rate of the current monitoring index of the rear-end server side is stored in a database;
on the back-end server side, if the resource utilization rate of one or more monitoring indexes of the back-end servers with a preset proportion exceeds a threshold value, adding the back-end server virtual machine created through Openstack as a standby to load balancing, and creating a new back-end server virtual machine through Openstack as a standby.
Preferably, at the back-end server side, if one or more monitoring indexes of the back-end servers with a preset proportion exceed a threshold value, checking whether a back-end server virtual machine serving as a spare exists or not through a third timing task;
if yes, adding the spare rear-end server virtual machine into the configuration of load balancing, and adding 1 to the elastic expansion grade of the rear-end server virtual machine; if not, ending the first timing task;
if the rear-end servers with the preset proportion have one or more monitoring indexes with the resource utilization rate lower than the threshold value, whether the elastic expansion level of the virtual machine of the rear-end servers is 0 or not is checked through the first timing task, if not, one rear-end server is selected as the rear-end server to be deleted, the rear-end server is marked as the state to be deleted, and the weight of the rear-end server to be deleted in all load balancing is set to be 0.
Preferably, when the resource utilization rate of the monitoring index of the load balancing side is regularly pulled through a second timing task at the rear-end server side, the pulled resource utilization rate of the monitoring index of the current load balancing side is stored in a database;
on the load balancing side, judging whether the load balancing cluster needs capacity expansion or capacity contraction based on the barrel effect;
if capacity expansion is needed, adding the load balancing virtual machine created through Openstack as a standby to the load balancing cluster, and creating a new load balancing virtual machine through Openstack as a standby;
and if capacity reduction is required, setting the weight of all the load balancing virtual machines in the load balancing to be 0, removing the load balancing from the load balancing cluster and deleting the rear-end server virtual machine when the current connection number of the load balancing is lower than a certain threshold, and reducing the elastic expansion level of the rear-end server virtual machine by 1.
Preferably, if capacity expansion is needed, whether a spare load balancing virtual machine exists or not is checked through a third timing task;
if yes, adding the spare load balancing virtual machine into the load balancing configuration, and adding 1 to the elastic expansion grade of the load balancing virtual machine; if not, ending the second timing task;
if the capacity needs to be reduced, judging whether a back-end server virtual machine to be deleted exists in the load balancing through a fourth timing task, and if the back-end server virtual machine does not exist, ending the fourth timing task;
if the virtual machine of the back-end server to be deleted exists, judging whether the load balancing connection number is lower than a threshold value through a fourth timing task, and if not, ending the fourth timing task;
and if the load balancing connection number is lower than a threshold value, removing the load balancing from the load balancing cluster and deleting the virtual machine of the back-end server to be deleted.
In a second aspect, the system for elastically scalable load balancing of the present invention dynamically adjusts the number of backend server virtual machines and the number of load balancing virtual machines by using the method for elastically scalable load balancing according to any one of the first aspect, and the system includes:
the configuration module is used for configuring load balance and a back-end server and configuring a plurality of timing tasks;
the back-end server side elastic expansion module is used for regularly pulling the resource utilization rate of the back-end server side monitoring index through a first timing task at the load balancing side, creating and adding a back-end server virtual machine to the load balancing through Openstack if the resource utilization rate of the back-end server side monitoring index is higher than a threshold value, and adding 1 to the elastic expansion level of the back-end server virtual machine;
the system comprises a balanced load side elastic telescopic module, wherein the balanced load side elastic telescopic module is used for regularly pulling the resource utilization rate of a load balancing side monitoring index through a second timing task of a rear-end server side, and if the resource utilization rate of the load balancing side monitoring index is not enough to support concurrency, a load balancing virtual machine is created and added to a load balancing cluster through Openstack.
Preferably, the configuration module is configured to configure load balancing and a backend server, and includes:
installing load balancing software on the virtual machine mirror image with balanced load, and configuring the load balancing;
configuring a software environment and a hardware environment required by creating a virtual machine through Openstack;
configuring a timing task and timing time of the timing task, wherein the timing task can access a back-end server and a load balancing side;
configuring a threshold value of a monitoring index, wherein the threshold value comprises an upper limit threshold value and a lower limit threshold value;
creating a load balancing virtual machine and a back-end server virtual machine as standby through Openstack;
the back-end server side elastic expansion module is used for executing the following operations:
when the resource utilization rate of the monitoring index of the rear-end server side is pulled regularly through a first timing task of the load balancing side, the pulled resource utilization rate of the current monitoring index of the rear-end server side is stored in a database;
on the back-end server side, if the resource utilization rate of one or more monitoring indexes of the back-end servers with a preset proportion exceeds a threshold value, adding a back-end server virtual machine which is created through Openstack and used as a standby to load balance, and creating a new back-end server virtual machine through Openstack and used as a standby;
at the back-end server side, if the back-end server with a preset proportion has one or more resource utilization rate monitoring indexes exceeding a threshold value, checking whether a back-end server virtual machine serving as a spare exists or not through a third timing task;
if yes, adding the spare rear-end server virtual machine into the configuration of load balancing, and adding 1 to the elastic expansion grade of the rear-end server virtual machine; if not, ending the first timing task;
if the rear-end servers with the preset proportion have one or more monitoring indexes with the resource utilization rate lower than a threshold value, checking whether the elastic expansion level of the virtual machine of the rear-end servers is 0 or not through a first timing task, if not, selecting one rear-end server as the rear-end server to be deleted, marking the rear-end server as the state to be deleted, and setting the weight of the rear-end server to be deleted in all load balancing as 0;
the load balancing elastic expansion module is used for executing the following operations:
when the resource utilization rate of the monitoring index of the load balancing side is pulled regularly through a second timing task at the rear-end server side, the pulled resource utilization rate of the monitoring index of the current load balancing side is stored in a database;
on the load balancing side, judging whether the load balancing cluster needs capacity expansion or capacity reduction based on the barrel effect;
if capacity expansion is needed, adding the load balancing virtual machine created through Openstack as a standby to the load balancing cluster, and creating a new load balancing virtual machine through Openstack as a standby;
if capacity reduction is needed, setting the weight of all load balancing virtual machines in load balancing to be 0, removing the load balancing from a load balancing cluster and deleting the rear-end server virtual machines when the current connection number of the load balancing is lower than a certain threshold value, and reducing the elastic expansion level of the rear-end server virtual machines by 1;
if capacity expansion is needed, checking whether a spare load balancing virtual machine exists through a third timing task;
if the virtual machine exists, adding the spare load balancing virtual machine into the load balancing configuration, and adding 1 to the elastic expansion grade of the load balancing virtual machine; if not, ending the second timing task;
if the capacity needs to be reduced, judging whether a back-end server virtual machine to be deleted exists in the load balancing through a fourth timing task, and if the back-end server virtual machine does not exist, ending the fourth timing task;
if the virtual machine of the back-end server to be deleted exists, judging whether the load balancing connection number is lower than a threshold value through a fourth timing task, and if not, ending the fourth timing task;
and if the number of the load balancing connections is lower than a threshold value, removing the load balancing from the load balancing cluster and deleting the virtual machine of the back-end server to be deleted.
The method and the system for balancing the elastic telescopic load have the following advantages that:
1. the resource utilization rate conditions of the load balancing end and the application back-end server end are detected at regular time through the timing task, the data are reported to the storage library, if the resource utilization rate continuously exceeds or is less than a set threshold value, a reserved virtual machine is added to the load balancing end or the back end, the dynamic capacity expansion and capacity contraction of the load balancing are realized, the stability and high availability of a service system are guaranteed, and meanwhile, the calculation resource cost is saved;
2. the timing tasks are cooperatively performed, and decoupling enables the thread not to continuously wait for subsequent operations of adding a reserved virtual machine and removing a rear-end virtual machine/load balance.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed for the embodiments or the prior art descriptions will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on the drawings without creative efforts.
The invention is further described below with reference to the accompanying drawings.
Fig. 1 is a block diagram of a workflow of a first timing task and a second timing task in a method for balancing elastic stretching load according to embodiment 1;
fig. 2 is a block diagram of a workflow of a third timing task in the method for elastic stretching load balancing according to embodiment 1;
fig. 3 is a work flow diagram of a fourth timing task in the method for resilient scaling load balancing according to embodiment 1.
Detailed Description
The present invention is further described below with reference to the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and can implement the present invention, but the embodiments are not intended to limit the present invention, and the embodiments and technical features of the embodiments can be combined with each other without conflict.
The embodiment of the invention provides a method and a system for balancing elastic telescopic load, which are used for solving the technical problem of how to realize dynamic elastic expansion and contraction of load balancing.
Example 1:
the method for elastically stretching and contracting load balancing comprises the steps of pulling the resource utilization rate of a current front-end server and a current back-end server through a timing task at a load balancing side, creating and adding the back-end server to the load balancing by utilizing Openstack if the resource utilization rate of the back-end server is high, pulling the resource use condition at the load balancing side through one timing at the back-end server side, and creating a load balancing virtual machine and adding the load balancing virtual machine to a load balancing cluster when the load balancing is not enough to support the current concurrency.
In one particular implementation, the method includes the steps of:
s100, configuring a load balance and a back-end server, and configuring a plurality of timing tasks;
s200, pulling the resource utilization rate of the monitoring index of the rear-end server side at regular time through a first timing task of the load balancing side, if the resource utilization rate of the monitoring index of the rear-end server side is higher than a threshold value, creating and adding a rear-end server virtual machine to the load balancing through Openstack, and adding 1 to the elastic expansion level of the rear-end server virtual machine;
s300, the resource utilization rate of the monitoring index of the load balancing side is pulled regularly through a second timing task of the rear-end server side, and if the resource utilization rate of the monitoring index of the load balancing side is not enough to support concurrency, a load balancing virtual machine is created and added to the load balancing cluster through Openstack.
In the specific implementation process of step S100, configuring a load balancing and backend server includes:
(1) installing load balancing software on the virtual machine mirror image with balanced load, and configuring the load balancing;
(2) configuring a software environment and a hardware environment required by creating a virtual machine through Openstack;
(3) configuring a timing task and timing time of the timing task, wherein the timing task can access a rear-end server and a load balancing side;
(4) configuring a threshold value of a monitoring index, wherein the threshold value comprises an upper limit threshold value and a lower limit threshold value;
(5) and creating a load balancing virtual machine and a back-end server virtual machine as standby through Openstack.
In the configuration, the virtual machine image with balanced load needs to be installed with load balancing software first, and according to a service scene, the four-layer LVS or the 7-layer Nginx and the like are selected and configured correctly.
All configuration information of load balancing should be synchronously stored in a file or a database, and for the Nginx, Nginx. conf configuration files in a load balancing cluster are not very different, so that load balancing virtual machines created by synchronous capacity expansion can be very quickly sent to the cluster to support service concurrence.
In the embodiment, the monitoring indexes of the back-end server side comprise memory occupancy rate, CPU occupancy rate and culture descriptor utilization rate; the load balancing side monitoring indexes comprise memory occupancy rates, CPU occupancy rates, culture descriptor utilization rates and the current connection number of load balancing.
When the load balancing timing task pulls the resource utilization of the back-end service server, the load balancing timing task mainly pulls the memory occupancy rate, the cpu occupancy rate, the file descriptor utilization rate and the like. When one or more performance indexes of the back-end servers exceeding the proportion exceed the range in the back-end of the load balancing, the back-end servers to be backed up should be configured in, and the backup servers are newly created to be used as reserves. And when one or more performance indexes of the backend servers exceeding the ratio are lower than the range, the back-end servers subjected to capacity expansion are subjected to capacity reduction. And removing the virtual machine at the back end from the upstream during capacity reduction, or setting the weight to be 0, wherein due to the strong performance of nginx, the nginx does not forward a new request to the node any more, the old connection is hung, and the established connection cannot be interrupted until time is out or actively disconnected. The safe exit of the service and the elegant gray scale online and offline can be ensured.
When the application side timing task pulls the performance index of load balancing, the monitored index dimensionality mainly includes: the cpu utilization rate, the memory occupancy rate, the current connection number of load balancing, the file descriptor utilization rate and the like are judged by a special algorithm to judge whether the cluster needs capacity expansion, and according to the barrel effect, the factor influencing the load balancing effect is the shortest wood stick, namely the index x with the utilization rate closest to the bottleneck. And determining two thresholds m and n, when x is greater than or equal to m, judging that the cluster needs to be subjected to capacity expansion, and when x is less than or equal to n, judging that the cluster needs to be subjected to capacity reduction. When the conditions are met, the timing task enables the backup load balancing virtual machines to be correctly configured according to other load balancing machines in use and added into the cluster. When elastic capacity shrinkage is needed, the weight of the load balancing is set to be very small or even 0, and when the current connection number of the load balancing is lower than a certain threshold value, the load balancing is removed from the cluster and the load balancing virtual machine is deleted.
Because traffic may be bursty or system design is not reasonable, load balancing should perform resilient scaling only when the threshold is continuously exceeded or undershot. The value of each pull should be stored in the database. And simultaneously storing the standby machine, the load balancing virtual machine to be contracted and the back-end service virtual machine, and recording the grade of elastic expansion.
In the embodiment, a plurality of timing tasks are set, the first timing task is mainly to realize pulling of monitoring indexes of virtual machines of the back-end server and storing the monitoring indexes into a database, statistics is carried out on a plurality of virtual machines reaching the maximum threshold value at the moment of load balancing, the previous records are combined, and if the monitoring indexes are continuously greater than the maximum threshold value, the reserved back-end server is added into the load balancing. And if the rear-end elastic expansion grade is continuously less than +1, randomly selecting a rear-end server, storing the rear-end server into a table to be deleted, and setting the weight of the rear-end server in all load balancing to be 0. The second timing task is the same, but the monitoring data of the load balancing side should be pulled. And the third timing task checks whether a reserve virtual machine exists, and if the reserve virtual machine does not exist, the reserve virtual machine is created for standby. And storing the created virtual machine into a reserved virtual machine table. And the fourth timing task checks whether the virtual machine is to be deleted, if the virtual machine is to be deleted, the connection is checked, the virtual machine is deleted after the connection is lower than a certain threshold value, and the elastic expansion level is-1. The plurality of timing tasks do not interfere with each other, and what is influenced is to set a flag bit to influence the execution of other timing tasks. And a plurality of timing tasks do not need to be started simultaneously, and the execution speed of each time is high, so that the interval guarantees that a large number of threads do not exist simultaneously to influence the performance. It should be noted that the level (number of times) of capacity expansion needs to be recorded to prevent the capacity reduction from being lower than the initial configuration. When the expansion level is 0, no matter how few services are, the back-end server and the load balancing virtual machine should not be reduced.
Based on the above, in a specific implementation of step S200, when the resource utilization rate of the monitoring index at the back-end server side is regularly pulled through the first timing task at the load balancing side, the pulled resource utilization rate of the current monitoring index at the back-end server side is stored in the database; and on the back-end server side, if one or more monitoring indexes of the back-end servers with a preset proportion exceed a threshold value, adding the back-end server virtual machine created through Openstack as a standby to load balancing, and creating a new back-end server virtual machine through Openstack as a standby.
At the back-end server side, if the back-end server with a preset proportion has one or more resource utilization rate monitoring indexes exceeding a threshold value, checking whether a back-end server virtual machine serving as a spare exists or not through a third timing task; if yes, adding the spare rear-end server virtual machine into the configuration of load balancing, and adding 1 to the elastic expansion grade of the rear-end server virtual machine; if not, ending the first timing task; if the resource utilization rate of one or more monitoring indexes of the rear-end servers with the preset proportion is lower than a threshold value, checking whether the elastic expansion level of the virtual machine of the rear-end servers is 0 or not through a first timing task, if not, selecting one rear-end server as the rear-end server to be deleted, marking the rear-end server as the state to be deleted, and setting the weight of the rear-end server to be deleted in all load balancing to be 0.
In the specific implementation of step S300, when the resource utilization rate of the load balancing side monitoring index is regularly pulled by the second timing task at the back-end server side, the resource utilization rate of the current load balancing side monitoring index that is pulled is stored in the database; on the load balancing side, judging whether the load balancing cluster needs capacity expansion or capacity reduction based on the barrel effect; if capacity expansion is needed, adding the load balancing virtual machine created through Openstack as a standby to the load balancing cluster, and creating a new load balancing virtual machine through Openstack as a standby; and if capacity reduction is required, setting the weight of all the load balancing virtual machines in the load balancing to be 0, removing the load balancing from the load balancing cluster and deleting the rear-end server virtual machine when the current connection number of the load balancing is lower than a certain threshold, and reducing the elastic expansion level of the rear-end server virtual machine by 1.
If capacity expansion is needed, checking whether a spare load balancing virtual machine exists or not through a third timing task; if the virtual machine exists, adding the spare load balancing virtual machine into the load balancing configuration, and adding 1 to the elastic expansion grade of the load balancing virtual machine; if not, ending the second timing task; if the capacity needs to be reduced, judging whether a back-end server virtual machine to be deleted exists in the load balancing through a fourth timing task, and if the back-end server virtual machine does not exist, ending the fourth timing task; if the virtual machine of the back-end server to be deleted exists, judging whether the load balancing connection number is lower than a threshold value through a fourth timing task, and if not, ending the fourth timing task; and if the load balancing connection number is lower than a threshold value, removing the load balancing from the load balancing cluster and deleting the virtual machine of the back-end server to be deleted.
In the use scene of the Openstack, an ECS instance can be created according to a specified virtual private network vpc, an instance mirror image, a specification navigator and the like, and the timing task does not need to wait for the virtual machine to be successfully created until the virtual machine is created, but stores the id of the ECS after the virtual machine is created. And checking whether the instance state is available when the timing task is executed next time. If not, the timing task does not need to be processed. And adding the instance state into the reserved virtual machine after the task at a certain time finds that the instance state is success. Different manufacturers need to ensure that the newly created virtual machine can be normally used according to actual conditions of the manufacturers, regardless of volume starting or mirror image starting. This eliminates the need for threads to track the creation of virtual machines.
As application implementation of the method in this embodiment, it is assumed that this load balancing provides a load balancing function for the application 1, and a backend server is configured behind the load balancing to provide a service of the application 1, and elastic expansion and contraction according to actual service access conditions are automatically achieved through four timing tasks, which specifically includes:
(1) relevant resources and hardware which are necessary for the Openstack virtual machine creation are prepared, and a correctly usable virtual machine can be created by using the mirror Image, the specification viewer, the virtual private network vpc, the host aggregation and the like of the application 1 and the load balancing virtual machine, and can execute corresponding business or a load balancing function;
(2) the database required by the patent is configured, and some parameters are configured. The timing time of the timing task can be configured according to the actual situation of the user. The resource use condition needing to be pulled can also be configured, such as CPU use ratio, memory use ratio, concurrent connection number, handle use number and the like; upper and lower threshold values are set. The timing task should have access to the back-end server and the load balancing side to pull regularly.
(2) Starting a timing task, wherein the timing task is carried out according to the line of the rear-end servers, when the actual use conditions of a plurality of rear-end servers are continuously greater than a threshold value, whether the rear-end servers are reserved or not is checked, if yes, the rear-end servers are added into the configuration in the load balance to be added into an application scene, the elastic expansion grade +1 of the rear-end virtual machines is set, and the purpose of setting the elastic expansion grade is that the number of the rear-end virtual machines is not lower than the original number even if the access amount is small; if the virtual machine is not reserved, the virtual machine is still in the process of creating, and the timing task is finished; if the continuous lower than the threshold value is made, whether the elastic expansion level is 0 or not is checked, if not, the capacity reduction can be carried out, at the moment, the weight of the load balancing to a certain virtual machine is set to be 0, and new service cannot enter the back-end virtual machine. Then, the timing task can be ended by marking the status of the timing task as a to-be-deleted status;
(3) the other two timing tasks are used for detecting the creation of a reserved virtual machine and detecting the state of a virtual machine to be deleted; if the reserved virtual machine state is still in the process of creation, ending the task; if the virtual machine is successfully established, the virtual machine is configured to a certain extent, and if the mirror image is good enough, the virtual machine can be directly placed into the directly used virtual machine if the virtual machine can be competent for tasks without configuration; the virtual machine to be deleted needs to monitor whether the current connection number is smaller than the set size, so as to ensure that the virtual machine is closed and deleted under the condition of the user with the least influence. The virtual machines mentioned in this section include back-end service virtual machines and load balancing virtual machines.
According to the method, the deletion server is created by combining a plurality of timing tasks and Openstack to realize load balancing and dynamic elastic expansion according to the actual service access concurrency, so that the user cost is reduced and the service efficiency is improved.
Example 2:
the invention discloses an elastic telescopic load balancing system which comprises a configuration module, a rear-end server side elastic telescopic module and a balanced load side elastic telescopic module, wherein the configuration module is used for configuring load balancing and a rear-end server and configuring a plurality of timing tasks; the back-end server side elastic expansion module is used for regularly pulling the resource utilization rate of the back-end server side monitoring index through a first timing task at the load balancing side, if the resource utilization rate of the back-end server side monitoring index is higher than a threshold value, creating and adding a back-end server virtual machine to the load balancing through Openstack, and adding 1 to the elastic expansion level of the back-end server virtual machine; the load balancing side elastic expansion module is used for regularly pulling the resource utilization rate of the load balancing side monitoring index through a second timing task at the rear end server side, and if the resource utilization rate of the load balancing side monitoring index is not enough to support concurrency, a load balancing virtual machine is created and added to the load balancing cluster through Openstack.
Wherein, the configuration module is used for configuring load balance and a back-end server, and comprises:
(1) installing load balancing software on the virtual machine mirror image with balanced load, and configuring the load balancing;
(2) configuring a software environment and a hardware environment required by creating a virtual machine through Openstack;
(3) configuring a timing task and timing time of the timing task, wherein the timing task can access a rear-end server and a load balancing side;
(4) configuring a threshold value of a monitoring index, wherein the threshold value comprises an upper limit threshold value and a lower limit threshold value;
(5) and creating a load balancing virtual machine and a back-end server virtual machine as standby through Openstack.
In the configuration, the virtual machine image with balanced load needs to be installed with load balancing software first, and according to a service scene, the four-layer LVS or the 7-layer Nginx and the like are selected and configured correctly.
All configuration information of load balancing should be synchronously stored in a file or a database, and for the Nginx, Nginx. conf configuration files in a load balancing cluster are not very different, so that load balancing virtual machines created by synchronous capacity expansion can be very quickly sent to the cluster to support service concurrence.
In the embodiment, the monitoring indexes of the back-end server side comprise memory occupancy rate, CPU occupancy rate and culture descriptor utilization rate; the load balancing side monitoring indexes comprise memory occupancy rate, CPU occupancy rate, culture descriptor utilization rate and the current connection number of load balancing.
The back-end server side elastic expansion module of this embodiment is configured to perform the following operations:
(1) when the resource utilization rate of the monitoring index of the rear-end server side is pulled regularly through a first timing task of the load balancing side, the pulled resource utilization rate of the current monitoring index of the rear-end server side is stored in a database;
(2) on the back-end server side, if the resource utilization rate of one or more monitoring indexes of the back-end servers with a preset proportion exceeds a threshold value, adding a back-end server virtual machine which is created through Openstack and used as a standby to load balance, and creating a new back-end server virtual machine through Openstack and used as a standby;
(3) at the back-end server side, if the back-end server with a preset proportion has one or more resource utilization rate monitoring indexes exceeding a threshold value, checking whether a back-end server virtual machine serving as a spare exists or not through a third timing task;
(4) if yes, adding the spare rear-end server virtual machine into the configuration of load balancing, and adding 1 to the elastic expansion grade of the rear-end server virtual machine; if not, ending the first timing task;
(5) if the resource utilization rate of one or more monitoring indexes of the rear-end servers with the preset proportion is lower than a threshold value, checking whether the elastic expansion level of the virtual machine of the rear-end servers is 0 or not through a first timing task, if not, selecting one rear-end server as the rear-end server to be deleted, marking the rear-end server as the state to be deleted, and setting the weight of the rear-end server to be deleted in all load balancing to be 0.
The load balancing elastic expansion module of the embodiment is used for executing the following operations:
(1) when the resource utilization rate of the monitoring index of the load balancing side is pulled regularly through a second timing task of the rear-end server side, the pulled resource utilization rate of the monitoring index of the current load balancing side is stored in a database;
(2) on the load balancing side, judging whether the load balancing cluster needs capacity expansion or capacity reduction based on the barrel effect;
(3) if capacity expansion is needed, adding the load balancing virtual machine created through Openstack as a standby to the load balancing cluster, and creating a new load balancing virtual machine through Openstack as a standby;
(4) if capacity reduction is needed, setting the weight of all load balancing virtual machines in load balancing as 0, removing the load balancing from the load balancing cluster and deleting the rear-end server virtual machines when the current connection number of the load balancing is lower than a certain threshold value, and reducing the elastic expansion level of the rear-end server virtual machines by 1;
(5) if capacity expansion is needed, checking whether a spare load balancing virtual machine exists through a third timing task;
(6) if yes, adding the spare load balancing virtual machine into the load balancing configuration, and adding 1 to the elastic expansion grade of the load balancing virtual machine; if not, ending the second timing task;
(7) if the capacity reduction is needed, judging whether a back-end server virtual machine to be deleted exists in the load balancing through a fourth timing task, and if the back-end server virtual machine does not exist, ending the fourth timing task;
(8) if the virtual machine of the back-end server to be deleted exists, judging whether the load balancing connection number is lower than a threshold value through a fourth timing task, and if not, ending the fourth timing task;
(9) and if the number of the load balancing connections is lower than a threshold value, removing the load balancing from the load balancing cluster and deleting the virtual machine of the back-end server to be deleted.
In this embodiment, when the back-end service server elastic expansion module pulls the resource utilization rate of the monitoring index, the memory occupancy rate, the cpu occupancy rate, the file descriptor utilization rate, and the like are mainly pulled. When one or more performance indexes of the back-end servers exceeding the proportion exceed the range in the back-end of the load balancing, the back-end servers to be backed up should be configured in, and the backup servers are created to be used as the reserve again. And when one or more performance indexes of the backend servers exceeding the ratio are lower than the range, the back-end servers subjected to capacity expansion are subjected to capacity reduction. And removing the virtual machine at the back end from the upstream during capacity reduction, or setting the weight to be 0, wherein due to the strong performance of nginx, the nginx does not forward a new request to the node any more, the old connection is hung, and the established connection cannot be interrupted until time is out or actively disconnected. The safe exit of the service can be ensured, and the online and offline of the service can be guaranteed in an elegant gray scale.
The resource utilization rate of the load balancing elastic expansion module for pulling the monitoring index is that the monitored index dimensionality mainly comprises: the cpu utilization rate, the memory occupancy rate, the current connection number of load balancing, the file descriptor utilization rate and the like are judged by a special algorithm to judge whether the cluster needs capacity expansion, and according to the barrel effect, the factor influencing the load balancing effect is the shortest wood stick, namely the index x with the utilization rate closest to the bottleneck. And determining two thresholds m and n, when x is larger than or equal to m, judging that the cluster needs to be subjected to capacity expansion, and when x is smaller than or equal to n, judging that the cluster needs to be subjected to capacity reduction. When the conditions are met, the timing task can correctly configure the backup load balancing virtual machine according to other load balancing machines in use and add the backup load balancing virtual machine into the cluster. When elastic capacity reduction is needed, the weight of the load balancing is set to be very small or even 0, and when the current connection number of the load balancing is lower than a certain threshold value, the load balancing is removed from the cluster and the load balancing virtual machine is deleted.
Because traffic may be bursty or unreasonably designed for the system, load balancing should perform resilient scaling only when the threshold is continuously exceeded or undershot. The value of each pull should be stored in the database. And simultaneously storing the standby machine, the load balancing virtual machine to be contracted and the back-end service virtual machine, and recording the grade of elastic expansion.
In the embodiment, a plurality of timing tasks are set, the first timing task is mainly to realize pulling of monitoring indexes of virtual machines of the back-end server and storing the monitoring indexes into a database, statistics is carried out on a plurality of virtual machines reaching the maximum threshold value at the moment of load balancing, the previous records are combined, and if the monitoring indexes are continuously greater than the maximum threshold value, the reserved back-end server is added into the load balancing. And the rear-end elastic expansion grade is +1, if the rear-end elastic expansion grade is continuously smaller than the rear-end elastic expansion grade, a rear-end server is randomly selected and stored in the table to be deleted, and the weight of the rear-end server in all load balancing is set to be 0. The second timing task is the same, but the monitoring data of the load balancing side is pulled. And the third timing task checks whether a reserve virtual machine exists, and if the reserve virtual machine does not exist, the reserve virtual machine is created for standby. And storing the created virtual machine into a reserved virtual machine table. And the fourth timing task checks whether the virtual machine is to be deleted, if the virtual machine is to be deleted, the connection is checked, the virtual machine is deleted after the connection is lower than a certain threshold value, and the elastic expansion level is-1. A plurality of timing tasks are not interfered with each other, and what is influenced is that a flag bit is set to influence the execution of other timing tasks. And a plurality of timing tasks do not need to be started simultaneously, and the execution speed of each time is high, so that the interval guarantees that a large number of threads do not exist simultaneously to influence the performance. It should be noted that the level (number of times) of capacity expansion needs to be recorded to prevent the capacity reduction from being lower than the initial configuration. When the expansion level is 0, no matter how few services are, the back-end server and the load balancing virtual machine should not be reduced.
The system of the embodiment can execute the method disclosed in embodiment 1.
While the invention has been particularly shown and described with reference to the preferred embodiments and drawings, it will be understood by those skilled in the art that the present invention is not limited to the embodiments disclosed, but rather that various other embodiments may be devised in combination with the embodiments and examples set forth herein and within the scope of the invention.
Claims (10)
1. The method for balancing the elastic telescopic load is characterized by comprising the following steps of:
configuring a load balance and a back-end server, and configuring a plurality of timing tasks;
regularly pulling the resource utilization rate of the monitoring index of the rear-end server side through a first timing task of the load balancing side, if the resource utilization rate of the monitoring index of the rear-end server side is higher than a threshold value, creating and adding a virtual machine of the rear-end server through Openstack to achieve load balancing, and adding 1 to the elastic expansion level of the virtual machine of the rear-end server;
and regularly pulling the resource utilization rate of the monitoring index of the load balancing side through a second timing task at the rear-end server side, and if the resource utilization rate of the monitoring index of the load balancing side is not enough to support concurrency, creating and adding a load balancing virtual machine to the load balancing cluster through Openstack.
2. The method of claim 1, wherein configuring the load balancing and backend server comprises:
installing load balancing software on the virtual machine mirror image with balanced load, and configuring the load balancing;
configuring a software environment and a hardware environment required by creating a virtual machine through Openstack;
configuring a timing task and timing time of the timing task, wherein the timing task can access a rear-end server and a load balancing side;
configuring a threshold value of a monitoring index, wherein the threshold value comprises an upper threshold value and a lower threshold value;
and creating a load balancing virtual machine and a back-end server virtual machine as standby through Openstack.
3. The method according to claim 2, wherein all configuration information is stored in the same configuration file or the same database when configuring the load balancing and the backend server.
4. The method of elastic scaling load balancing according to claim 1, 2 or 3, characterized in that the back-end server side monitors metrics including memory occupancy, CPU occupancy, and cultural descriptor usage;
the load balancing side monitoring indexes comprise memory occupancy rate, CPU occupancy rate, culture descriptor utilization rate and the current connection number of load balancing.
5. The method according to any one of claims 1, 2 or 3, wherein when the resource utilization rate of the monitoring index at the back-end server side is pulled regularly by the first timing task at the load balancing side, the pulled resource utilization rate of the monitoring index at the current back-end server side is stored in the database;
on the back-end server side, if the resource utilization rate of one or more monitoring indexes of the back-end servers with a preset proportion exceeds a threshold value, adding the back-end server virtual machine created through Openstack as a standby to load balancing, and creating a new back-end server virtual machine through Openstack as a standby.
6. The method for balancing elastic stretching load according to claim 5, characterized in that at the back-end server side, if one or more monitoring indexes of the back-end servers with a predetermined proportion exceed a threshold value, checking whether a back-end server virtual machine as a backup exists through a third timing task;
if yes, adding the spare rear-end server virtual machine into the configuration of load balancing, and adding 1 to the elastic expansion grade of the rear-end server virtual machine; if not, ending the first timing task;
if the resource utilization rate of one or more monitoring indexes of the rear-end servers with the preset proportion is lower than a threshold value, checking whether the elastic expansion level of the virtual machine of the rear-end servers is 0 or not through a first timing task, if not, selecting one rear-end server as the rear-end server to be deleted, marking the rear-end server as the state to be deleted, and setting the weight of the rear-end server to be deleted in all load balancing to be 0.
7. The method for balancing elastic stretching load according to claim 6, characterized in that when the resource utilization rate of the monitoring index at the load balancing side is pulled regularly by the second timing task at the back-end server side, the pulled resource utilization rate of the monitoring index at the current load balancing side is stored in the database;
on the load balancing side, judging whether the load balancing cluster needs capacity expansion or capacity reduction based on the barrel effect;
if capacity expansion is needed, adding the load balancing virtual machine created through Openstack as a standby to the load balancing cluster, and creating a new load balancing virtual machine through Openstack as a standby;
and if capacity reduction is required, setting the weight of all the load balancing virtual machines in the load balancing to be 0, removing the load balancing from the load balancing cluster and deleting the rear-end server virtual machine when the current connection number of the load balancing is lower than a certain threshold, and reducing the elastic expansion level of the rear-end server virtual machine by 1.
8. The method for balancing elastic telescopic load according to claim 7, wherein if capacity expansion is required, checking whether a virtual machine for load balancing exists as a backup through a third timing task;
if the virtual machine exists, adding the spare load balancing virtual machine into the load balancing configuration, and adding 1 to the elastic expansion grade of the load balancing virtual machine; if not, ending the second timing task;
if the capacity reduction is needed, judging whether a back-end server virtual machine to be deleted exists in the load balancing through a fourth timing task, and if the back-end server virtual machine does not exist, ending the fourth timing task;
if the virtual machine of the back-end server to be deleted exists, judging whether the load balancing connection number is lower than a threshold value through a fourth timing task, and if not, ending the fourth timing task;
and if the number of the load balancing connections is lower than a threshold value, removing the load balancing from the load balancing cluster and deleting the virtual machine of the back-end server to be deleted.
9. A system for resilient scaling load balancing, characterized in that the number of backend server virtual machines and load balancing virtual machines is dynamically adjusted by a method for resilient scaling load balancing according to any of claims 1-8, the system comprising:
the configuration module is used for configuring load balancing and a back-end server and configuring a plurality of timing tasks;
the system comprises a rear-end server side elastic expansion module, a load balancing module and a back-end server side elastic expansion module, wherein the rear-end server side elastic expansion module is used for regularly pulling the resource utilization rate of a rear-end server side monitoring index through a first timing task at the load balancing side, creating and adding a rear-end server virtual machine to load balancing through Openstack if the resource utilization rate of the rear-end server side monitoring index is higher than a threshold value, and adding 1 to the elastic expansion level of the rear-end server virtual machine;
the load balancing side elastic telescopic module is used for regularly pulling the resource utilization rate of the load balancing side monitoring index through a second timing task of the rear-end server side, and if the resource utilization rate of the load balancing side monitoring index is not enough to support concurrency, a load balancing virtual machine is created and added to the load balancing cluster through Openstack.
10. The system according to claim 9, wherein the configuration module is configured for configuring load balancing and backend servers, and comprises:
installing load balancing software on the virtual machine mirror image with balanced load, and configuring the load balancing;
configuring a software environment and a hardware environment required by creating a virtual machine through Openstack;
configuring a timing task and timing time of the timing task, wherein the timing task can access a back-end server and a load balancing side;
configuring a threshold value of a monitoring index, wherein the threshold value comprises an upper threshold value and a lower threshold value;
establishing a load balancing virtual machine and a back-end server virtual machine as standby through Openstack;
the back-end server side elastic expansion module is used for executing the following operations:
when the resource utilization rate of the monitoring index of the rear-end server side is regularly pulled through a first timing task of the load balancing side, the pulled resource utilization rate of the monitoring index of the current rear-end server side is stored in a database;
on the back-end server side, if the resource utilization rate of one or more monitoring indexes of the back-end servers with a preset proportion exceeds a threshold value, adding a back-end server virtual machine which is created through Openstack and used as a standby to load balance, and creating a new back-end server virtual machine through Openstack and used as a standby;
at the back-end server side, if the back-end server with a preset proportion has one or more resource utilization rate monitoring indexes exceeding a threshold value, checking whether a back-end server virtual machine serving as a spare exists or not through a third timing task;
if yes, adding the spare rear-end server virtual machine into the configuration of load balancing, and adding 1 to the elastic expansion grade of the rear-end server virtual machine; if not, ending the first timing task;
if the rear-end servers with the preset proportion have one or more monitoring indexes with the resource utilization rate lower than a threshold value, checking whether the elastic expansion level of the virtual machine of the rear-end servers is 0 or not through a first timing task, if not, selecting one rear-end server as the rear-end server to be deleted, marking the rear-end server as the state to be deleted, and setting the weight of the rear-end server to be deleted in all load balancing as 0;
the load balancing elastic expansion module is used for executing the following operations:
when the resource utilization rate of the monitoring index of the load balancing side is pulled regularly through a second timing task of the rear-end server side, the pulled resource utilization rate of the monitoring index of the current load balancing side is stored in a database;
on the load balancing side, judging whether the load balancing cluster needs capacity expansion or capacity reduction based on the barrel effect;
if capacity expansion is needed, adding the load balancing virtual machine created through Openstack as a standby to the load balancing cluster, and creating a new load balancing virtual machine through Openstack as a standby;
if capacity reduction is needed, setting the weight of all load balancing virtual machines in load balancing to be 0, removing the load balancing from a load balancing cluster and deleting the rear-end server virtual machines when the current connection number of the load balancing is lower than a certain threshold value, and reducing the elastic expansion level of the rear-end server virtual machines by 1;
if capacity expansion is needed, checking whether a spare load balancing virtual machine exists through a third timing task;
if the virtual machine exists, adding the spare load balancing virtual machine into the load balancing configuration, and adding 1 to the elastic expansion grade of the load balancing virtual machine; if not, ending the second timing task;
if the capacity reduction is needed, judging whether a back-end server virtual machine to be deleted exists in the load balancing through a fourth timing task, and if the back-end server virtual machine does not exist, ending the fourth timing task;
if the virtual machine of the back-end server to be deleted exists, judging whether the load balancing connection number is lower than a threshold value through a fourth timing task, and if not, ending the fourth timing task;
and if the load balancing connection number is lower than a threshold value, removing the load balancing from the load balancing cluster and deleting the virtual machine of the back-end server to be deleted.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210253038.4A CN114785793A (en) | 2022-03-15 | 2022-03-15 | Elastic expansion load balancing method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210253038.4A CN114785793A (en) | 2022-03-15 | 2022-03-15 | Elastic expansion load balancing method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114785793A true CN114785793A (en) | 2022-07-22 |
Family
ID=82424860
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210253038.4A Pending CN114785793A (en) | 2022-03-15 | 2022-03-15 | Elastic expansion load balancing method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114785793A (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016058318A1 (en) * | 2014-10-17 | 2016-04-21 | 中兴通讯股份有限公司 | Elastic virtual machine (vm) resource scaling method, apparatus and system |
| CN109067867A (en) * | 2018-07-30 | 2018-12-21 | 北京航空航天大学 | The virtualization container service elastic telescopic method of data-oriented central loading monitoring |
| CN109086131A (en) * | 2018-06-25 | 2018-12-25 | 郑州云海信息技术有限公司 | A kind of virtual machine elastic telescopic control method and control device |
| CN110753112A (en) * | 2019-10-23 | 2020-02-04 | 北京百度网讯科技有限公司 | Elastic expansion method and device of cloud service |
| WO2020083051A1 (en) * | 2018-10-22 | 2020-04-30 | 中国移动通信有限公司研究院 | Network element load balancing method and network device |
| CN112000435A (en) * | 2020-08-21 | 2020-11-27 | 浪潮云信息技术股份公司 | Multi-activity load balancing method and system based on Openstack |
| CN112416506A (en) * | 2019-08-21 | 2021-02-26 | 中移(苏州)软件技术有限公司 | Container management method, device and computer storage medium |
-
2022
- 2022-03-15 CN CN202210253038.4A patent/CN114785793A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016058318A1 (en) * | 2014-10-17 | 2016-04-21 | 中兴通讯股份有限公司 | Elastic virtual machine (vm) resource scaling method, apparatus and system |
| CN109086131A (en) * | 2018-06-25 | 2018-12-25 | 郑州云海信息技术有限公司 | A kind of virtual machine elastic telescopic control method and control device |
| CN109067867A (en) * | 2018-07-30 | 2018-12-21 | 北京航空航天大学 | The virtualization container service elastic telescopic method of data-oriented central loading monitoring |
| WO2020083051A1 (en) * | 2018-10-22 | 2020-04-30 | 中国移动通信有限公司研究院 | Network element load balancing method and network device |
| CN112416506A (en) * | 2019-08-21 | 2021-02-26 | 中移(苏州)软件技术有限公司 | Container management method, device and computer storage medium |
| CN110753112A (en) * | 2019-10-23 | 2020-02-04 | 北京百度网讯科技有限公司 | Elastic expansion method and device of cloud service |
| CN112000435A (en) * | 2020-08-21 | 2020-11-27 | 浪潮云信息技术股份公司 | Multi-activity load balancing method and system based on Openstack |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112199194B (en) | Resource scheduling method, device, equipment and storage medium based on container cluster | |
| US10387202B2 (en) | Quality of service implementation in a networked storage system with hierarchical schedulers | |
| US10331491B2 (en) | Virtual data center resource mapping method and device | |
| US8639816B2 (en) | Distributed computing based on multiple nodes with determined capacity selectively joining resource groups having resource requirements | |
| US20190163371A1 (en) | Next generation storage controller in hybrid environments | |
| US8434078B2 (en) | Quick deployment method | |
| CN110515724B (en) | Resource allocation method, device, monitor and machine-readable storage medium | |
| US20040039815A1 (en) | Dynamic provisioning system for a network of computers | |
| US20080007765A1 (en) | Load distribution control system and method | |
| EP1654649B1 (en) | On demand node and server instance allocation and de-allocation | |
| WO2012048030A2 (en) | Automatic replication of virtual machines | |
| CN110221920B (en) | Deployment method, device, storage medium and system | |
| EP4029197B1 (en) | Utilizing network analytics for service provisioning | |
| CN111459642B (en) | Fault processing and task processing method and device in distributed system | |
| CN109039933B (en) | Cluster network optimization method, device, equipment and medium | |
| US8527565B2 (en) | Selecting and reassigning a blade for a logical partition for service scheduling of a blade server | |
| CN114675927A (en) | Service instance deployment method and device, electronic equipment and storage medium | |
| CN114785793A (en) | Elastic expansion load balancing method and system | |
| US9110823B2 (en) | Adaptive and prioritized replication scheduling in storage clusters | |
| CN113923222B (en) | Data processing method and device | |
| CN115934311A (en) | Data redistribution method and electronic equipment | |
| CN116635831A (en) | Data storage method and device and related equipment | |
| CN112398668B (en) | IaaS cluster-based cloud platform and node switching method | |
| CN106844487B (en) | A kind of distributed memory system replicated based on library | |
| CN117149099B (en) | Calculation and storage split server system and control method |
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 |