CN108632357B - Data center network area division method, device and equipment - Google Patents

Data center network area division method, device and equipment Download PDF

Info

Publication number
CN108632357B
CN108632357B CN201810321126.7A CN201810321126A CN108632357B CN 108632357 B CN108632357 B CN 108632357B CN 201810321126 A CN201810321126 A CN 201810321126A CN 108632357 B CN108632357 B CN 108632357B
Authority
CN
China
Prior art keywords
network
energy storage
data center
center network
connection
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.)
Active
Application number
CN201810321126.7A
Other languages
Chinese (zh)
Other versions
CN108632357A (en
Inventor
赵志军
金艳
徐晨
金军
沈元聪
周强
罗超
赵振敏
曹军威
杨刚
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huachuang Branch Of Jiaxing Hengguang Electric Power Construction Co ltd
Tsinghua University
Jiaxing Power Supply Co of State Grid Zhejiang Electric Power Co Ltd
Original Assignee
Huachuang Branch Of Jiaxing Hengguang Electric Power Construction Co ltd
Tsinghua University
Jiaxing Power Supply Co of State Grid Zhejiang Electric Power Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Huachuang Branch Of Jiaxing Hengguang Electric Power Construction Co ltd, Tsinghua University, Jiaxing Power Supply Co of State Grid Zhejiang Electric Power Co Ltd filed Critical Huachuang Branch Of Jiaxing Hengguang Electric Power Construction Co ltd
Priority to CN201810321126.7A priority Critical patent/CN108632357B/en
Publication of CN108632357A publication Critical patent/CN108632357A/en
Application granted granted Critical
Publication of CN108632357B publication Critical patent/CN108632357B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
    • G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/06—Management of faults, events, alarms or notifications
    • H04L41/0654—Management of faults, events, alarms or notifications using network fault recovery
    • H04L41/0659—Management of faults, events, alarms or notifications using network fault recovery by isolating or reconfiguring faulty entities
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/06—Management of faults, events, alarms or notifications
    • H04L41/0677—Localisation of faults
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/12—Discovery or management of network topologies

Landscapes

  • Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Probability & Statistics with Applications (AREA)
  • Supply And Distribution Of Alternating Current (AREA)

Abstract

The invention provides a data center network area division method, a device and equipment. The method comprises the following steps: acquiring a network-energy storage characteristic matrix of a data center network according to structural parameters of the data center network; the network-energy storage characteristic matrix is used for describing connection and energy storage relation among servers of the data center network; performing spectral clustering with the characteristic value quantity of K on the network-energy storage characteristic matrix based on the preset number K of the regions to obtain the region corresponding to each target device of the data center network; and performing area division according to the area corresponding to each target device, and dividing the data center network into K areas. The invention divides the region, shields the detailed characteristics of regional energy scheduling and can realize fast and high-efficiency energy management scheduling.

Description

Data center network area division method, device and equipment
Technical Field
The embodiment of the invention relates to the field of network division, in particular to a method, a device and equipment for dividing a data center network area.
Background
At present, with the rapid development of energy internet, a service concept taking a user as a core and a high-efficiency and rapid data interconnection platform thereof have great impact on the operation concepts of production, management, consumption and the like of the traditional energy industry. Compared with the traditional power grid, the energy internet emphasizes the information and data interconnection of production and consumption units taking users as cores, so that new energy consumption is promoted. The basic purpose is that through data interconnection, the power generation demand of new energy enterprises and the power consumption demand of power consumers are butted, and supply and demand are fully matched through power control, so that the utilization of new energy is promoted, and the use efficiency of energy is improved.
An important information communication infrastructure of the energy internet is a data center based on big data and cloud computing technology. The data center is a product of deep fusion of an information communication technology, a power electronic technology, a computer technology and an internet technology, can realize large-scale and high-speed information data processing, transmission and storage, is a main support for intelligent operation of an energy internet, and is an intelligence brain for the operation of a modern social system.
A data center may be referred to as a "cloud," which includes dense, homogeneous or heterogeneous computing, communication, and storage resources. Whether the cloud is a public cloud, a private cloud or a mixed cloud, the virtual machine is used as a resource providing unit and specifically corresponds to computing, communication and storage entities of different models. With the continuous development of data centers, the number and performance of related resources are in an explosive growth trend, and the heterogeneity is also increased, so that while the performance is greatly improved, the processing, transmission delay and complexity of related services are undoubtedly increased, but the requirements (such as bandwidth, processing delay and the like) on the quality of the services are continuously reduced, which brings great challenges to the management and scheduling of the data centers. Based on the heterogeneity and the quantity density of the computing entities, it is necessary to perform regional layering and hierarchical management on the computing entities, so that the data center is divided into essential key technologies, and the data center is a main support for performing energy management, task management, fault management and energy conservation and consumption reduction.
The data center area division is similar to the traditional power grid area division which is generally based on region intervals and the like. However, with the continuous advance of urbanization, industrial parks and urban areas with a huge geographic range are formed continuously. At this time, the conventional region division method according to the region interval is not applicable.
Disclosure of Invention
Aiming at the problems in the prior art, the embodiment of the invention provides a data center network region dividing method, device and equipment, which are suitable for the development of an urbanization process, partition is carried out according to network structure parameters of a super-large data center, the detailed characteristics of regional energy scheduling are shielded, and energy management and scheduling can be faster and more efficient.
In a first aspect, an embodiment of the present invention provides a method for dividing a data center network area, where the method includes:
acquiring a network-energy storage characteristic matrix of a data center network according to structural parameters of the data center network; the network-energy storage characteristic matrix is used for describing connection and energy storage relation among servers of the data center network;
performing spectral clustering with the characteristic value quantity of K on the network-energy storage characteristic matrix based on the preset area quantity K to obtain an area corresponding to each target device of the data center network, wherein K is an integer greater than 1;
and performing area division according to the area corresponding to each target device, and dividing the data center network into K areas.
In a second aspect, an embodiment of the present invention provides a data center network area dividing device, including:
the network-energy storage characteristic module is used for acquiring a network-energy storage characteristic matrix of the data center network according to the structural parameters of the data center network; the network-energy storage characteristic matrix is used for describing connection and energy storage relation among servers of the data center network;
the spectral clustering module is used for performing spectral clustering with the characteristic value quantity of K on the network-energy storage characteristic matrix based on the preset number K of regions to obtain a region corresponding to each target device of the data center network, wherein K is an integer greater than 1; and
and the area division module is used for carrying out area division according to the area corresponding to each target device and dividing the data center network into K areas.
In a third aspect, an embodiment of the present invention provides a data center network area dividing device, where the data center network area dividing device includes:
at least one processor; and
at least one memory communicatively coupled to the processor, wherein:
the memory stores program instructions executable by the processor, and the processor calls the program instructions to execute the data center network area partitioning method according to the first aspect of the embodiments of the present invention and the method according to any optional embodiment thereof.
In a fourth aspect, a non-transitory computer-readable storage medium is provided, where the non-transitory computer-readable storage medium stores computer instructions for executing the data center network area partitioning method according to the first aspect of the embodiments of the present invention and the method according to any optional embodiment thereof.
According to the data center network region division method provided by the embodiment of the invention, the spectral clustering of the network-energy storage characteristic matrix is carried out through the connection and energy storage relation disclosed by the structural parameters of the data center network, the problems of network connection, energy, electric power and electric efficiency are considered during region division, the method is suitable for the development of the urbanization process, and the optimal network region division is realized; by dividing the network region through the embodiment of the invention, the detailed characteristics of regional energy scheduling are shielded, and the energy management and scheduling can be faster and more efficient.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of a method for dividing a data center network area according to an embodiment of the present invention;
FIG. 2 is a block diagram of a data center network partition management according to an embodiment of the present invention;
fig. 3 is a schematic diagram of a framework of a data center network area dividing device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some embodiments, but not all embodiments, of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Fig. 1 is a schematic flow chart of a data center network region division method according to an embodiment of the present invention, where the data center network region division method shown in fig. 1 includes:
s100, acquiring a network-energy storage characteristic matrix of a data center network according to structural parameters of the data center network; the network-energy storage characteristic matrix is used for describing connection and energy storage relation among servers of the data center network
Specifically, the structural parameters include connection bandwidth between servers, connection delay, energy storage device capacity, power transmission power, and power transmission efficiency.
Specifically, the data center network in the embodiment of the invention refers to an energy internet forming a data center based on big data and a cloud computing technology. The energy internet comprises a plurality of servers, each server is connected with an energy storage device, and the energy internet further comprises an energy router, and the energy router is used for achieving network routing.
In step S100, network connection bandwidth between every two servers in the data center network, network delay between every two servers, capacity of energy storage devices associated with each server, maximum power of power transmission between the energy storage devices, power transmission efficiency between the energy storage devices, and the like need to be counted. And then constructing a feature matrix according to the obtained network structure parameters.
S200, performing spectral clustering with the characteristic value quantity of K on the network-energy storage characteristic matrix based on the preset number K of regions to obtain a region corresponding to each target device of the data center network, wherein K is an integer greater than 1;
specifically, the preset number K of areas indicates that the data center network is scheduled to be divided into K areas; performing spectral clustering with K eigenvalues on the network-energy storage characteristic matrix to obtain an N-dimensional vector S1×N，S1×NAll the elements are [1, K ]]The integers between the two correspond to the area of each target device of the data center network respectively.
Specifically, the target device may be an energy router.
And S300, performing area division according to the area corresponding to each target device, and dividing the data center network into K areas.
Specifically, after the area corresponding to each target device is obtained in step S200, the target devices in the K areas are determined, so that the data center network is divided into K areas according to the area to which each target device belongs.
The embodiment of the invention carries out the spectral clustering of the network-energy storage characteristic matrix through the connection and energy storage relation disclosed by the structural parameters of the data center network, gives consideration to the problems of network connection, energy, electric power and electric efficiency during the area division, is suitable for the development of the urbanization process and realizes the optimal network area division; the embodiment of the invention divides the network area, shields the detailed characteristics of area energy scheduling, and can realize faster and more efficient energy management and scheduling
In an optional embodiment, in step S100, the obtaining a network-energy storage feature matrix of the data center network according to the structural parameters of the data center network specifically includes:
s100.1, acquiring a network connection characteristic matrix according to connection bandwidth and connection time delay between servers of a data center network;
specifically, assume the number N of servers in the data center network and number them IDC1,IDC2,…IDCN. Assume server IDCjIDC with serveriAnd (i ≠ j) network connection is formed, the network connection bandwidth is respectively tested, and the analogy is repeated, so that the network bandwidth between any two servers in the data center network is obtained.
In particular, assume a server IDCjIDC with serveriAnd (i ≠ j) is connected with each other through a network, the network connection time delay (mu s) is respectively tested, and the rest is done in the same way, so that the network time delay (ms) between any two servers in the data center is obtained.
S100.2, acquiring an energy storage connection characteristic matrix according to the capacity of energy storage equipment of a server of the data center network, the power transmission power among the energy storage equipment and the circuit transmission efficiency;
in particular, assume a server IDC1Capacity of connected energy storage device is p1(kW/h), server IDC1,IDC2,…IDCNAcquiring the capacity of energy storage equipment connected to each server for each node;
in particular, assume a server IDCjEnergy storage device and server IDCiAnd (i ≠ j) energy storage devices are connected through power, so that the maximum power of power transmission is obtained respectively, and by analogy, the power transmission power between the energy storage devices of any two servers in the data center network is obtained.
Assume server IDCjEnergy storage device and server IDCiAnd (i ≠ j) energy storage devices are connected through power, so that the transmission efficiency of power transmission is obtained respectively, and by analogy, the power transmission efficiency between the energy storage devices of any two servers in the data center network is obtained.
And S100.3, acquiring a network-energy storage characteristic matrix of the data center network according to the network connection characteristic matrix and the energy storage connection characteristic matrix.
Specifically, the network connection characteristic matrix and the energy storage connection characteristic matrix are superposed to obtain a network-energy storage characteristic matrix C of the data center networkN×N：
CN×N＝TBN×NPQEN×N，
Wherein, TBN×NFor network connection feature matrices, PQEN×NThe characteristic matrix is connected to the energy storage.
Based on the above embodiment, in step S100.1, the obtaining a network connection feature matrix according to the connection bandwidth and the connection delay between the servers of the data center network specifically includes:
for any two servers with connection relation in the data center network, acquiring the connection bandwidth between the two servers to obtain a network connection bandwidth matrix BN×N：
Figure BDA0001625291560000071
Wherein, bij,i＝1,...,N,j＝1,...,NIs the connection bandwidth between server i and server j;
for any two servers in the data center network with connection relation, obtaining the connection time delay between the two servers to obtain a network connection time delay matrix TN×N：
Figure BDA0001625291560000072
Wherein, tij,i＝1,...,N,j＝1,...,NIs the connection delay between server i and server j;
according to the network connection bandwidth matrix BN×NAnd said network connection delay matrix TN×NObtaining a network connection characteristic matrix TBN×N：
TBN×N＝BN×N·/TN×N。
Based on the above embodiment, step S100.2 is to obtain an energy storage connection feature matrix according to the energy storage device capacity of the server of the data center network, the power transmission power between the energy storage devices, and the circuit transmission efficiency, and specifically includes:
acquiring the capacity of energy storage equipment of each server of the data center network, and diagonalizing the capacity of the energy storage equipment to obtain a diagonal matrix PN×N＝diag(p1,p2…pN) Wherein p isi,i＝1,...,NCapacity of the energy storage device to which server i is connected;
for energy storage devices of two servers with any connection relation, acquiring power transmission power between the energy storage devices of the two servers to obtain a power transmission power matrix QN×N：
Figure BDA0001625291560000073
Wherein q isij,i＝1,...,N,j＝1,...,NTransmitting power for power between the energy storage device of the server i and the energy storage device of the service j;
for the energy storage devices of two servers with any connection relation, obtaining the power transmission efficiency between the energy storage devices of the two servers to obtain a power transmission efficiency matrix EN×N：
Figure BDA0001625291560000081
Wherein e isij,i＝1,...,N,j＝1,...,NThe power transmission efficiency between the energy storage device of the server i and the energy storage device of the service j;
according to the diagonal matrix PN×NThe power transmission power matrix QN×NAnd the power transmission efficiency matrix EN×NObtaining an energy storage connection characteristic matrix PQEN×N：
PQEN×N＝PN×N·QN×N·EN×N。
Fig. 2 is a schematic diagram of data center network partition management according to an embodiment of the present invention, please refer to fig. 2, in the above embodiment, a feature matrix is constructed by obtaining structural parameters of a data center network, that is, a connection bandwidth between any two servers, a connection delay between any two servers, an energy storage device capacity of each server, an electric power transmission power between energy storage devices of any two servers, and an electric power transmission efficiency between energy storage devices of any two servers, and the feature matrix includes a network connection feature matrix and a network connection feature matrix, and a network-energy storage feature matrix is established by the network connection feature matrix and the network connection feature matrix, and network area division is performed by performing spectral clustering on the network-energy storage feature matrix; the network needs to be divided into several areas, and if the network needs to be divided into K areas, spectral clustering with the characteristic value of K is performed on the network-energy storage characteristic matrix, so that the key equipment is divided into corresponding areas, and area division is completed.
Based on the above embodiment, in step S300, the performing area division according to the area corresponding to each target device, and dividing the data center network into K areas further includes:
the management entity of each region of the data center network respectively manages and schedules the energy in each region, and the whole energy supply and demand balance is realized in a decentralized cooperation mode; and/or
The management entity of each region of the data center network respectively selects the matched services in the respective region and optimizes the resources in the region; and the scheduling and distribution of related services are realized among the areas in a negotiation or game mode.
After the area division of the data center network is completed through the embodiment, energy management and service management in each area can be realized according to the divided areas. Specifically, each region may select a computer (or corresponding computing resource) as a management entity (or referred to as an administrator) of the region, so as to manage the operation and scheduling of resource entities in the region.
When the data center network needs energy management and scheduling, according to the region division and the generated energy and load approximation of the nodes in each region, the energy management and scheduling can be firstly carried out in the region to realize supply and demand balance, and when the supply and demand balance cannot be realized in the region, the supply and demand balance is realized through energy scheduling among the regions. Because the detailed features of regional energy scheduling are masked, regional partitioning can make energy management and scheduling faster and more efficient. Specifically, when energy consumption in the data center network needs to be managed and controlled, each area management entity manages and schedules energy in an area, and the overall energy supply and demand balance of the data center network is realized in a decentralized cooperation mode.
When service allocation and scheduling are performed, the data center network needs to complete intelligent service management and scheduling according to the matching of services and computing, communication and storage resources. After the area division is based, the homogeneity of the related resources in the area brings great convenience to the parallel operation and scheduling of the related programs, and resource waste is avoided. Tasks with certain characteristics are allocated to resource entities with matched characteristics in one area, so that observability, controllability and manageability of service are improved, convenience is brought to fault processing, and overall performance of the data center is improved. Specifically, when the services of the data center need to be managed and controlled, each regional management entity selects the matched services, and then optimizes the resources in the allocated region to complete related calculation, communication and storage tasks. The scheduling and distribution of related services can be realized by negotiation or game among the areas.
Based on the above embodiment, in step S300, the performing area division according to the area corresponding to each target device, and dividing the data center network into K areas further includes:
when a fault occurs, the management entity of the fault area performs fault detection on the jurisdiction area and sends state information to the adjacent area according to the fault detection condition so as to assist the fault location of the adjacent area.
After the area division of the data center network is completed through the embodiment, the complexity of fault location is greatly reduced and the location precision can be greatly improved after the area division is performed, and the reliability of fault location and the fault isolation and recovery performance are greatly enhanced based on the homogeneity of relevant resource entities in the area, so that a complex fault location algorithm and a large amount of consumption of computing resources are avoided. Specifically, when a fault occurs, each regional management entity performs fault judgment on the regional resource, and as the scale of the region is far smaller than that of the whole data center network, the fault judgment becomes simpler and more efficient. And according to the fault detection condition, the area management entity sends state information to the adjacent area to assist the fault location of the adjacent area. After fault positioning is realized, the local area management entity can realize the isolation and repair of the fault device, and the fault processing efficiency is greatly improved.
Further, when the goal of energy saving and consumption reduction of the data center network needs to be achieved, different areas make self-adaptive energy saving and consumption reduction strategies according to the characteristics of the areas, and respectively suitable energy saving and consumption reduction technologies are adopted to achieve the maximization of the energy saving effect. Meanwhile, for the energy-saving tasks which are difficult to complete in a single area, the energy conservation and the consumption reduction can be realized through the advantage complementation in a cooperation mode among the areas.
When the network state changes significantly (topology change, resource allocation, physical composition, etc.), the repartitioning of the management area of the data center network needs to be triggered, and the area management entity needs to be reassigned, so that the automatic and programmed management is realized, and the high efficiency of the operation of the data center network is ensured.
The embodiment of the present invention further provides a data center network area dividing device, including:
the network-energy storage characteristic module is used for acquiring a network-energy storage characteristic matrix of the data center network according to the structural parameters of the data center network; the network-energy storage characteristic matrix is used for describing connection and energy storage relation among servers of the data center network;
the spectral clustering module is used for performing spectral clustering with the characteristic value quantity of K on the network-energy storage characteristic matrix based on the preset number K of regions to obtain a region corresponding to each target device of the data center network, wherein K is an integer greater than 1; and
and the area division module is used for carrying out area division according to the area corresponding to each target device and dividing the data center network into K areas.
The apparatus of the embodiment of the present invention may be used to implement the technical solution of the data center network area partitioning method embodiment shown in fig. 1, and the implementation principle and the technical effect are similar, which are not described herein again.
Fig. 3 is a schematic frame diagram of a data center network area dividing device according to an embodiment of the present invention. Referring to fig. 3, the apparatus includes: a processor (processor)310, a communication Interface (communication Interface)320, a memory (memory)330 and a bus 340, wherein the processor 310, the communication Interface 320 and the memory 330 complete communication with each other through the bus 340. The processor 310 may call logic instructions in the memory 330 to perform methods comprising: acquiring a network-energy storage characteristic matrix of a data center network according to structural parameters of the data center network; the network-energy storage characteristic matrix is used for describing connection and energy storage relation among servers of the data center network; performing spectral clustering with the characteristic value quantity of K on the network-energy storage characteristic matrix based on the preset area quantity K to obtain an area corresponding to each target device of the data center network, wherein K is an integer greater than 1; and performing area division according to the area corresponding to each target device, and dividing the data center network into K areas.
An embodiment of the present invention discloses a computer program product, which includes a computer program stored on a non-transitory computer readable storage medium, the computer program including program instructions, when the program instructions are executed by a computer, the computer can execute the methods provided by the above method embodiments, for example, the method includes: acquiring a network-energy storage characteristic matrix of a data center network according to structural parameters of the data center network; the network-energy storage characteristic matrix is used for describing connection and energy storage relation among servers of the data center network; performing spectral clustering with the characteristic value quantity of K on the network-energy storage characteristic matrix based on the preset area quantity K to obtain an area corresponding to each target device of the data center network, wherein K is an integer greater than 1; and performing area division according to the area corresponding to each target device, and dividing the data center network into K areas.
Embodiments of the present invention provide a non-transitory computer-readable storage medium, which stores computer instructions, where the computer instructions cause the computer to perform the methods provided by the above method embodiments, for example, the methods include: acquiring a network-energy storage characteristic matrix of a data center network according to structural parameters of the data center network; the network-energy storage characteristic matrix is used for describing connection and energy storage relation among servers of the data center network; performing spectral clustering with the characteristic value quantity of K on the network-energy storage characteristic matrix based on the preset area quantity K to obtain an area corresponding to each target device of the data center network, wherein K is an integer greater than 1; and performing area division according to the area corresponding to each target device, and dividing the data center network into K areas.
Those of ordinary skill in the art will understand that: the implementation of the above-described apparatus embodiments or method embodiments is merely illustrative, wherein the processor and the memory may or may not be physically separate components, i.e. may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as a usb disk, a removable hard disk, a ROM/RAM, a magnetic disk, an optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute the method according to the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims (9)

1. A method for dividing a data center network area is characterized by comprising the following steps:
acquiring a network-energy storage characteristic matrix of a data center network according to structural parameters of the data center network; the network-energy storage characteristic matrix is used for describing connection and energy storage relation among servers of the data center network;
performing spectral clustering with the characteristic value quantity of K on the network-energy storage characteristic matrix based on the preset area quantity K to obtain an area corresponding to each target device of the data center network, wherein K is an integer greater than 1;
dividing the data center network into K areas according to the area corresponding to each target device;
the acquiring a network-energy storage characteristic matrix of the data center network according to the structural parameters of the data center network specifically includes:
acquiring a network connection characteristic matrix according to connection bandwidth and connection time delay between servers of a data center network;
acquiring an energy storage connection characteristic matrix according to the capacity of energy storage equipment of a server of a data center network, the power transmission power among the energy storage equipment and the circuit transmission efficiency;
and acquiring a network-energy storage characteristic matrix of the data center network according to the network connection characteristic matrix and the energy storage connection characteristic matrix.
2. The method according to claim 1, wherein the obtaining a network connection feature matrix according to a connection bandwidth and a connection delay between servers of a data center network specifically includes:
for any two servers with connection relation in the data center network, acquiring the connection bandwidth between the two servers to obtain a network connection bandwidth matrix BN×N：
Figure FDA0002731902840000011
Wherein, bij,i＝1,...,N,j＝1,...,NIs the connection bandwidth between server i and server j;
for any two servers with connection relation in the data center network, acquiring the connection between the two serversTime delay to obtain a network connection time delay matrix TN×N：
Figure FDA0002731902840000021
Wherein, tij,i＝1,...,N,j＝1,...,NIs the connection delay between server i and server j;
according to the network connection bandwidth matrix BN×NAnd said network connection delay matrix TN×NObtaining a network connection characteristic matrix TBN×N：
TBN×N＝BN×N·/TN×N。
3. The method according to claim 1, wherein obtaining the energy storage connection feature matrix according to the energy storage device capacity of the server of the data center network, the power transmission power between the energy storage devices, and the circuit transmission efficiency specifically includes:
acquiring the capacity of energy storage equipment of each server of the data center network, and diagonalizing the capacity of the energy storage equipment to obtain a diagonal matrix PN×N＝diag(p1,p2…pN) Wherein p isi,i＝1,...,NCapacity of the energy storage device to which server i is connected;
for energy storage devices of two servers with any connection relation, acquiring power transmission power between the energy storage devices of the two servers to obtain a power transmission power matrix QN×N：
Figure FDA0002731902840000022
Wherein q isij,i＝1,...,N,j＝1,...,NTransmitting power for power between the energy storage device of the server i and the energy storage device of the service j;
for the energy storage devices of two servers with any connection relation, acquiring the electric power between the energy storage devices of the two serversTransmission efficiency, obtaining a power transmission efficiency matrix EN×N：
Figure FDA0002731902840000023
Wherein e isij,i＝1,...,N,j＝1,...,NThe power transmission efficiency between the energy storage device of the server i and the energy storage device of the service j;
according to the diagonal matrix PN×NThe power transmission power matrix QN×NAnd the power transmission efficiency matrix EN×NObtaining an energy storage connection characteristic matrix PQEN×N：
PQEN×N＝PN×N·QN×N·EN×N。
4. The method according to claim 1, wherein the obtaining a network-energy storage feature matrix of the data center network according to the network connection feature matrix and the energy storage connection feature matrix specifically includes:
multiplying the network connection characteristic matrix and the energy storage connection characteristic matrix to obtain a network-energy storage characteristic matrix C of the data center networkN×N：
CN×N＝TBN×NPQEN×N。
5. The method according to claim 1, wherein the dividing the data center network into K regions according to the region corresponding to each target device further comprises:
the management entity of each region of the data center network respectively manages and schedules the energy in each region, and the whole energy supply and demand balance is realized in a decentralized cooperation mode; and/or
The management entity of each region of the data center network respectively selects the matched services in the respective region and optimizes the resources in the region; and the scheduling and distribution of related services are realized among the areas in a negotiation or game mode.
6. The method according to claim 1 or 5, wherein the area division is performed according to the area corresponding to each target device, the data center network is divided into K areas, and then the method further comprises:
when a fault occurs, the management entity of the fault area performs fault detection on the jurisdiction area and sends state information to the adjacent area according to the fault detection condition so as to assist the fault location of the adjacent area.
7. A data center network area partitioning apparatus, comprising:
the network-energy storage characteristic module is used for acquiring a network-energy storage characteristic matrix of the data center network according to the structural parameters of the data center network; the network-energy storage characteristic matrix is used for describing connection and energy storage relation among servers of the data center network;
the spectral clustering module is used for performing spectral clustering with the characteristic value quantity of K on the network-energy storage characteristic matrix based on the preset number K of regions to obtain a region corresponding to each target device of the data center network, wherein K is an integer greater than 1; and
the area division module is used for carrying out area division according to the area corresponding to each target device and dividing the data center network into K areas;
the network-energy storage feature module is specifically configured to:
acquiring a network connection characteristic matrix according to connection bandwidth and connection time delay between servers of a data center network;
acquiring an energy storage connection characteristic matrix according to the capacity of energy storage equipment of a server of a data center network, the power transmission power among the energy storage equipment and the circuit transmission efficiency;
and acquiring a network-energy storage characteristic matrix of the data center network according to the network connection characteristic matrix and the energy storage connection characteristic matrix.
8. A data center network area partitioning apparatus, comprising:
at least one processor; and
at least one memory communicatively coupled to the processor, wherein:
the memory stores program instructions executable by the processor, the processor invoking the program instructions to perform the method of any of claims 1 to 6.
9. A non-transitory computer-readable storage medium storing computer instructions that cause a computer to perform the method of any one of claims 1 to 6.
CN201810321126.7A 2018-04-11 2018-04-11 Data center network area division method, device and equipment Active CN108632357B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810321126.7A CN108632357B (en) 2018-04-11 2018-04-11 Data center network area division method, device and equipment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810321126.7A CN108632357B (en) 2018-04-11 2018-04-11 Data center network area division method, device and equipment

Publications (2)

Publication Number Publication Date
CN108632357A CN108632357A (en) 2018-10-09
CN108632357B true CN108632357B (en) 2021-02-02

Family

ID=63704926

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810321126.7A Active CN108632357B (en) 2018-04-11 2018-04-11 Data center network area division method, device and equipment

Country Status (1)

Country Link
CN (1) CN108632357B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111628893B (en) * 2020-05-27 2022-07-12 北京星辰天合科技股份有限公司 Fault processing method and device of distributed storage system and electronic equipment

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103632203A (en) * 2013-09-23 2014-03-12 国家电网公司 Distribution network power supply area division method based on comprehensive evaluation
CN105356450A (en) * 2015-10-28 2016-02-24 国家电网公司西北分部 Power grid subarea division method based on dynamic electricity prices
CN105790279A (en) * 2016-04-28 2016-07-20 西华大学 Reactive voltage partitioning method based on spectral clustering
CN106203867A (en) * 2016-07-19 2016-12-07 国家电网公司 Grid division methods based on power distribution network assessment indicator system and cluster analysis
CN107492887A (en) * 2017-09-04 2017-12-19 清华大学 A kind of region partitioning method and system of wide area energy internet

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100470995C (en) * 2007-03-23 2009-03-18 清华大学 Distributed computing method of the features of the power system
US9300501B2 (en) * 2013-04-12 2016-03-29 Broadcom Corporation Spatial null creation using massive MIMO (M-MIMO)
CN106329520B (en) * 2016-09-21 2018-12-14 河海大学 A kind of UPFC modeling method based on PSASP

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103632203A (en) * 2013-09-23 2014-03-12 国家电网公司 Distribution network power supply area division method based on comprehensive evaluation
CN105356450A (en) * 2015-10-28 2016-02-24 国家电网公司西北分部 Power grid subarea division method based on dynamic electricity prices
CN105790279A (en) * 2016-04-28 2016-07-20 西华大学 Reactive voltage partitioning method based on spectral clustering
CN106203867A (en) * 2016-07-19 2016-12-07 国家电网公司 Grid division methods based on power distribution network assessment indicator system and cluster analysis
CN107492887A (en) * 2017-09-04 2017-12-19 清华大学 A kind of region partitioning method and system of wide area energy internet

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
基于遗传算法的地区电网停电恢复;盛四清;《电力系统自动化》;20010430;全文 *
开放架构下能源胡亮网新能源接入功率控制方法研究;任光，曹军威，胡紫巍;《南方电网技术》;20161225;全文 *

Also Published As

Publication number Publication date
CN108632357A (en) 2018-10-09

Similar Documents

Publication Publication Date Title
Pires et al. A virtual machine placement taxonomy
CN105912397A (en) Resources management method and device
CN104503826B (en) The virtual machine mapping method and device of a kind of cloud computation data center
CN110661842B (en) Resource scheduling management method, electronic equipment and storage medium
Cerroni Multiple virtual machine live migration in federated cloud systems
CN108304256B (en) Task scheduling method and device with low overhead in edge computing
Song et al. Gaia scheduler: A kubernetes-based scheduler framework
CN113098711A (en) Power distribution Internet of things CPS (control system) management and control method and system based on cloud edge cooperation
CN110519776A (en) Balanced cluster and federated resource distribution method in a kind of mist computing system
WO2023098374A1 (en) Network resource deployment method and apparatus, and electronic device and storage medium
CN113645146B (en) New stream density-based software defined network controller load balancing method and system
CN110430068A (en) A kind of Feature Engineering method of combination and device
EP3979563B1 (en) Inter-domain data interaction method and apparatus
Ruan et al. Resource allocation and distributed uplink offloading mechanism in fog environment
CN109614190A (en) Resources configuration management method and platform
Saxena et al. Communication cost aware resource efficient load balancing (care-lb) framework for cloud datacenter
CN108632357B (en) Data center network area division method, device and equipment
CN111767139A (en) Cross-region multi-data-center resource cloud service modeling method and system
Chen et al. Blockchain-based optical network slice rental approach for IoT
CN106874108A (en) Thin cloud is minimized in mobile cloud computing use number technology
Belgacem et al. New virtual machine placement approach based on the micro genetic algorithm in cloud computing
CN110908800B (en) Scheduling optimization method based on scheduling entropy in edge cloud collaborative environment
Swarnakar et al. A novel improved hybrid model for load balancing in cloud environment
Yi et al. Hfedmtl: Hierarchical federated multi-task learning
CN104866611B (en) A kind of practical method for fast searching of power system Active Splitting section

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