CN117311992A - Method for predicting and automatically dynamically balancing internal resources of cluster based on established resources - Google Patents
Method for predicting and automatically dynamically balancing internal resources of cluster based on established resources Download PDFInfo
- Publication number
- CN117311992A CN117311992A CN202311597960.6A CN202311597960A CN117311992A CN 117311992 A CN117311992 A CN 117311992A CN 202311597960 A CN202311597960 A CN 202311597960A CN 117311992 A CN117311992 A CN 117311992A
- Authority
- CN
- China
- Prior art keywords
- cluster
- service
- resources
- resource
- snapshot
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 230000008859 change Effects 0.000 claims abstract description 5
- 230000008569 process Effects 0.000 claims description 5
- 238000011084 recovery Methods 0.000 claims description 4
- 238000013499 data model Methods 0.000 claims description 3
- 238000012423 maintenance Methods 0.000 abstract description 4
- 238000013480 data collection Methods 0.000 abstract description 2
- 230000000694 effects Effects 0.000 description 4
- 238000007405 data analysis Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000007726 management method 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
- 230000000737 periodic effect Effects 0.000 description 1
- 238000013468 resource allocation Methods 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5061—Partitioning or combining of resources
- G06F9/5072—Grid computing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5083—Techniques for rebalancing the load in a distributed system
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a method for predicting and automatically dynamically balancing internal resources of a cluster based on established resources, which realizes the dynamic balance of the internal resources of the cluster through a prediction algorithm and automatic adjustment of the cluster, thereby meeting the requirements of business on service resources. When the service is lowered aiming at the service resource requirement, the corresponding service resource is released, and the initial state of the cluster is restored. The invention is based on data collection and cluster service resource prediction, on the premise that the cluster scale does not change, the invention can effectively cope with the requirement of the service on specific service resources, and on the premise that the service resources are released by the service, the cluster can recover itself. By adopting the scheme of the invention, the internal resources of the cluster can be self-balanced, and the cluster is more intelligent. In a specific scene, the cost and expense caused by the capacity expansion of the clusters due to short-term service expansion can be reduced, and the maintenance cost of the clusters can be greatly reduced.
Description
Technical Field
The invention belongs to the technical field of big data, relates to cluster resource adjustment technology, and particularly relates to a method for predicting and automatically dynamically balancing internal resources of a cluster based on established resources.
Background
In recent years, as the service is continuously expanded, the cluster size is also increased, but at present, once the cluster service is distributed and created, the cluster service is difficult to adjust according to the service requirement. Under the conditions of service expansion and resource shortage, the method ensures the smooth expansion, stability and reliability of cluster service resources, and is an important subject of large data platform departments.
Most of the current solutions are to expand the capacity in a special period. The traditional capacity expansion mode only increases physical nodes, thereby achieving capacity expansion of cluster service resources, and the scheme can solve the problem of shortage of cluster resources, but needs human intervention and cannot fully and efficiently utilize the cluster resources; in addition, if the demand of the service for the cluster service resources increases rapidly, the operation of expanding and accommodating the cluster is only needed before and after a certain period of time, which is complicated and causes the stability of the cluster to be reduced due to more human intervention. Meanwhile, when some service resources are insufficient, the whole cluster resources are only laterally expanded at present to solve the resource problem. However, the entire cluster is laterally expanded, and although the problem of partial service resource shortage can be solved, for the cluster, all the resources of the cluster cannot be fully utilized.
At present, the solution to the rapid increase of the cluster resources by the service is not a cluster resource capacity expansion mode. Cluster expansion can be broadly divided into two main categories: the first class represents traditional manufacturers, and physical hosts are used for cluster deployment, so that the capacity expansion scheme is to increase cluster physical host resources; the second type is cloud on the cluster, such as Amazon's cloud cluster service resource, when the capacity expansion is needed, the number of containers is directly adjusted by the cloud technology, so that the capacity expansion effect of the cluster resource is achieved. However, the nature of capacity expansion of the clusters is to increase the number of nodes of the clusters and enlarge the scale of the clusters (a container can be understood as a special physical host); of course, the cloud cluster can also independently upgrade the Memory and other resources of the container on the premise of unchanged cluster scale, thereby achieving the effect of expanding the resources.
However, the above scheme, whether expanding the number of nodes (including containers or physics) of the cluster or adjusting the resources of the nodes, is always to increase the overall resources of the cluster. Although the scheme can solve the problem of resource shortage caused by business expansion, the scheme has obvious defects, such as continuous expansion of cluster scale, and thus, the maintenance difficulty is exponentially increased; in addition, the capacity of the service resources is only expanded according to specific service requirements, and the utilization rate of the whole service resources is not optimal according to the whole cluster; further more human intervention is required, resulting in reduced cluster stability, etc.
Disclosure of Invention
In order to solve the problems, the invention discloses a method for predicting and automatically dynamically balancing the resources in a cluster based on established resources, and the method realizes the dynamic balancing of the resources in the cluster through a prediction algorithm and automatic adjustment of the cluster, thereby meeting the requirements of business on service resources. When the service is lowered aiming at the service resource requirement, the corresponding service resource is released, and the initial state of the cluster is restored.
In order to achieve the above purpose, the technical scheme of the invention is as follows:
the invention provides a method for predicting and automatically dynamically balancing internal resources of a cluster based on established resources, which comprises the following steps:
step 1, dividing service types of a distributed computing system, periodically acquiring resource amount data of each type of service, and modeling acquired data of each service by using a Cluster Oracle;
step 2, predicting the service resource quantity required by one or more distributed computing systems in a future period based on the data model obtained in the step 1 and the collected historical resource usage data of one or more distributed computing systems, and obtaining an optimal allocation result of cluster service resources by combining the current cluster scale total resource quantity;
step 3, using Cluster Snapshot Service service to store the snapshot of the current cluster state;
step 4, triggering automatic dynamic balance of the cluster through an automatic dynamic balance resource framework of the cluster according to the optimal distribution result of the cluster service resources calculated in the step 2 at a proper time point after the current cluster snapshot is stored;
and 5, when detecting that the utilization rate of one or more distributed computing system resources reaches the adjustment threshold, triggering a recovery mode of the cluster automatic dynamic balance resource framework according to the cluster snapshot recorded in the step 3 at a proper time point to recover the cluster service resource distribution state and the service state.
Further, in the step 2, the optimal allocation result is:
total cluster scale:
；
wherein the method comprises the steps ofThe total number of the hosts is m, and the total number of Memory resources is m; />For the adjusted host number, +.>The Memory resource is adjusted;
the various service resources change into:
；
for the number of hosts before adjustment, +.>Memory resources before adjustment; if->、/>If the number is positive, the service will expand, if +.>、/>Negative, the shrinkage is indicated.
Further, the snapshot saving in the step 3 includes: the clusters distribute snapshots and business task snapshots.
Further, the cluster distribution snapshot is used for storing the distribution, service configuration and service state of various service hosts of the current cluster; the service task snapshot is used for storing snapshots of all services running on various services, including running states and resource occupation conditions.
Further, in the step 4, the process of automatic dynamic balancing of the cluster includes:
(1) Suspending all running traffic for all services, stopping running one or more distributed computing systems and all services involved;
(2) According to the optimal distribution result of the clusters, readjusting the resource scale of the distributed computing system to expand or contract the capacity of one or more distributed computing system resources;
(3) And recovering all online services according to the service task snapshot.
Furthermore, in the automatic dynamic balance process of the cluster, the total resources of the cluster are unchanged, and the total number of resources of other services is adjusted.
Further, in the step 5, the process of recovering the cluster service resource distribution state and the service state includes:
(1) Suspending all running traffic for all services, stopping one or more distributed computing systems and all services involved;
(2) According to the cluster backup snapshot recorded in the step 3, readjusting the resource scale of the distributed computing system, and recovering the cluster to the cluster distribution state before dynamic balance;
(3) And recovering all online services according to the service task snapshot.
Further, after the cluster is restored to the cluster distribution state before dynamic balance, the total number of the clusters is unchanged, the total number of the service hosts of the one or more distributed computing systems is adjusted to the original number, and the total number of the resources of other services is restored to the initial value of the step 1.
The beneficial effects of the invention are as follows:
the invention is based on data collection and cluster service resource prediction, on the premise that the cluster scale does not change, the invention can effectively cope with the requirement of the service on specific service resources, and on the premise that the service resources are released by the service, the cluster can recover itself. By adopting the scheme of the invention, the internal resources of the cluster can be self-balanced, and the cluster is more intelligent. In a specific scene, the cost and expense caused by the capacity expansion of the clusters due to short-term service expansion can be reduced, and the maintenance cost of the clusters can be greatly reduced.
The invention can release human resources without human intervention; and cluster resources can be fully utilized, capacity expansion is not required under the unnecessary condition, and the requirement of business on cluster service resources can be met.
The method of the invention can solve the rapid increase demand of service expansion aiming at specific service resources on the premise of not changing the cluster scale and the total resources of the clusters, and simultaneously greatly reduce the cost caused by the cluster expansion due to the short-term expansion of the service and the difficulty of cluster operation and maintenance.
Detailed Description
The technical scheme provided by the present invention will be described in detail with reference to the following specific examples, and it should be understood that the following specific examples are only for illustrating the present invention and are not intended to limit the scope of the present invention.
Assuming that the current cluster size for realizing the scheme of the invention is c (representing the number of cluster nodes), and the total resource of the cluster Memory is m; including various services such as HDFS, hive, HBase, yarn, presto, spark, etc., respectively,Representing the Memory resource amount (the same as other resources of the cluster) used by various services,Indicating the number of nodes for each type of service. The following formula can thus be derived:
；
the invention provides a method for predicting and automatically dynamically balancing cluster internal resources based on established resources, which mainly comprises the following steps:
step 1, aiming at a HDFS, yarn, hive, HBase, presto distributed computing system and the like, firstly dividing the distributed computing system into a storage type, a computing type, a middleware type and the like according to different characteristics of service types, and carrying out targeted and periodic resource quantity acquisition on each type of service. For example, for storage type services, metrics such as disk capacity, system IO, etc. are of interest; for the computing type service, the indexes such as system CPU, memory usage and the like are focused. Collected data for each service was analyzed and modeled using Cluster Oracle (CO, related to framework capabilities). In the following, the computing type resource Yarn is taken as an example, and the resources of one or more systems can be predicted and dynamically balanced at the same time according to the need.
And 2, based on the data model obtained in the step 1 and the condition of historical usage calculation type resources Yarn, adopting different fitting and prediction algorithms (adopting data analysis and prediction model algorithms or integrating and integrating the data analysis and prediction algorithms in an integrated mode so as to achieve decoupling, and determining which fitting and prediction algorithm is selected by a user), and predicting the service resource quantity required by Yarn and the service resource quantity available for the cluster in three months in the future (the prediction time can be configured according to the requirement) by combining the historical usage data of the Yarn and the current cluster scale total resource quantity, so as to obtain the optimal allocation result of the cluster service resources.
The allocation adjustment result is shown in the following formula:
total cluster scale:
；
wherein the method comprises the steps ofThe total number of the hosts is m, and the total number of Memory resources is m;
the various service resources change into:
if it is、/>Positive numbers indicate that the service is to be expanded, and vice versa.
For the adjusted host number, +.>The Memory resource is adjusted;
assuming that according to the collection and analysis of the service resource amount used by the service, the computing resource demand of the service in the future is predicted to show an expansion trend, especially in terms of computing type resources Yarn, in the current c_i host, memory resource m_i of Yarn reaches or exceeds the limit of the host, and the increasing demand of the service cannot be met, and expansion is needed. When the demand for computing resources in the coming three months shows a decreasing trend, the capacity reduction can be performed.
More specifically, during capacity expansion, the invention preferably adjusts each service resource internally by automatically adjusting the cluster resource mechanism to release the service host resource which is relatively less used by the service; at necessary moment, the cluster service can be temporarily put off the shelf, and more cluster resources are released, so that the effect of dynamically balancing the resources in the cluster is achieved.
Step 3, using Cluster Snapshot Service (CSS) service to snapshot save the current cluster state. This step mainly includes two aspects, namely cluster distribution snapshots and business task snapshots. The cluster distribution snapshot is mainly used for storing the distribution, service configuration, service state and the like of various service hosts of the current cluster so as to facilitate subsequent resource allocation and management. The service task snapshot is used for storing snapshots of all services operated on various services, including information such as operation states, resource occupation conditions and the like.
And step 4, triggering automatic dynamic balance of the cluster after the current cluster snapshot is stored. By balancing the capacity through a Cluster Dynamic Balancing Service (CDBS) framework, at an appropriate point in time (determined by the user to minimize the impact on the traffic), CDBS balancing clusters are triggered according to the cluster optimal distribution results calculated in step 2.
At this time, it is necessary to suspend all running services of all services (step 3 has made a snapshot of the service tasks) and then stop running Yarn and all services involved. And then, according to the optimal cluster distribution result, the scale of the Yarn resources is re-enlarged and adjusted so as to achieve the Yarn resource capacity expansion effect. For example, the total number of clusters c is unchanged, and the Yarn resource is formed by the original、/>Dynamic expansion to->、/>The method comprises the steps of carrying out a first treatment on the surface of the Wherein the increase of Yarn->、/>The amount of reduction of other service resources, how the host allocates, is determined by the automatic dynamic balance resource framework. And finally, recovering all online services according to the service task snapshot so as to dynamically balance cluster resources and meet the rapid increase demand of specific services on the resources.
And 5, when the Cluster prediction model detects that the utilization rate of the computing type service Yarn service resources is reduced and reaches an adjustment threshold, the Cluster Oracle (CO) triggers a recovery mode of a CDBS (Cluster automatic dynamic balance resource framework). At the appropriate point in time (determined by the user, the service impact minimum point in time), the CDBS triggers a cluster recovery action according to the cluster snapshot recorded in step 3.
First, all running traffic for all services is suspended (step 3 has taken a traffic task snapshot), and then the running of Yarn and all services involved is stopped. Next, the Yarn resources are readjusted according to the cluster backup snapshot recorded in the step 3And (5) restoring the clusters to the cluster distribution state before dynamic balance. After the cluster service resource distribution state and the service state are restored, the total number c of the clusters is unchanged, but the total number of resources of the Yarn service host is reduced to be original、/>And the total number of resources of the other services is restored to the initial value of step 1. And finally, recovering all online services according to the service task snapshot.
It should be noted that the foregoing merely illustrates the technical idea of the present invention and is not intended to limit the scope of the present invention, and that a person skilled in the art may make several improvements and modifications without departing from the principles of the present invention, which fall within the scope of the claims of the present invention.
Claims (8)
1. The method for predicting and automatically dynamically balancing the resources in the cluster based on the established resources is characterized by comprising the following steps:
step 1, dividing service types of a distributed computing system, periodically acquiring resource amount data of each type of service, and modeling acquired data of each service by using a Cluster Oracle;
step 2, predicting the service resource quantity required by one or more distributed computing systems in a future period based on the data model obtained in the step 1 and the collected historical resource usage data of one or more distributed computing systems, and obtaining an optimal allocation result of cluster service resources by combining the current cluster scale total resource quantity;
step 3, using Cluster Snapshot Service service to store the snapshot of the current cluster state;
step 4, triggering automatic dynamic balance of the cluster through an automatic dynamic balance resource framework of the cluster according to the optimal distribution result of the cluster service resources calculated in the step 2 at a proper time point after the current cluster snapshot is stored;
and 5, when detecting that the utilization rate of one or more distributed computing system resources reaches the adjustment threshold, triggering a recovery mode of the cluster automatic dynamic balance resource framework according to the cluster snapshot recorded in the step 3 at a proper time point to recover the cluster service resource distribution state and the service state.
2. The method for predicting and automatically dynamically balancing resources in a cluster based on a given resource according to claim 1, wherein in the step 2, the optimal allocation result is:
total cluster scale:
；
wherein the method comprises the steps ofThe total number of the hosts is m, and the total number of Memory resources is m; />In order to adjust the number of hosts to be accommodated,the Memory resource is adjusted;
the various service resources change into:
；
for the number of hosts before adjustment, +.>Memory resources before adjustment; if->、/>If the number is positive, the service will expand, if +.>、/>Negative, the shrinkage is indicated.
3. The method for predicting and automatically dynamically balancing resources within a cluster based on established resources according to claim 1, wherein the saving of the snapshot in step 3 comprises: the clusters distribute snapshots and business task snapshots.
4. The method for predicting and automatically dynamically balancing resources in a cluster based on established resources according to claim 3, wherein the cluster distribution snapshot is used for storing distribution, service configuration and service status of various service hosts in the current cluster; the service task snapshot is used for storing snapshots of all services running on various services, including running states and resource occupation conditions.
5. The method for automatically dynamically balancing resources within a cluster based on a prediction of a given resource according to claim 1, wherein in step 4, the process of automatically dynamically balancing the cluster comprises:
(1) Suspending all running traffic for all services, stopping running one or more distributed computing systems and all services involved;
(2) According to the optimal distribution result of the clusters, readjusting the resource scale of the distributed computing system to expand or contract the capacity of one or more distributed computing system resources;
(3) And recovering all online services according to the service task snapshot.
6. The method for predicting and automatically dynamically balancing resources in a cluster based on established resources according to claim 5, wherein the total resources of the cluster are unchanged and the total number of resources of other services is adjusted during the automatic dynamic balancing of the cluster.
7. The method for dynamically balancing resources within a cluster based on a prediction of a given resource according to claim 1, wherein the process of recovering the distribution state and the traffic state of the service resources of the cluster in step 5 comprises:
(1) Suspending all running traffic for all services, stopping one or more distributed computing systems and all services involved;
(2) According to the cluster backup snapshot recorded in the step 3, readjusting the resource scale of the distributed computing system, and recovering the cluster to the cluster distribution state before dynamic balance;
(3) And recovering all online services according to the service task snapshot.
8. The method of claim 7, wherein after the cluster is restored to the cluster distribution state before dynamic balancing, the total number of clusters is unchanged, the total number of the service hosts of the one or more distributed computing systems is adjusted to the original number, and the total number of the resources of the other services is restored to the initial value of step 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311597960.6A CN117311992B (en) | 2023-11-28 | 2023-11-28 | Method for predicting and automatically dynamically balancing internal resources of cluster based on established resources |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311597960.6A CN117311992B (en) | 2023-11-28 | 2023-11-28 | Method for predicting and automatically dynamically balancing internal resources of cluster based on established resources |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117311992A true CN117311992A (en) | 2023-12-29 |
| CN117311992B CN117311992B (en) | 2024-02-13 |
Family
ID=89250201
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311597960.6A Active CN117311992B (en) | 2023-11-28 | 2023-11-28 | Method for predicting and automatically dynamically balancing internal resources of cluster based on established resources |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117311992B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110737529A (en) * | 2019-09-05 | 2020-01-31 | 北京理工大学 | cluster scheduling adaptive configuration method for short-time multiple variable-size data jobs |
| WO2020206705A1 (en) * | 2019-04-10 | 2020-10-15 | 山东科技大学 | Cluster node load state prediction-based job scheduling method |
| CN112905334A (en) * | 2021-02-02 | 2021-06-04 | 深信服科技股份有限公司 | Resource management method, device, electronic equipment and storage medium |
| CN113795826A (en) * | 2019-06-27 | 2021-12-14 | 英特尔公司 | Automated resource management for distributed computing |
| CN114189482A (en) * | 2021-12-14 | 2022-03-15 | 郑州阿帕斯数云信息科技有限公司 | Control method, device and system for cluster resources |
-
2023
- 2023-11-28 CN CN202311597960.6A patent/CN117311992B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020206705A1 (en) * | 2019-04-10 | 2020-10-15 | 山东科技大学 | Cluster node load state prediction-based job scheduling method |
| CN113795826A (en) * | 2019-06-27 | 2021-12-14 | 英特尔公司 | Automated resource management for distributed computing |
| CN110737529A (en) * | 2019-09-05 | 2020-01-31 | 北京理工大学 | cluster scheduling adaptive configuration method for short-time multiple variable-size data jobs |
| CN112905334A (en) * | 2021-02-02 | 2021-06-04 | 深信服科技股份有限公司 | Resource management method, device, electronic equipment and storage medium |
| CN114189482A (en) * | 2021-12-14 | 2022-03-15 | 郑州阿帕斯数云信息科技有限公司 | Control method, device and system for cluster resources |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117311992B (en) | 2024-02-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110888714B (en) | Scheduling method, scheduling device and computer readable storage medium for containers | |
| CN102868763B (en) | The dynamic adjusting method that under a kind of cloud computing environment, virtual web application cluster is energy-conservation | |
| CN104639626B (en) | A kind of multistage load estimation and cloud resource elasticity collocation method and monitoring configuration system | |
| CN113010260B (en) | Container number elastic expansion method and container number elastic expansion method system | |
| CN107249035B (en) | Shared repeated data storage and reading method with dynamically variable levels | |
| CN110289994B (en) | Cluster capacity adjusting method and device | |
| CN111381928B (en) | Virtual machine migration method, cloud computing management platform and storage medium | |
| CN104407926B (en) | A kind of dispatching method of cloud computing resources | |
| CN105868004B (en) | Scheduling method and scheduling device of service system based on cloud computing | |
| CN103581332B (en) | HDFS framework and pressure decomposition method for NameNodes in HDFS framework | |
| CN106250305A (en) | Self-adaptive control method for data acquisition period of monitoring system in cloud computing environment | |
| CN102413186B (en) | Resource scheduling method and device based on private cloud computing, and cloud management server | |
| CN103455577A (en) | Multi-backup nearby storage and reading method and system of cloud host mirror image file | |
| CN102307241A (en) | Cloud calculation resource disposition method based on dynamic prediction | |
| TW202133055A (en) | Method for establishing system resource prediction and resource management model through multi-layer correlations | |
| CN108132837A (en) | A kind of distributed type assemblies dispatch system and method | |
| CN111104226A (en) | Intelligent management system and method for multi-tenant service resources | |
| CN112165436A (en) | Flow control method, device and system | |
| CN108924202A (en) | A kind of the data disaster tolerance method and relevant apparatus of distributed type assemblies | |
| CN110083306A (en) | A kind of distributed objects storage system and storage method | |
| CN114564147A (en) | Self-adaptive adjustment method and system of data stream, hard disk and storage medium | |
| CN106973091B (en) | Distributed memory data redistribution method and system, and master control server | |
| CN116700903A (en) | Memory adjusting method, system, equipment and medium of cloud computing-oriented virtual machine | |
| CN109669636B (en) | Distributed intelligent storage system | |
| CN117149527B (en) | System and method for backing up and recovering server data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |