CN103078950A - Method for increasing starting speed of cluster physical computers under cloud environment - Google Patents
Method for increasing starting speed of cluster physical computers under cloud environment Download PDFInfo
- Publication number
- CN103078950A CN103078950A CN2013100179360A CN201310017936A CN103078950A CN 103078950 A CN103078950 A CN 103078950A CN 2013100179360 A CN2013100179360 A CN 2013100179360A CN 201310017936 A CN201310017936 A CN 201310017936A CN 103078950 A CN103078950 A CN 103078950A
- Authority
- CN
- China
- Prior art keywords
- cluster
- power
- starting
- computer
- power supply
- 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
Images
Landscapes
- Power Sources (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention provides a method for increasing starting speed of cluster physical computers under a cloud environment. A system comprises a node power counting unit and a starting control unit, wherein the node power counting unit is used for counting peak power of each cluster physical computer during starting and the power of each cluster physical machine during normal work; and the starting control unit is used for arranging an optimal starting strategy according to starting peak power of each computer so as to shorten starting time on the whole. The method comprises the following steps of: firstly, counting a rated load of a cluster power supply source and calculating maximum quantity of the computers allowed to be started each time by accumulating the acquired starting peak power of the computer through the node power counting unit; and secondly, sequentially arranging optimal quantity of the computers allowed to be started in the rated load of the cluster power supply source according to the rated load of the cluster power supply source, which is counted by the node power counting unit, and the starting peak power of each computer by the starting control unit, so as to shorten the starting time on the whole.
Description
Technical field
The present invention relates to the cloud computing technology field, specifically improve the method for cluster physical machine starting up speed under a kind of cloud environment.
Background technology
The development of current cloud computing and increasing industry expansion of application, task quantity in the execution of the environment of cloud computing also is more and more, the physical equipment quantity of therefore managing under the cloud computing environment also is more and more huger, thereby the management that causes at these cluster computers needs to spend the more time.
In the scope that the load of cluster power supply allows, improve to greatest extent the whole starting up speed of cluster physical computer.Computer is when start, the power of computer has a peak value, power during this peakedness ratio computer normal operation wants high, when a start of the computer in the cluster, the stack of each computer power peak value has the possibility above the load of cluster power supply, can cause the bad of electric power system like this.For safety opening terminal in the scope that allows in the load of cluster power supply, many keepers have taked the method for one one order start now, although this method does not have the danger that exceeds the load of cluster power supply, but cluster needs the more time of waste on the starting up speed of integral body.Converge the starting up speed of calculating physical machine cluster under the environment in order to improve, we propose a kind of method for starting computer based on computer booting peak value statistics, in the scope that cluster allows for electric loading, be many material device power-up at one time to greatest extent, reach the purpose that improves on the whole cluster physical machine starting up speed.
How a large amount of equipment reduces the time of expending when starting in the process of safeguarding, need again to ensure the also problem with regard to having become us to be concerned about of cluster power supply safety simultaneously.
The scheme of current computer manufacturer, service equipment under cloud environment, computer equipment in the cluster is adopted the mode of sequence starting when starting, although the power supply surge pressure in the time of can avoiding like this all devices and start simultaneously, but have very large defective in the efficient that starts, it is too many to expend time in.There is randomness, in the time of so just the equipment that has the hardware device problem might being used for preventing that load from increasing, distributes the at first computer of load.Cause easily systematic function to reduce, even the problem such as service disconnection.Converge the starting up speed of calculating physical machine cluster under the environment in order to improve, we propose a kind of general, flexibly, method for starting computer based on computer booting peak value statistics, in the scope that cluster allows for electric loading, be many material device power-up at one time to greatest extent, reach the purpose that improves on the whole cluster physical machine starting up speed.
Summary of the invention
Technical assignment of the present invention is to solve the deficiencies in the prior art, and the method that improves cluster physical machine starting up speed under a kind of cloud environment is provided.
Technical scheme of the present invention realizes that in the following manner system comprises node power statistic unit (1), start control unit (2), wherein:
Worker when the peak power when node power statistic unit (1) is used for the start of statistical cluster physical machine and cluster computer normal operation leads;
Start control unit (2) arranges best start strategy according to the opening computer peak power, reduces the available machine time on the whole;
Concrete steps are as follows:
1) node power statistic unit (1) statistical cluster power supply nominal load, the computer booting power peak that gets access to by adding up of peak power, calculates each maximum quantity that allows start;
2) cluster power supply nominal load and every opening computer peak power of counting on according to node power statistic unit (1) of start control unit (2), the orderly optimal number of starting shooting in the cluster power supply nominal load that is arranged in, reduce the available machine time on the whole, the time waste of avoiding tradition to start shooting and cause successively, the bulk velocity that provides cloud computing machine cluster material machine to start shooting.
Outstanding beneficial effect of the present invention: the present invention relates to a kind of cloud computing applied technical field, be specifically related to improve under a kind of cloud environment the method for cluster physical machine starting up speed.Under cloud environment, in the management process of physical machine, provide a kind of general, flexibly, the starting-up method of physical machine, in the scope that the load of cluster power supply allows, improve to greatest extent the whole starting up speed of cluster physical computer.Save time thereby reach, improve the purpose of start efficient.
Description of drawings
Fig. 1 is that one one common of cluster physical machine starts schematic diagram successively;
Fig. 2 is many parallel starting schematic diagrames in cluster power supply nominal load allowed band.
Embodiment
With reference to the accompanying drawings method of the present invention is done and described in detail.
Architecture of the present invention mainly comprises: node power statistic unit (1), start shooting control unit (2) wherein:
Worker when the peak power when node power statistic unit (1) mainly is the start of statistical cluster physical machine and cluster computer normal operation leads, for the start control unit improves the start peak-data of each node computer and the rated power of computer.
Start control unit (2) arranges best start strategy according to the opening computer peak power, reduces the available machine time on the whole.
Concrete steps are as follows:
1) node power statistic unit (1) statistical cluster power supply nominal load, the computer booting power peak that gets access to by adding up of peak power, calculates each maximum quantity that allows start;
2) cluster power supply nominal load and every opening computer peak power of counting on according to node power statistic unit (1) of start control unit (2), the orderly optimal number of starting shooting in the cluster power supply nominal load that is arranged in, reduce the available machine time on the whole, the time waste of avoiding tradition to start shooting and cause successively, the bulk velocity that provides cloud computing machine cluster material machine to start shooting.
Embodiment:
The rated power of cluster power supply unit is 260KW.Computer in the cluster has 300, if the order start of or a rack of a rack although do not have the danger that exceeds power supply unit rated power, can expend on the available machine time of whole cluster too much.Statistics according to statistic unit, power is 800W during every equipment operation, but the peak power when start is 1KW, if so simultaneously start, the peak power of so whole cluster is the rated power 260KW of the 300KW power supply unit that surpasses cluster, and the possibility of damage equipment is just arranged.
In order to ensure safety, reach simultaneously the purpose that guarantees starting up speed, the start control unit can select 260 computers to start shooting simultaneously, peak power when starting shooting like this is 260KW, within the rated power of cluster power supply unit, after peak value, more remaining 40 equipment are started shooting simultaneously.Thereby reach in the situation that guarantees whole group system safety, reduce the purpose of available machine time.
Except the disclosed technical characterictic of specification of the present invention, be the public office technology of those skilled in the art.
Claims (1)
1. improve the method for cluster physical machine starting up speed under the cloud environment, it is characterized in that system comprises node power statistic unit (1), start control unit (2), wherein:
Worker when the peak power when node power statistic unit (1) is used for the start of statistical cluster physical machine and cluster computer normal operation leads;
Start control unit (2) arranges best start strategy according to the opening computer peak power, reduces the available machine time on the whole;
Concrete steps are as follows:
1) node power statistic unit (1) statistical cluster power supply nominal load, the computer booting power peak that gets access to by adding up of peak power, calculates each maximum quantity that allows start;
2) cluster power supply nominal load and every opening computer peak power of counting on according to node power statistic unit (1) of start control unit (2), the orderly optimal number of starting shooting in the cluster power supply nominal load that is arranged in, reduce the available machine time on the whole, the time waste of avoiding tradition to start shooting and cause successively, the bulk velocity that provides cloud computing machine cluster material machine to start shooting.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013100179360A CN103078950A (en) | 2013-01-18 | 2013-01-18 | Method for increasing starting speed of cluster physical computers under cloud environment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013100179360A CN103078950A (en) | 2013-01-18 | 2013-01-18 | Method for increasing starting speed of cluster physical computers under cloud environment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103078950A true CN103078950A (en) | 2013-05-01 |
Family
ID=48155352
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2013100179360A Pending CN103078950A (en) | 2013-01-18 | 2013-01-18 | Method for increasing starting speed of cluster physical computers under cloud environment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103078950A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112486661A (en) * | 2019-09-12 | 2021-03-12 | 赵伟 | Automatic control method and system for peak-shifting startup of computer cluster |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101001171A (en) * | 2006-06-20 | 2007-07-18 | 华为技术有限公司 | Single board and method for controlling power supply of it |
| CN101599840A (en) * | 2009-07-07 | 2009-12-09 | 浪潮电子信息产业股份有限公司 | The method of server time-sharing power in a kind of large-scale cluster |
| CN102546288A (en) * | 2010-11-16 | 2012-07-04 | 美国博通公司 | Method and system for measuring aspects of a network having a plurality of network components |
-
2013
- 2013-01-18 CN CN2013100179360A patent/CN103078950A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101001171A (en) * | 2006-06-20 | 2007-07-18 | 华为技术有限公司 | Single board and method for controlling power supply of it |
| CN101599840A (en) * | 2009-07-07 | 2009-12-09 | 浪潮电子信息产业股份有限公司 | The method of server time-sharing power in a kind of large-scale cluster |
| CN102546288A (en) * | 2010-11-16 | 2012-07-04 | 美国博通公司 | Method and system for measuring aspects of a network having a plurality of network components |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112486661A (en) * | 2019-09-12 | 2021-03-12 | 赵伟 | Automatic control method and system for peak-shifting startup of computer cluster |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10014717B2 (en) | Power supply device and power supply method in data center | |
| CN104601492A (en) | Method and device for controlling business flow under NFV framework | |
| CN108255582A (en) | Method, system, equipment and the storage medium of java virtual machine garbage reclamations | |
| US10528113B2 (en) | Low latency computer system power reduction | |
| CN103365392B (en) | Internal storage management system | |
| US10331198B2 (en) | Dynamically adapting to demand for server computing resources | |
| CN104699485A (en) | Massive program management method and massive program construction method | |
| US9760454B2 (en) | High availability method and system for improving the utility of physical servers in cloud computing resource pool | |
| EP4429066A1 (en) | Charging and discharging control method and apparatus for energy storage system, controller, and energy storage system | |
| CN110601245A (en) | Method and system for checking transient overvoltage after extra-high voltage direct current fault | |
| CN104123183A (en) | Cluster assignment dispatching method and device | |
| CN103944940A (en) | Equipment and method for dynamically configuring proxy server | |
| CN103078950A (en) | Method for increasing starting speed of cluster physical computers under cloud environment | |
| CN105141009A (en) | Power supply management method, power supply management chip and electronic system | |
| CN105573860A (en) | STD (Suspend to disk) management method and STD management system used in computer system | |
| CN104008062A (en) | Internal memory management method and internal memory management device | |
| CN104281587A (en) | Connection establishing method and device | |
| CN107967172B (en) | Spark task dynamic migration method and system for heterogeneous storage | |
| CN108139969A (en) | A kind of internal memory configuring method, apparatus and system | |
| CN102521065B (en) | Method for dynamic loading of configuration file through FPGA (field programmable gate array) | |
| CN105137350A (en) | Wind turbine generator set performance monitoring method based on eight states | |
| US20190155359A1 (en) | Method for automatically managing the electricity consumption of a server farm | |
| US11204634B2 (en) | Power monitoring and reduction | |
| CN104460924A (en) | Design method for realizing off-peak power-up of server through BIOS | |
| CN103973811A (en) | High-availability cluster management method capable of conducting dynamic migration |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20130501 |
|
| WD01 | Invention patent application deemed withdrawn after publication |