CN103617090A - Energy saving method based on distributed management - Google Patents
Energy saving method based on distributed management Download PDFInfo
- Publication number
- CN103617090A CN103617090A CN201310661388.5A CN201310661388A CN103617090A CN 103617090 A CN103617090 A CN 103617090A CN 201310661388 A CN201310661388 A CN 201310661388A CN 103617090 A CN103617090 A CN 103617090A
- Authority
- CN
- China
- Prior art keywords
- node
- virtual machine
- management
- utilization factor
- physical
- 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
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses an energy saving method based on distributed management. The method comprises local management and overall management, management tasks are distributed to the local management of all nodes through the distributed management, and the overall management is only used for collecting the information of the local management and making decisions. The local management is mainly used for monitoring abnormal conditions, and migration is carried out by selecting a virtual machine with a too high or too low cpu use ratio at the physical node with abnormality. The overall management is mainly used for judging and evaluating all the physical nodes, and at last, the physical node which the virtual machine should be migrated to is obtained. A data center with a distributed virtual machine management method is adopted in a large-scale cloud computing system, reliable service quality can be guaranteed, and meanwhile the high efficiency and energy saving of the data center are achieved.
Description
Technical field
The present invention relates to cloud computing technical field of virtualization, be specifically related to a kind of saving energy consumption method based on distributed management.
Technical background
Cloud computing, is exactly to utilize system architecture technology that thousands of station servers are integrated in simple terms, for user provides resource flexibly, distributes and task scheduling ability.
In recent years, cloud computing had become research and the application focus of computation schema.Cloud computing mode needs the infrastructure of large-scale calculations to support.Yet the infrastructure of these large-scale calculations has but consumed a large amount of electric power resources, and this electrical source consumption is at increase year after year.According to the estimation of IDC, the angle of controlling from operating cost, in IT industry, Energy Consumption Cost has reached 25% of hardware purchase cost.And data just with 52% to meet average growth rate per annum constantly soaring.Traffic pressure when enterprise and big-and-middle-sized mechanism in the face of constantly changing, and while being the data of exponential rapid growth, need to protect data center ring, the characteristic of energy-conservation aspect give enough considerations and attention.
Here there are several key words: the one, ultra-large, comprise the quantity of the quantity of machine, user's quantity and concurrent task; The 2nd, resource consolidation, the server resource of thousands of can gather does something, such as storage mass data, or processes a large-scale task; The 3rd, to pay flexibly with fast, large-scale server resource can be allocated flexibly, by application demand, resolves into several virtual resource pools, supports rapidly a large amount of concurrent requests or operation.Yet the infrastructure of these large-scale calculations has but consumed a large amount of electric power resources, and this electrical source consumption is at increase year after year.In fact, most electrical source consumption comes from the consumption of idling-resource.Studies show that 70% left and right that when physical node of the data center power supply that (CPU utilization factor is 0) consumes is at one's leisure full load, (CPU utilization factor is 100%) consumes.Virtual is the important technology that reduces electrical source consumption.It can make a plurality of virtual machine instance operate in same physical machine.Monitor of virtual machine, such as Xen, KVM etc., can support operating virtual machine by network, to carry out real-time migration in different physical machine, and have no sensation for user the break period of real-time migration.Therefore, data center can realize energy-conservation by the way that the lower virtual machine of some loads is incorporated in less physical machine.
At present, cloud computing data center mainly adopts centralized management method to integrate virtual machine, is about to the virtual machine on All hosts in data center and carries out unified management, thereby reach energy-conservation object.Although centralized management method can obtain certain energy-saving effect, but still there are some problems: (1) is because centralized management method adopts Continuous optimization, and every suboptimization is all that all virtual machines are carried out once to combination again, so will certainly cause a large amount of virtual machine (vm) migrations in management process.And the migration of virtual machine can increase the consumption of node cpu resource, the number of times that migration occurs is more, and the cpu resource of consumption is just more, and then the electrical source consumption of data center is also more.On the other hand, the virtual machine in migration is declined by long service performance will.So be necessary to reduce the generation of virtual machine (vm) migration number of times in management process.(2) centralized management method is only applicable to the cluster of small-scale, and when system extension arrives large-scale cloud computing system, centralized management method will be difficult to manage.
Summary of the invention
The technical problem to be solved in the present invention is: centralized management method adopts optimization in real time at the virtual machine of management process Zhong Dui data center, its objective is and guarantees that in each configuration of virtual machine be constantly all optimum.Therefore, adopt the data center of centralized management method will produce a large amount of virtual machine (vm) migrations, also increased the electrical source consumption producing because of migration simultaneously.
The technical solution adopted in the present invention is:
A kind of power-economizing method based on distributed management, described method comprises local management and global administration, be different from centralized management method system is carried out to unified management, the local management Shang，Er global administration that distributed management is distributed to each node by managerial role is just responsible for collecting the information of local management and carrying out decision-making; Wherein:
Local management is mainly monitoring abnormal conditions, is occurring that abnormal physical node gets on to select the too high or too low virtual machine of cpu utilization factor to move;
Local management continues the CPU utilization factor of monitoring local node, and from operating in the monitor of virtual machine (VMM node, Virtual Machine Monitor) on, obtain the information of each virtual machine, its objective is according to the demand of resource is adjusted to the size of virtual machine and determine which virtual machine which time need to move away from this node.
Global administration judges, assesses all physical nodes, finally obtains the physical node that virtual machine should move to, and accepts the physical node of virtual machine;
Global administration operates on a host node, collects the nodal information from local management, and selects best node to receive the virtual machine that needs migration.
In order to improve the CPU utilization factor of node, system is provided with a threshold ones, when the CPU of node utilization factor lower than arrange threshold ones time, system will all be moved away the virtual machine on this node also closed node, to reach energy-conservation object.
In addition, in order to improve the service quality of system, system is also provided with Yi Ge wealthy family limit value, and this is because if the utilization factor of physical node CPU is too high, and on this node, a plurality of virtual machines will reduce the service quality to user because of contention resource so.When the CPU of node utilization factor is greater than the wealthy family limit value of setting, system will select the one or more virtual machines on this node to move, thereby reduces the CPU utilization factor of node.
Different from centralized management method, in distributed management process, do not adopt real-time optimization, but by the local management being distributed on node, the CPU utilization factor of node is monitored, if the CPU utilization factor of node is within wealthy family's limit value of setting and threshold ones, virtual machine just can not move so, if certain node determination is abnormal, when the utilization factor of this node cpu is greater than wealthy family's limit value or is less than threshold ones, system will be from occurring that abnormal physical node selects virtual machine to move; Choose other physical nodes and receive selected virtual machine.
The selection of virtual machine in described method, once monitor of virtual machine captures extremely, local management just starts from occurring that abnormal physical node selects virtual machine to move; Abnormal for eliminating, conventionally one or more virtual machines can be moved away from there is abnormal physical node.
If there is abnormal physical node, be the too high situation of CPU utilization factor, local management need to move one or more virtual machines so.The number of times occurring in order to reduce migration, first local management falls to sort by the current C PU utilization factor of all virtual machines on this node, then select wherein that virtual machine of CPU utilization factor maximum to move, if after migration, this node cpu utilization factor is still too high, continue that virtual machine that the sequence of migration CPU utilization factor is taken second place, until the abnormal elimination of this node.
If there is abnormal physical node, be the too low situation of CPU utilization factor, problem just becomes relatively simple so, and local management will also be closed all virtual machine (vm) migrations on node to other nodes by this node, to reach energy-conservation object.
The selection of physical node in described method, when the selection of virtual machine finishes, global administration will select physical node to receive the virtual machine that needs migration: first global administration judges in advance to all physical nodes, see whether to exist after receiving this virtual machine and there will not be abnormal physical node, if do not meet the physical node of such condition, so just not having migration occurs, if there is the physical node satisfying condition, first global administration carries out electrical source consumption assessment to all physical nodes that satisfy condition so, calculate these physical nodes poor with electrical source consumption afterwards before sink virtual machine, then therefrom select and produce the poor physical node of minimal power consumption, and by virtual machine (vm) migration to this node.
Beneficial effect of the present invention is:
Adopt the data center of distributed virtual machine management method in large-scale cloud computing system, when can guarantee reliability services quality,, realize the energy-efficient of data center.And the setting of high lower bound threshold value is a direction of the energy-conservation research of the data center of cloud computing from now on.
Accompanying drawing explanation
Fig. 1 is distribution management structure schematic diagram of the present invention.
Embodiment
With reference to the accompanying drawings, the present invention will be described in conjunction with the embodiments:
As shown in the figure, distribution management method comprises local management and global administration.Local management is mainly monitoring abnormal conditions, is occurring that abnormal physical node gets on to select the too high or too low virtual machine of cpu utilization factor to move.Global administration judges, assesses all physical nodes, finally obtains the physical node that virtual machine should move to, and accepts the physical node of virtual machine.
Local management continues the CPU utilization factor of monitoring local node, and from operating in monitor of virtual machine VMM node, obtain the information of each virtual machine, its objective is according to the demand of resource is adjusted to the size of virtual machine and determine which virtual machine which time need to move away from this node.Global administration operates on a host node, and it collects the nodal information from local management, and selects best node to receive the virtual machine that needs migration.Be different from centralized management method system is carried out to unified management, the local management Shang，Er global administration that distribution management method is distributed to each node by managerial role is just responsible for collecting the information of local management and carrying out decision-making.In order to improve the CPU utilization factor of node, system is provided with a threshold ones.When the CPU of node utilization factor lower than arrange threshold ones time, system will all be moved away the virtual machine on this node also closed node, to reach energy-conservation object.In addition,, in order to improve the service quality of system, system is also provided with Yi Ge wealthy family limit value.This is because if the utilization factor of physical node CPU is too high, and on this node, a plurality of virtual machines will reduce the service quality to user because of contention resource so.When the CPU of node utilization factor is greater than the wealthy family limit value of setting, system will select the one or more virtual machines on this node to move, thereby reduces the CPU utilization factor of node.
Claims (7)
1. the power-economizing method based on distributed management, it is characterized in that: described method comprises local management and global administration, the local management Shang，Er global administration that wherein said distributed management is distributed to each node by managerial role is just responsible for collecting the information of local management and carrying out decision-making;
Local management is mainly monitoring abnormal conditions, is occurring that abnormal physical node gets on to select the too high or too low virtual machine of cpu utilization factor to move;
Global administration judges, assesses all physical nodes, finally obtains the physical node that virtual machine should move to.
2. a kind of power-economizing method based on distributed management according to claim 1, it is characterized in that: in order to improve the CPU utilization factor of node, system is provided with a threshold ones, when the CPU of node utilization factor lower than arrange threshold ones time, system will all be moved away the virtual machine on this node and closed node, to reach energy-conservation object.
3. a kind of power-economizing method based on distributed management according to claim 1 and 2, it is characterized in that: in order to improve the service quality of system, system is also provided with Yi Ge wealthy family limit value, when the CPU of node utilization factor is greater than the wealthy family limit value of setting, system will select the one or more virtual machines on this node to move, thereby reduces the CPU utilization factor of node.
4. a kind of power-economizing method based on distributed management according to claim 3, it is characterized in that: the selection of virtual machine in described method, once monitor of virtual machine captures extremely, local management just starts from occurring that abnormal physical node selects virtual machine to move; Abnormal for eliminating, one or more virtual machines are moved away from there is abnormal physical node.
5. a kind of power-economizing method based on distributed management according to claim 4, it is characterized in that: if there is abnormal physical node, be the too high situation of CPU utilization factor, local management need to move one or more virtual machines so, the number of times occurring in order to reduce migration, first local management falls to sort by the current C PU utilization factor of all virtual machines on this node, then select wherein that virtual machine of CPU utilization factor maximum to move, if after migration, this node cpu utilization factor is still too high, continue that virtual machine that the sequence of migration CPU utilization factor is taken second place, until the abnormal elimination of this node.
6. a kind of power-economizing method based on distributed management according to claim 4, it is characterized in that: if there is abnormal physical node, be the too low situation of CPU utilization factor, local management will also be closed all virtual machine (vm) migrations on node to other nodes by this node, to reach energy-conservation object.
7. a kind of power-economizing method based on distributed management according to claim 4, it is characterized in that: the selection of physical node in described method, when the selection of virtual machine finishes, global administration will select physical node to receive the virtual machine that needs migration: first global administration judges in advance to all physical nodes, see whether to exist after receiving this virtual machine and there will not be abnormal physical node, if do not meet the physical node of such condition, so just not having migration occurs, if there is the physical node satisfying condition, first global administration carries out electrical source consumption assessment to all physical nodes that satisfy condition so, calculate these physical nodes poor with electrical source consumption afterwards before sink virtual machine, then therefrom select and produce the poor physical node of minimal power consumption, and by virtual machine (vm) migration to this node.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310661388.5A CN103617090A (en) | 2013-12-10 | 2013-12-10 | Energy saving method based on distributed management |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310661388.5A CN103617090A (en) | 2013-12-10 | 2013-12-10 | Energy saving method based on distributed management |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103617090A true CN103617090A (en) | 2014-03-05 |
Family
ID=50167793
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310661388.5A Pending CN103617090A (en) | 2013-12-10 | 2013-12-10 | Energy saving method based on distributed management |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103617090A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104572307A (en) * | 2015-01-30 | 2015-04-29 | 无锡华云数据技术服务有限公司 | Method for flexibly scheduling virtual resources |
| CN106301843A (en) * | 2015-05-28 | 2017-01-04 | 亿阳信通股份有限公司 | A kind of cloud platform safeguards system and method |
| CN107566319A (en) * | 2016-06-30 | 2018-01-09 | 中央大学 | The instant transfer method of virtual machine |
| CN109062669A (en) * | 2018-08-07 | 2018-12-21 | 郑州云海信息技术有限公司 | Virtual machine migration method and system under a kind of Random Load |
| CN111930315A (en) * | 2020-08-21 | 2020-11-13 | 北京天融信网络安全技术有限公司 | Data access method, data access device and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1329809A1 (en) * | 2002-01-18 | 2003-07-23 | Hewlett-Packard Company, A Delaware Corporation | Distributed computing system and method |
| CN101382912A (en) * | 2008-09-02 | 2009-03-11 | 中国科学院软件研究所 | Method for optimizing resource sensitive property orienting application server and system thereof |
| US20130047153A1 (en) * | 2011-08-19 | 2013-02-21 | Hitachi, Ltd. | Method and apparatus to improve efficiency in the use of resources in data center |
| CN102981910A (en) * | 2012-11-02 | 2013-03-20 | 曙光云计算技术有限公司 | Realization method and realization device for virtual machine scheduling |
-
2013
- 2013-12-10 CN CN201310661388.5A patent/CN103617090A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1329809A1 (en) * | 2002-01-18 | 2003-07-23 | Hewlett-Packard Company, A Delaware Corporation | Distributed computing system and method |
| CN101382912A (en) * | 2008-09-02 | 2009-03-11 | 中国科学院软件研究所 | Method for optimizing resource sensitive property orienting application server and system thereof |
| US20130047153A1 (en) * | 2011-08-19 | 2013-02-21 | Hitachi, Ltd. | Method and apparatus to improve efficiency in the use of resources in data center |
| CN102981910A (en) * | 2012-11-02 | 2013-03-20 | 曙光云计算技术有限公司 | Realization method and realization device for virtual machine scheduling |
Non-Patent Citations (1)
| Title |
|---|
| 马飞等: "云计算中高效节能的虚拟机分布式管理", 《计算机工程》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104572307A (en) * | 2015-01-30 | 2015-04-29 | 无锡华云数据技术服务有限公司 | Method for flexibly scheduling virtual resources |
| CN104572307B (en) * | 2015-01-30 | 2019-03-05 | 无锡华云数据技术服务有限公司 | The method that a kind of pair of virtual resource carries out flexible scheduling |
| CN106301843A (en) * | 2015-05-28 | 2017-01-04 | 亿阳信通股份有限公司 | A kind of cloud platform safeguards system and method |
| CN107566319A (en) * | 2016-06-30 | 2018-01-09 | 中央大学 | The instant transfer method of virtual machine |
| CN109062669A (en) * | 2018-08-07 | 2018-12-21 | 郑州云海信息技术有限公司 | Virtual machine migration method and system under a kind of Random Load |
| CN111930315A (en) * | 2020-08-21 | 2020-11-13 | 北京天融信网络安全技术有限公司 | Data access method, data access device and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhou et al. | Minimizing SLA violation and power consumption in Cloud data centers using adaptive energy-aware algorithms | |
| Zhou et al. | Virtual machine placement algorithm for both energy-awareness and SLA violation reduction in cloud data centers | |
| Buyya et al. | Sustainable cloud computing: foundations and future directions | |
| CN103617090A (en) | Energy saving method based on distributed management | |
| CN104038392A (en) | Method for evaluating service quality of cloud computing resources | |
| CN103761146A (en) | Method for dynamically setting quantities of slots for MapReduce | |
| Arthi et al. | Energy aware cloud service provisioning approach for green computing environment | |
| CN107203255A (en) | Power-economizing method and device are migrated in a kind of network function virtualized environment | |
| CN104360724B (en) | A kind of heat dissipating method of the blade server based on job scheduling | |
| CN102339233A (en) | Cloud computing centralized management platform | |
| Zhang et al. | An energy and SLA-aware resource management strategy in cloud data centers | |
| Niu et al. | JouleMR: Towards cost-effective and green-aware data processing frameworks | |
| CN104468710A (en) | Mixed big data processing system and method | |
| CN103970256A (en) | Energy saving method and system based on memory compaction and CPU dynamic frequency modulation | |
| CN102855157A (en) | Method for comprehensively scheduling load of servers | |
| CN105528054A (en) | Cluster system comprehensive scheduling energy saving method and device | |
| CN107197013B (en) | Energy-saving system for enhancing cloud computing environment | |
| Hema et al. | An Efficient Framework for Utilizing Underloaded Servers in Compute Cloud. | |
| Wang et al. | An efficient energy-aware and service quality improvement strategy applied in cloud computing | |
| Patel et al. | Efficient resource allocation strategy to improve energy consumption in cloud data centers | |
| Sun et al. | Based on QoS and energy efficiency virtual machines consolidation techniques in cloud | |
| CN102521715A (en) | Method and system capable of controlling resource allocation of application system | |
| Manzoor et al. | An analysis of energy-efficient approaches used for virtual machines and data centres | |
| Son et al. | Multi-objective optimization method for resource scaling in cloud computing | |
| CN107193362B (en) | Energy-saving device for enhancing cloud computing environment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140305 |
|
| RJ01 | Rejection of invention patent application after publication |