WO2016095162A1 - Dispositif et procédé pour déterminer une opération pour régler un nombre de machines virtuelles - Google Patents

Dispositif et procédé pour déterminer une opération pour régler un nombre de machines virtuelles Download PDF

Info

Publication number
WO2016095162A1
WO2016095162A1 PCT/CN2014/094226 CN2014094226W WO2016095162A1 WO 2016095162 A1 WO2016095162 A1 WO 2016095162A1 CN 2014094226 W CN2014094226 W CN 2014094226W WO 2016095162 A1 WO2016095162 A1 WO 2016095162A1
Authority
WO
WIPO (PCT)
Prior art keywords
vms
load
indicator
determining
performance
Prior art date
Application number
PCT/CN2014/094226
Other languages
English (en)
Chinese (zh)
Inventor
唐朋成
Original Assignee
华为技术有限公司
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 华为技术有限公司 filed Critical 华为技术有限公司
Priority to PCT/CN2014/094226 priority Critical patent/WO2016095162A1/fr
Priority to CN201480028886.9A priority patent/CN106170767B/zh
Publication of WO2016095162A1 publication Critical patent/WO2016095162A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines

Definitions

  • the present invention relates to cloud computing technologies, and in particular, to an apparatus and method for determining a virtual machine number adjustment operation.
  • Cloud Computing technology integrates computing, storage, service components, network software/hardware resources distributed on the network, and provides users with convenient and fast services based on resource virtualization.
  • the functions of network elements in the network can be implemented by one or more virtual machines.
  • EPC network elements in the EPC network such as: Mobility Management Entity (MME), Packet Data Network Gateway (Packet)
  • MME Mobility Management Entity
  • Packet Packet Data Network Gateway
  • PGW Packet Data Network Gateway
  • SGW Serving GateWay
  • HSS Home Subscriber Server
  • VNFs virtualized network functions
  • the number of VMs in the VNF can be adjusted as the business changes. This automatic adjustment process is called “automatic capacity expansion/reduction:” Users can set the corresponding application programming interface (API) provided by the vendor. The VM number adjustment program.
  • API application programming interface
  • FIG. 1 shows an existing VM number adjustment scheme.
  • the central processing unit (CPU) utilization is used to measure the load of the system, and the number of VMs is adjusted according to the system load.
  • an action corresponding to the threshold is performed, for example, adding a VM, and the length is a period of the expansion cooling time, regardless of whether the load exceeds the upper expansion threshold.
  • the action corresponding to the threshold is performed, For example, if one VM is reduced, and then the length is the condensed cooldown time, the VM can no longer be reduced regardless of whether the system load exceeds the lower threshold.
  • the number of VMs that can be increased or decreased can only be fixed if the capacity is increased or reduced. In some scenarios, for example, in the case of sudden service, the number of VMs can only be fixed. The number of VMs is slow to expand.
  • An embodiment of the present invention provides an apparatus and method for determining a VM number adjustment operation, to provide a flexible configuration scheme of a VM.
  • an embodiment of the present invention provides an apparatus for determining a virtual machine VM number adjustment operation, including:
  • a data acquisition module configured to acquire a load indicator of a current time system and a number of VMs used by the system, where the system includes one or more VMs;
  • a first determining module configured to determine, according to a load indicator of the system at the current time and a number of VMs used by the system, a first operation to be performed on the system
  • the first operation includes: adding m VMs, reducing n VMs, or keeping the number of VMs unchanged, where m, n are positive integers, and m, n are systems of the first decision module according to current time
  • the load indicator and the number of VMs used by the system are determined.
  • the first determining module is specifically configured to:
  • the performance indicator of the system after performing the first operation is predicted by the first decision module according to the load indicator of the system at the current time, and the number of VMs used by the system after performing the first operation;
  • the number of VMs used by the system after performing the first operation is predicted by the first decision module according to the number of VMs used by the system at the current time and the determined first operation.
  • the data collection module is further configured to: before the first determining module determines the first operation, obtain the current time Performance indicators of the system;
  • the first decision module is specifically configured to:
  • the last time the system is in the current state the cumulative reward value obtained by performing the first operation on the system, the immediate reward value obtained by performing the first operation on the system at the current time, and the last time
  • the system is in a next state, performing a maximum of the accumulated reward values obtained by each of the first set of executable first operations for the system, wherein the next state is The state in which the system is located after performing the first operation;
  • the reward value is determined by the first decision module according to the number of VMs used by the system, and the performance index of the system under the load indicated by the load indicator of the system at the current time, the system The fewer the number of VMs used, the higher the performance metric of the system indicates the higher the performance of the system, the greater the reward value.
  • the first determining module is specifically configured to:
  • the largest cumulative return value is selected.
  • the device further includes: a VM quantity adjustment mode selection module, configured to select a mode for adjusting the number of VMs;
  • the first decision module is specifically configured to:
  • the method for adjusting the number of VMs selected by the VM quantity adjustment mode selection module is: adjusting the number of VMs used by the system according to the load indicator of the system at the current time and the number of VMs used by the system,
  • the first operation to be performed on the system is determined based on the load indicator of the system at the current time and the number of VMs used by the system.
  • the apparatus further includes a second determining module, configured to:
  • the method for adjusting the number of VMs selected by the VM quantity adjustment mode selection module is: determining whether the number of VMs needs to be adjusted according to the upper limit threshold of the load index of the system and the lower limit threshold of the load indicator of the system,
  • the second operation includes: adding p VMs, reducing q VMs, or keeping the number of VMs unchanged, the p being a preset upward adjustment step, and the q being a preset downward adjustment step , p, q are positive integers.
  • the data collection module is further configured to: acquire a load indicator of the system in each of the T historical moments before the current time, where T is a positive integer;
  • the load indicator upper limit threshold and the load indicator lower limit threshold are determined by the second decision module according to the following conditions according to the load index of the system in each of the T historical moments before the current time:
  • the performance indicators of the system meet the preset performance index requirements, and the number of VMs used is the least.
  • the second determining module is specifically configured to determine the load indicator upper limit threshold and the load indicator lower threshold according to the following steps:
  • the input return ratio is determined by performing the following steps:
  • predicting that the system performs the second operation after performing the second operation on the system according to the combination of the candidate load index upper threshold and the lower limit of the load indicator The number of VMs; and predicting performance indicators of the system after performing the second operation according to the load indicator of the system at the historical moment and the predicted number of VMs used by the system after performing the second operation;
  • an embodiment of the present invention provides a method for determining a virtual machine VM number adjustment operation, including:
  • the first operation includes: adding m VMs, reducing n VMs, or keeping the number of VMs unchanged, where m and n are positive integers, and m and n are load indicators of the system according to the current time and the The number of VMs used by the system is determined.
  • determining a first operation to be performed on the system includes:
  • the performance indicator of the system after performing the first operation is predicted according to the load indicator of the system at the current time, and the number of VMs used by the system after performing the first operation;
  • the number of VMs used by the system after an operation is predicted based on the number of VMs used by the system at the current time, and the determined first operation.
  • the method before determining the first operation to be performed on the system, the method further includes: acquiring a performance indicator of the system at the current moment;
  • Determining the first action to be performed on the system including:
  • the cumulative reward value obtained by the system in the current state for performing the first operation on the system, the immediate reward value obtained by performing the first operation on the system at the current time, and the last time The system is in the next state, performing the executable first on the system a maximum of the cumulative reward values obtained by each of the first operations in the set of operations, wherein the next state is a state in which the system is after performing the first operation;
  • the reward value is determined according to the number of VMs used by the system, and the performance index of the system under the load indicated by the load indicator of the system at the current time, and the number of VMs used by the system is increased. Less, the performance metric of the system indicates that the higher the performance of the system, the greater the reward value.
  • determining an immediate reward value obtained by performing the first operation on the system at a current moment includes:
  • determining, in the next state in which the system is located after performing the first operation, performing the The maximum of the cumulative reward values obtained for each of the first operations in the set of first operations performed including:
  • the largest cumulative return value is selected.
  • the method further includes: selecting a VM quantity adjustment The way;
  • Determining the first action to be performed on the system including:
  • the number of selected VMs is adjusted, the number of VMs used by the system is adjusted according to the load indicator of the system and the number of VMs used by the system.
  • the sixth possible implementation manner if the selected number of VMs is adjusted, according to the upper limit threshold of the load index of the system and the lower threshold of the load indicator of the system, To determine whether the number of VMs needs to be adjusted, after selecting the method of adjusting the number of VMs, it also includes:
  • the operation includes: adding p VMs, reducing q VMs, or keeping the number of VMs unchanged, the p being a preset upward adjustment step, the q being a preset downward adjustment step, p, q being A positive integer.
  • the load indicator upper limit threshold and the load indicator lower limit threshold are based on T historical moments before the acquired current time
  • the load indicators of the system described in each historical moment are determined according to the following conditions:
  • the performance indicator of the system meets the preset performance index requirement, and the number of VMs used is the least, where T is a positive integer.
  • the load indicator upper limit threshold and the load indicator lower threshold are specifically determined according to the following steps:
  • the input return ratio is determined by performing the following steps:
  • the load indicator of the system is predicted, and the prediction is based on the negative of the candidate.
  • the number of VMs is adjusted according to the load index of the system and the number of VMs used by the system.
  • the number of VMs is adjusted according to the load index of the system and the number of VMs used by the system.
  • the number of VMs can be accurately determined. The value enables the number of VM adjustments to be determined in real time based on the load change of the system and the number of VMs currently used, improving the real-time performance of the VM adjustment.
  • the performance index of the system meets the preset performance index requirement, and the number of VMs used by the system is the least. It can balance the performance of the system with the number of VMs of the system, and improve the VM utilization while ensuring the performance requirements of the system.
  • the method of reinforcement learning is adopted, and when the value of the adjustment is determined, the accumulated return value of the historical time before the current time is considered, and The possible return value at the next moment is also considered, so that it is no longer necessary to set the cooling time in the method shown in FIG. 1, thereby improving the real-time performance of the VM number adjustment.
  • unnecessary expansion and shrinkage operations can be reduced, and the ping-pong effect is slowed down.
  • the upper limit of the load index may be adjusted according to the historical load of the system. Threshold and load indicator lower threshold, flexible setting of VM number adjustment threshold according to system load, evaluation of different thresholds to select optimal threshold, improve system VM utilization, and ensure system performance index requirements.
  • the load index upper limit threshold and the load index lower limit threshold are adjusted according to the historical load of the system, after the VM number is adjusted, the performance index of the system meets the preset performance index requirement, and the number of VMs used is the least, and the system can be considered.
  • the performance metrics and the number of VMs required for the system are considered.
  • the load index upper limit threshold and the load index lower limit threshold are adjusted according to the historical load of the system, the available thresholds are respectively evaluated for different system load conditions, and the optimal threshold is selected, so that the VM used in the adjustment system is adjusted. When the number is adjusted, the optimal threshold is selected to improve the utilization efficiency of the VM.
  • FIG. 1 is a schematic diagram of a virtual machine quantity adjustment scheme
  • FIG. 2 is a schematic structural diagram of an apparatus for determining a VM number adjustment operation according to an embodiment of the present invention
  • FIG. 3 is a schematic structural diagram of a second apparatus for determining a VM number adjustment operation according to an embodiment of the present disclosure
  • FIG. 4 is a schematic structural diagram of an apparatus for determining a VM number adjustment operation according to an embodiment of the present invention
  • Figure 5 is a schematic diagram of a method of reinforcement learning
  • FIG. 6 is a flowchart of Embodiment 1 of the present invention.
  • Figure 7 is a schematic diagram of a cumulative return table
  • FIG. 8 is a schematic diagram of a process of updating a cumulative return table according to Embodiment 1 of the present invention.
  • Embodiment 9 is a schematic diagram of a data fitting process in Embodiment 2 of the present invention.
  • FIG. 10 is a flowchart of Embodiment 3 of the present invention.
  • FIG. 11 and FIG. 12 are diagrams showing the apparatus for adjusting the number of VMs and the VNF in the fourth embodiment of the present invention. Schematic diagram
  • FIG. 13 and FIG. 14 are schematic diagrams of service scenarios applicable to an embodiment of the present invention in Embodiment 5 of the present invention.
  • FIG. 15 is a schematic structural diagram of a fourth apparatus for determining a VM number adjustment operation according to an embodiment of the present disclosure
  • FIG. 16 is a flowchart of a method for determining a VM number adjustment operation according to an embodiment of the present invention.
  • An embodiment of the present invention provides an apparatus and method for determining a VM number adjustment operation, to provide a flexible configuration scheme of a VM.
  • the data collection module acquires the load indicator of the current time system and the number of VMs used by the system, wherein the system includes one or more VMs; the first decision module The data collection module determines a first operation to be performed on the system according to the acquired load indicator of the current time system and the number of VMs used by the system; wherein the first operation includes: adding m VMs, reducing n VMs, or maintaining VMs The quantity is unchanged, where m and n are positive integers, and m and n are determined by the first decision module according to the load index of the current time system and the number of VMs used by the system.
  • the number of VMs that are increased or decreased can be determined according to the load index of the current time system and the number of VMs used by the system, and is no longer a fixed setting value, and is more flexible and adaptable.
  • the load condition of the system and the usage of the VM are used to flexibly determine whether to increase or decrease the number of VMs to adapt to different scenarios.
  • the system includes one or more VMs implemented by the one or more VMs.
  • the system can be any VNF implemented by the VM, such as various network elements in the foregoing network: MME, PGW, HSS, and the like.
  • the device for determining the VM number adjustment operation provided by the embodiment of the present invention may be located outside the system or may be located in the system. For details, refer to Embodiment 4.
  • the device for determining the number of VMs is determined by the embodiment of the present invention, and is used to determine the operation of adjusting the number of VMs, and sends the result of the determination to the device for performing the VM number adjustment, and the device for performing the VM number adjustment is provided according to the embodiment of the present invention.
  • the determination result sent by the device that determines the VM number adjustment performs an operation of adjusting the number of VMs to the system.
  • the device performing the VM number adjustment may be the VNF manager and the virtualization infrastructure in the Network Funcitons Virtualization-Management and Orchestration (NFV-MANO) in FIG.
  • the Virtualised Infrastructure Manager (VIM) is configured by the two devices in conjunction with the number of VMs executing the system VNF.
  • VMM Virtualised Infrastructure Manager
  • the method for adjusting the number of VMs disclosed in the above-mentioned standard documents is only an optional implementation manner, and the embodiments of the present invention are directed to providing an apparatus and method for determining the number of VMs to adjust the number of VMs.
  • the decision of the operation does not limit the execution of the VM number adjustment operation. It can be considered that the decision scheme provided by the embodiment of the present invention can be applied to any execution scheme of the VM number adjustment including the manner disclosed by the above standard documents.
  • the system's load indicators may include, but are not limited to, the various service load indicators and various resource load indicators listed in Table 1 below.
  • System performance indicators may include, but are not limited to, the indicators listed in Table 2:
  • Performance Business establishment success rate Service establishment delay Packet transmission delay shake Packet loss rate Block error rate Bit error rate
  • the device for determining the VM number adjustment operation has the following three types:
  • the device 10 shown in Figure 2 includes:
  • the first decision module 202 determines the first operation to be performed on the system according to the load indicator of the current time system acquired by the data collection module 201 and the number of VMs used by the system;
  • the first operation includes:
  • the above m and n are determined by the first decision module 202 according to the load index of the current time system and the number of VMs used by the system.
  • the device 11 shown in FIG. 3 includes:
  • the implementation of the data collection module 201 and the first decision module 202 may refer to the apparatus 10 for determining the VM number adjustment operation.
  • the VM quantity adjustment mode selection module 204 is configured to select the VM quantity adjustment mode, that is, the first
  • the decision module 202 determines whether the decision is still made by the second decision module 203.
  • the second decision module 203 is specifically configured to:
  • the second operation includes: adding p VMs, reducing q VMs, or keeping the number of VMs unchanged; p is a preset upward adjustment step, q is a preset downward adjustment step, and p and q are A positive integer.
  • the difference from the first operation includes: the number of VMs in the second operation is preset.
  • the number of VMs is adjusted according to the load index of the current time system acquired by the data collection module 201 and the system used. The number of VMs is determined.
  • the device 12 shown in FIG. 4 includes: a data acquisition module 201 and a second decision module 203.
  • the implementation of the data acquisition module 201 and the second decision module 203 can be adjusted by referring to the second determination of the number of VMs as shown in FIG. Operated device 11.
  • FIG. 2 is a schematic structural diagram of an apparatus 10 for determining a VM number adjustment operation according to an embodiment of the present invention. As shown in FIG. 2, the device 10 includes:
  • the data collection module 201 is configured to acquire a load indicator of the current time system and a quantity of VMs used by the system, where the system includes one or more VMs;
  • the first decision module 202 is configured to determine, according to the load indicator of the current time system acquired by the data collection module 201 and the number of VMs used by the system, the first operation to be performed on the system;
  • the first operation includes: adding m VMs, reducing n VMs, or keeping the number of VMs unchanged, where m and n are positive integers, and m and n are load indexes of the first decision module 202 according to the acquired current time system. And the number of VMs used by the system is determined.
  • the data collection module 201 can acquire the load indicator of the system and the number of VMs used by the system in a preset period. For example: 1 minute or 5 minutes.
  • the data collection module 201 can collect one of the various indicators listed in Table 1 as a load indicator of the system, such as: CPU utilization.
  • various indicators listed in Table 1 may be collected, and various indicators are weighted and normalized, and the system load indicators are comprehensively calculated.
  • max_User_num is the maximum number of users of the system
  • ⁇ is a weighting factor, 0 ⁇ 1, 0 ⁇ Disk_occupation_rate ⁇ 1.
  • the first decision module 202 may determine a plurality of options for performing the first operation to be performed on the system according to the obtained load index of the system at the current time and the number of VMs used by the system, and several of them are illustrated here. . Regardless of the manner in which it is used, the purpose of flexibly setting the number of adjusted VMs can be achieved as compared with the method shown in FIG.
  • Table 3 The correspondence table between the load index of the pre-stored system and the number of VMs used by the system is shown in Table 3. This table is used to indicate the number of VMs that the system should use if the system performance is guaranteed under the system load indicated by the system's load indicator.
  • Table 3 the correspondence between the system load and the number of VMs that the system should use.
  • the performance index of the system can be measured by a specific performance indicator in Table 2, such as: service establishment delay, or various performance indicators, in a similar manner to formula 1. Perform a weighted summation to calculate.
  • the performance metrics of the system can be compared with preset performance metric thresholds to determine whether the performance metrics of the system are guaranteed. For example, if the performance indicator of the system is the packet loss rate, it is determined that the packet loss rate is less than the preset packet loss rate threshold, which ensures the performance index of the system. For example, if the performance index of the system is the success rate of the service establishment, the performance of the system is ensured when the service establishment success rate is greater than the preset service establishment success rate threshold.
  • the first decision module 202 checks the table 3, determines the number of VMs that the system corresponding to the load indicator of the current time system of the acquired system, and the number of VMs used by the system at the current time obtained by the root data collection module 201, and determines the first to be executed on the system. Operation, that is, increase or decrease the number of VMs, or keep the number of VMs unchanged, and if the number of VMs is increased or decreased, the specific value of the number of VMs is increased or decreased, so that after the first operation is performed on the system, the system uses The number of VMs is equal to the number of VMs that the system should use for the current load indicator of the system.
  • the first decision module 202 obtains the current load index of the system as 25%, and looks up Table 3 to determine that the number of VMs that the system should use is three. If the number of VMs used by the system at the current time is five, Be sure to reduce 2 VMs.
  • the first decision module 202 obtains the current load index of the system at 75%, and looks up Table 3 to determine that the number of VMs that the system should use is eight, and if the current time is obtained, the number of VMs used by the system is four. Then determine to add 4 VMs.
  • the data collection module 201 can also obtain the load indicator of the system at the previous time.
  • the first decision module 202 determines the change of the load indicator of the system according to the current time acquired by the data collection module 201 and the load index of the system at the previous time. Trends, if it is determined that the system's load index is incremented, the number of VMs used by the adjusted system can be larger than the number of VMs that should be used by the system in Table 3, and the number of system buffer VMs is preset.
  • the data collection module 201 obtains the current time, and the load indicators of the system in the first four moments are: 20%, 30%, 35%, 40%, and 45%, respectively, and the number of VMs used by the system at the current time is 2, On the basis of adding 3 VMs, taking into account the increasing trend of the system's load indicators, and then increasing the preset system buffer VM number (assumed to be 1), a total of 4 VMs are added, and the number of VMs used by the system is adjusted. Is 6.
  • This method can be applied to the scenario of sudden business, and can rapidly increase or decrease the number of VMs to meet the load of the system.
  • the first decision module 202 determines a first operation to be performed on the system such that the performance metric of the system conforms to the preset performance indicator requirement after the first operation is performed, and the number of VMs used is the least;
  • the performance indicator of the system after performing the first operation is predicted by the first decision module 202 according to the load indicator of the current time system and the number of VMs used by the system after the first operation is performed; after the first operation is performed, the system
  • the number of VMs used is predicted by the first decision module 202 based on the number of VMs used by the system at the current time, and the determined first operation to be performed on the system.
  • the preset performance indicator requirement may be a preset Service Level Agreement (SLA) requirement.
  • SLA refers to the agreement between the operator and the customer to ensure the performance and reliability of the service at a certain cost. Often this cost is a major factor driving the quality of service provided by operators.
  • three performance indicators are specified in the SLA: service processing delay, packet loss rate of service packets, and bandwidth for processing services.
  • the benchmark values of these three performance indicators are 100 milliseconds delay, 1% packet loss rate, and 100 Mbps bandwidth.
  • the service processing delay is less than 100 ms and the packet loss rate is less than 1%, the bandwidth is greater than 100 Mbps.
  • the penalty method can be based on the time setting of the violation of the SLA.
  • the penalty for violating the service processing delay per unit time is p_1
  • the penalty for the violation of the packet loss rate is p_2
  • the penalty for violating the bandwidth is p_3.
  • the data collection module 201 is further configured to: before the first decision module 202 determines that the first operation to be performed on the system, obtain a performance indicator of the current time system;
  • the first decision module 202 can employ a method of Reinforcement Learning to determine the first operation to be performed on the system.
  • the main idea of reinforcement learning is to learn and choose the best action that can achieve the goal. 5, when the body of each operation to make a i in its environment, return the value r i, the main return from the learned value obtained in a subsequent operation to produce the maximum value of the cumulative reward.
  • the return value is divided into the cumulative return value and the immediate return value.
  • the accumulated return value is the cumulative result of the return value at some time before the current time, for example, the result of the summation;
  • the immediate return value is the return value obtained after the first operation a i is performed at the current time.
  • the main body is the first decision module 202, and the environment is the above “system”, and the reward value may be determined by the number of VMs used by the system and the performance index of the system under the load indicated by the current load indicator of the system.
  • the immediate return value r can be defined by the following formula 2:
  • Performance is the performance index of the system, which can be a specific performance indicator in Table 2.
  • Standards such as: business establishment success rate, reciprocal of business establishment delay; can also integrate various performance indicators, using weighted summation in a similar way to formula 1, to calculate: the weighted results meet: weighted summation performance indicators
  • the performance index after summation is larger; the larger the performance index such as the service establishment success rate, the larger the performance index after the weighted summation.
  • Vm_number is the number of VMs used by the system.
  • the first decision module 202 determines an immediate reward value obtained by performing the first operation by:
  • the instantaneous reward value obtained by performing the first operation on the system is determined according to the predicted number of VMs used by the system after the first operation and performance indicators of the system.
  • the first determining module 202 may determine, according to the manner, performing the cumulative operation of each of the first operations in the executable first operation set in the next state in which the system is located after the performing the first operation
  • the largest cumulative return value is selected.
  • FIG. 3 is a schematic structural diagram of a second apparatus 11 for determining a VM number adjustment operation according to an embodiment of the present invention. As shown in FIG. 3, the device 11 includes:
  • the implementation of the data collection module 201 and the first decision module 202 can refer to the first type above.
  • the implementation of the corresponding module in the device 10 for determining the VM number adjustment operation; the VM number adjustment mode selection module 204 is configured to select the VM number adjustment mode, that is, whether the decision is made by the first decision module 202 or the second decision module 203.
  • the second decision module 203 is specifically configured to:
  • the second operation includes: adding p VMs, reducing q VMs, or keeping the number of VMs unchanged; p is a preset upward adjustment step, q is a preset downward adjustment step, and p and q are A positive integer.
  • the difference from the first operation includes: the number of VMs in the second operation is preset.
  • the number of VMs is adjusted according to the load index of the current time system acquired by the data collection module 201 and the system used. The number of VMs is determined.
  • the scheme similar to the VM number adjustment shown in FIG. 1 is that the load index of the current time system is compared with a preset threshold, and whether the number of VMs is adjusted is determined according to the comparison result.
  • the threshold is determined according to the load index of the historical time system.
  • the data collection module 201 is further configured to: acquire a load indicator of each of the T historical moments before the current time, where T is a positive integer;
  • the load index upper limit threshold and the load index lower limit threshold are determined by the second decision module 203 according to the load index of each historical time system in the T historical moments before the current time acquired by the data collection module 201, and satisfy:
  • the performance index of the system after performing the second operation meets the preset performance index requirement, and the number of VMs used is the least, where T is a positive integer.
  • the preset performance index requirements may also be the requirements of the aforementioned SLA.
  • the second determining module 203 is specifically configured to determine a load indicator upper limit threshold and a load indicator lower limit threshold according to the following steps:
  • the input return ratio is determined by performing the following steps:
  • the load indicator of each historical time system acquired by the data collection module 201 predict the number of VMs used by the system after performing the second operation on the system according to the combination of the candidate load index upper threshold and the lower limit of the load index;
  • the load indicator of the historical time system and the predicted number of VMs used after the second operation of the system, and the performance index of the system after the second operation is predicted by the predictor;
  • the input return ratio is determined according to the number of VMs used by the system after performing the second operation and the performance index corresponding to each of the performance indicators, wherein the number of performance indicators that meet the preset performance index requirements of each performance indicator is larger. The smaller the total number of VMs used by the system at each historical moment, the higher the return on investment ratio;
  • the third embodiment is an example of a method for the second decision module 203 to determine the lower limit threshold of the load index and the upper limit threshold of the load indicator.
  • FIG. 4 is a schematic structural diagram of a third apparatus 12 for determining a VM number adjustment operation according to an embodiment of the present invention. As shown in Figure 4, the device 12 includes:
  • the data acquisition module 201 and the second decision module 203 wherein the implementation of the data acquisition module 201 and the second decision module 203 can refer to the implementation of the corresponding module in the device 11 for determining the VM number adjustment operation as shown in FIG.
  • Embodiment 1 is an example of a method for the first decision module 202 to determine the VM number operation
  • the second embodiment is an example of a method for fitting the data analysis module 205;
  • the third embodiment is an example of a method for the second decision module 203 to determine the VM number adjustment operation
  • the fourth embodiment describes the relationship between the device for determining the VM number adjustment operation and the VNF provided by the embodiment of the present invention
  • the fifth embodiment of the present invention provides a service scenario to which the embodiment of the present invention is applied.
  • Embodiment 1 describes a method by which the first decision module 202 adjusts the number of VMs.
  • the apparatus for determining the VM number adjustment operation may further include a data analysis module 205, configured to acquire, from the data collection module 201, a load indicator, a performance indicator, and a system used by the system at a historical time.
  • the number of VMs based on the obtained data, fits a function of the system's performance metrics to the system's load metrics and the number of VMs used by the system for use by the first decision module 202 in calculating the immediate return value (see steps). S604).
  • the possible state s i of the system constitutes a state set: ⁇ s i
  • i 1..M ⁇ , and in each state, there are N kinds of executable actions for the system, that is, the first operation There are N possible choices, N kinds of executable actions constitute an Action Set: ⁇ a i
  • i 1..N ⁇ , M, N are positive integers.
  • the immediate return value of the jth action a j corresponding to the i-th state s i is r i,j , and the accumulated return value is q i,j ; the load index of the system, the number of VMs used by the system, and the performance of the system
  • Equation 3 The functional relationship between the indicators is shown in Equation 3:
  • performance is the performance index of the system
  • workload is the load indicator of the system
  • vm_number is the number of VMs used by the system.
  • Equation 4 The immediate return value is shown in Equation 4 below:
  • the Q table in Fig. 6 is the cumulative return value table, referred to as the "cumulative return table", as shown in FIG.
  • the method of Embodiment 1 specifically includes the following steps:
  • S601 Initialize the state set, the action set, and the cumulative return table.
  • the state set contains all possible states of the system;
  • the action set is a set of first operations that the system can perform;
  • the cumulative return table is used to store the accumulated reward values of each state corresponding to each action;
  • the first decision module 202 obtains the load indicator, the performance indicator, and the number of VMs used by the system at the current time t from the data collection module 201.
  • the first decision module 202 determines to perform state matching, and determines the state s t of the current time system according to the load index, the performance index, and the number of VMs used by the system at the current time t obtained in step S602;
  • the state set includes a limited number of M states, and the M states can be defined by dividing the load indicator value of the system into A intervals, and dividing the value of the system performance index into B intervals.
  • Step S603 when determining the current state of the system, can determine the load index, the performance index of the system, and the number of VMs used by the system respectively, and then find the range from the state in which the three belong respectively.
  • the state s t at which the system is currently located referred to as the "current state”.
  • the load index workload current time t of the system performing an action after a j new performance metrics performance systems ', and performing the operation after a j System number of the VM vm_number', determines the execution state after the operation a j system is located , recorded as s t+k ;
  • the acquisition state s t + returns all accumulated value of k (i.e., in a state s t + k, each movement a 1, a 2, ..., a N corresponding accumulation
  • q t+k,j is the cumulative return value obtained by executing the action a j when the system is in the state s t+k last time
  • the accumulated reward values are updated by steps S604 and S605.
  • the difference between the accumulated return value and the immediate return value is that the accumulated return value not only considers the immediate return value of the current moment, but also considers the accumulation of all the return values of the state during the historical running process, and possible The maximum cumulative return value for a state. That is, the cumulative return value not only considers the impact of the historical state on the current state, but also the impact of the possible next state on the current state.
  • the above steps S604 to S605 can also refer to the schematic diagram shown in FIG. 8.
  • For the action a j in the action set calculate the performance index performance' of the system after executing the action, calculate the immediate return value r t,j according to the number of new VMs vm_number' of the system after performing the action, and performance';
  • the immediate return value r t,j the maximum cumulative return value q t+k,l corresponding to each action in the Q table in the next state S t+k after the action is performed, and the current state in the Q table of the current state t
  • the accumulated return value q t, j obtained by performing the action a j is calculated, and the updated q′ t,j is calculated , and the accumulated return value of the execution action a j corresponding to the state t in the Q table is updated with the value.
  • S607 Determine the cooling time. During the cooling time, the system no longer increases or decreases the number of VMs.
  • Step S607 is an optional step of ensuring the reliability of the system by the cooling time.
  • the cooling time may include: expansion cooling time, shrinkage cooling time, and anti-oscillation cooling time.
  • the capacity expansion cooling time is the interval between the two expansion operations (that is, the number of VMs is increased). Each time the capacity expansion/reduction capacity is determined, the expansion cooling time timer is decremented by 1. If the timer value is greater than zero, The new capacity expansion action (that is, the current number of VMs is not changed) is allowed. If the timer is less than or equal to zero, the capacity expansion decision is allowed. Once the decision to implement the capacity expansion is made, the expansion cooling time timer is reset.
  • the implementation of the reduction cooling time is similar to the expansion cooling time, but only for the shrinking action.
  • Anti-oscillation cooling time refers to the time between the decision to expand the capacity and the time to allow the refinement decision to be made again. After each expansion decision, the timer will be reset. After each expansion/reduction decision, the timer will be decremented by 1. If the timer value is greater than zero, even if the decision is reduced, it is not allowed. This action is performed (ie, the number of current VMs is kept constant), and if the timer value is less than or equal to zero, the shrinkage decision is allowed to be performed.
  • Step S607 is an optional step. If step S607 is not performed, after step S606 is performed, step S608 is directly executed: adjusting the number of VMs used by the system.
  • the number of VMs may be dynamically changed according to the load indicator of the current time system and the number of VMs used by the system at the current time.
  • the performance indicators of the system may be consistent with the preset performance. Under the requirements of the index, the number of VMs used by the system is the least, which improves the utilization efficiency of the VM.
  • the number of VM adjustments can be dynamically changed.
  • the number of VMs that are expanded or contracted corresponding to the upper expansion threshold or the lower expansion threshold is determined, and cannot be adjusted in real time according to the load index of the system.
  • the reward value of adding or deleting different VM numbers is quantitatively evaluated, and the action with the largest return value is selected as the action to be expanded or contracted to the system.
  • the first decision module 202 determines the first operation, it can be determined not only whether to increase, decrease or maintain the number of VMs, but also accurately determine the adjusted value of the number of VMs, and realize the determination according to the load index of the system.
  • the value of the VM quantity adjustment when determining the adjusted value, considers the cumulative return value of the historical moment before the current time, and also considers the possible return value at the next moment, so that it is no longer necessary to be in the method shown in FIG. , set the cooling time, which improves the real-time performance of the VM quantity adjustment.
  • the number of VM adjustments can be adapted to different system loads and system resource occupations.
  • the second embodiment introduces a method for data fitting by the data analysis module 205.
  • the method can be applied to any device for determining the VM number adjustment operation provided by the embodiment of the present invention, and is used for fitting a functional relationship between the performance index of the system and the load index of the system and the number of VMs used by the system.
  • the data collection module 201 collects the load index of the system, the number of VMs used by the system, and the performance index of the system in the historical time period before the current time.
  • the data acquisition module The block 201 may store the collected data in a history database, which may be located in the data collection module 201 or independent of the data collection module 201.
  • the historical time period may be several days in the past, several weeks, several months, or other longer or shorter time periods, the length of which may be determined according to the actual needs of the system operation.
  • the data collection module 201 may periodically collect the foregoing data, where the data may be an instantaneous value of a fixed average value, a maximum value, or a sampling time within a sampling period.
  • the data collection module 201 may initiate a data query request to the monitoring module for monitoring the foregoing data. After receiving the query request, the monitoring module sends the requested data to the data collection module 201; or, the monitoring module cycle The data is actively reported to the data collection module 201.
  • the data analysis module 205 extracts the load index, the performance index, and the number of VMs used by the system in a historical time period from the historical database, and uses the fitting algorithm to calculate the performance index performance and the load indicator workload and the number of VMs used by the system.
  • machine learning algorithms such as linear fitting, polynomial fitting, etc. may be employed.
  • the relationship between the performance indicator and the load indicator and the number of VMs used by the system is obtained according to the collected historical time data, so that the first decision module 202 and the second decision module 203 can be provided according to the second embodiment.
  • the functional relationship predicts the return value of different operations and selects the operation with the highest return. According to the requirements of different precisions and complexity, different fitting algorithms can be selected in this embodiment during the implementation.
  • the second decision module 203 determines the upper limit threshold of the load index and the lower limit threshold of the load indicator. Program.
  • the data collection module 201 is further configured to acquire T historical times before the current time. Inscribed, the load indicator of the system at each historical moment, where T is a positive integer.
  • the data collection module 201 can periodically collect the load index of the system by using a preset data collection period Period.
  • the T load indicators may be load indicators collected at T historical moments within the pre-T*Period of the current time, or may be load indicators collected during any T acquisition periods before the current time.
  • the T historical moments may be historical moments in the past several days, weeks, or months.
  • the data collection module 201 may store the collected load indicators of the system at the T historical moments in a historical database.
  • the historical database may be located in the data collection module 201 or independent of the data collection module 201.
  • the load indicator of the T historical moments may be a load indicator collected in all collection periods of the past day: w 1 , w 2 , . . . , w T
  • the change of the load index is basically in units of days.
  • the change of the system load index of the previous day the change of the current system's load index can be basically predicted.
  • the optimal load index upper threshold and the lower limit of the load index are selected, so that the adjustment of the number of VMs of the current system is adapted to the change of the load index of the system, and the number of VMs is minimized. , to meet the requirements of the default performance indicators.
  • the load index of the T historical moments may be the load index collected in all the collection periods in the past week, and it is considered that the change of the system load indicator is basically in units of weeks; or, the load indicators of the T historical moments may be past For the load indicators collected in all collection periods of one month, it is considered that the changes in the system load indicators are basically in units of months.
  • the possible values of the upper limit threshold of the load index are th 1 , th 2 ,...,th U
  • the possible values of the lower limit of the load index are tl 1 , tl 2 ,..., tl L
  • the threshold update process is
  • the historical load indicator is a reference, and a pair of optimal thresholds are selected from all possible threshold pairs (tl i , th j ), and the evaluation criterion adopted is the input return ratio, and each performance indicator meets the preset performance index.
  • the more the required performance indicators the smaller the total number of VMs used in each historical moment, and the higher the return on investment ratio.
  • the specific calculation process is shown in Figure 10 and includes the following steps:
  • the second decision module 203 acquires the load indicators w 1 , w 2 , . . . , w T of the T historical time system from the data collection module 201.
  • the data collection module 201 stores the collected load indicators in their own historical database
  • the second decision module 203 selects a set of threshold pairs (tl i , th j ) from the candidate pair of thresholds, and uses the threshold pair as the lower limit of the load index threshold and the upper limit of the load index, and the load of the system for the T historical moments.
  • the indicators w 1 , w 2 , . . . , w T determine the second operation to be performed on the system;
  • the second operation to be performed on the system is determined with the selected threshold pair (tl i , th j ), and the number of VMs used by the system is predicted to be v t after performing the second operation.
  • the performance index p t is calculated according to the following formula 6:
  • ROI i,j h((v 1 ,v 2 ,...,v T ),(p 1 ,p 2 ,...,p T )).
  • the ROI i,j may be defined as a ratio of the total time of the system performance indicator to the preset performance indicator threshold in the T historical moments, and the total time of the VM running.
  • the ROI i, j can be calculated by the following formula 8:
  • step S1004 selecting, from the U ⁇ L ROIs calculated in step S1003, a threshold pair corresponding to the largest ROI as the optimal threshold pair, and using the upper limit threshold of the load index in the optimal threshold pair as the optimal upper limit threshold of the load index.
  • the lower limit of the load index is used as the optimal lower limit threshold of the load indicator;
  • S1005 Adjust the number of VMs used by the system according to the selected optimal load index lower threshold and the optimal load index upper threshold.
  • Embodiment 3 provides a method for determining a load index upper limit threshold and a load indicator lower limit threshold. This method can solve the problem that the threshold is difficult to determine.
  • the second decision module 203 can periodically run the steps S1001 to S1003 mentioned in this embodiment to adjust the upper limit threshold of the load index and the lower limit of the load index, so that when the number of VMs used by the system is adjusted, the optimal threshold is used. Adjust to improve the utilization efficiency of VM.
  • the load index upper threshold and the load index lower threshold are manually specified, and the user is instructed to invoke the cloud service provider's Application Programming Interface (API) to threshold.
  • API Application Programming Interface
  • the configured threshold cannot be adapted to the current system load. Blind settings may result in wasted resources or performance indicators cannot meet the requirements.
  • the threshold can adapt to the change of the system load, and the utilization efficiency of the VM is improved.
  • the fourth embodiment describes the relationship between the device for determining the VM number adjustment operation and the VNF provided by the embodiment of the present invention.
  • the apparatus for determining the VM number adjustment operation may be deployed inside each VNF, as shown in FIG. 11; or may be deployed outside the VNF to manage and arrange the VNF.
  • the VNF Manager (VNFM), as shown in Figure 12.
  • the device for adjusting the operation may be located in a service control unit for performing service control or a resource control unit for performing resource control in the VNF, where the unit for performing service control is used to establish a service, release a service, and adjust a service.
  • the unit for resource control is used to control, add, delete, etc. resources including the VM.
  • each of VNF-1, VNF-2, and VNF3 has a means for determining the VM number adjustment operation.
  • the apparatus for determining the VM number adjustment operation provided by the embodiment of the present invention is deployed in a network element other than the VNF that is responsible for managing and orchestrating the VNF, for example, in the VNF manager. Since a VNF is managed by a VNF manager, for different VNFs, the apparatus for determining the VM number adjustment operation provided by the embodiment of the present invention needs to be deployed in its corresponding VNF manager to implement different VNFs. The number of VMs is adjusted.
  • the service scenarios applicable to the embodiments of the present invention may include two basic service scenarios (Basic Scenario 1 and Basic Scenario 2) as shown in FIG. 13, and the integrated scenario shown in FIG. 13
  • the basic scenario 1 shown in FIG. 13 is a daily busy hour idle feature.
  • the main feature of the scenario is that the system load will show obvious peaks and troughs with time, and the duration of the peaks and troughs. Longer. For example: every day from 8 am to 10 pm is busy, other time periods are idle;
  • the basic scenario 2 shown in FIG. 13 is a bursty business scenario.
  • the main feature of the scene is that it also includes peaks and troughs, but the duration of the peaks and troughs is shorter. For example, a short-term business peak occurred between 12:30 and 12:45.
  • Figure 14 shows an integrated scenario, which is a combination of the two basic business scenarios shown in Figure 13.
  • the embodiments of the present invention are applicable to any of the foregoing service scenarios, and can implement accurate and real-time adjustment of the number of VMs.
  • the device for determining the VM number adjustment operation provided by the embodiment of the present invention is described.
  • the following describes the fourth device for determining the VM number adjustment operation provided by the embodiment of the present invention.
  • a fourth apparatus for determining a VM number adjustment operation includes:
  • a memory 1501 configured to store a program for adjusting the number of VMs
  • the processor 1502 is configured to call the program in the memory 1501 to determine an operation of adjusting the number of VMs performed by the system.
  • the processing performed by the processor 1502 may be specifically referred to the processing in the apparatus for determining the VM number adjustment operations of the first, second, and third types provided by the embodiments of the present invention, and the repeated description is omitted.
  • processor 1502 performs the following processing:
  • the first operation includes: adding m VMs, reducing n VMs, or keeping the number of VMs unchanged, where m and n are positive integers, and m and n are load indicators of the system according to the current time and the number of VMs used by the system. definite.
  • the bus architecture may include any number of interconnected buses and bridges, specifically linked by one or more processors represented by processor 1502 and various circuits of memory represented by memory 1501.
  • the bus architecture can also link various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and, therefore, will not be further described herein.
  • the bus interface provides an interface.
  • the user interface 1503 may also be an interface capable of externally connecting the required devices, including but not limited to a keypad, a display, a speaker, a microphone, a joystick, and the like.
  • the device for determining the VM number adjustment operation provided by the embodiment of the present invention is introduced. Based on the same inventive concept as the device, the embodiment of the present invention further provides a method for determining the VM number adjustment operation, which can be implemented by referring to the device. Implementation, repetition will not be repeated.
  • FIG. 16 is a flowchart of a method for determining a VM number adjustment operation according to an embodiment of the present invention. As shown in FIG. 16, the method includes:
  • S1601 Obtain the load indicator of the current time system and the number of VMs used by the system, where The system includes one or more VMs;
  • S1603 Determine, according to the current load indicator of the system and the number of VMs used by the system, the first operation to be performed on the system;
  • the first operation includes: adding m VMs, reducing n VMs, or keeping the number of VMs unchanged, where m and n are positive integers, and m and n are load indicators of the system according to the current time and the number of VMs used by the system. definite.
  • determining a first operation to be performed on the system including:
  • the performance indicator of the system after performing the first operation is predicted according to the load indicator of the current time system and the number of VMs used by the system after the first operation is performed; the number of VMs used by the system after performing the first operation is based on The number of VMs used by the system at the current time, as well as the determined first operation predicted.
  • the method before determining the first operation to be performed on the system, the method further includes: acquiring a performance indicator of the current time system;
  • Determine the first action to perform on the system including:
  • the return value is determined according to the number of VMs used by the system and the performance index of the system under the load indicated by the current load indicator of the system.
  • determining an immediate reward value obtained by performing the first operation on the system at the current moment including:
  • the instantaneous reward value obtained by performing the first operation on the system at the current time is determined according to the predicted number of VMs used by the system after the first operation and the performance index of the system.
  • determining a maximum value of the accumulated reward values obtained by performing each of the first operations of the set of executable first operations on the system in the next state in which the system is located after performing the first operation include:
  • the largest cumulative return value is selected.
  • the method before determining the first operation to be performed on the system, the method further includes: selecting a manner of adjusting the number of VMs;
  • Determine the first action to perform on the system including:
  • the number of selected VMs is adjusted, the number of VMs used by the system is adjusted according to the current load index of the system and the number of VMs used by the system.
  • the first operation to be performed on the system is determined according to the acquired load indicator of the current time system and the number of VMs used by the system.
  • the method further includes:
  • the load indicator upper limit threshold and the load indicator lower limit threshold are determined according to the following conditions according to the load index of each historical time system among the T historical moments before the acquired current time:
  • the performance indicators of the system meet the preset performance index requirements, and the number of VMs used is the least, where T is a positive integer.
  • the upper limit threshold of the load indicator and the lower limit of the load indicator are specifically determined according to the following steps:
  • the input return ratio is determined by performing the following steps:
  • For the load indicator of each historical time system acquired predict the number of VMs used by the system after performing the second operation on the system according to the combination of the candidate load index upper threshold and the lower limit of the load index; and according to the historical time The load indicator and the predicted number of VMs used by the system after performing the second operation, predicting performance indicators of the system after performing the second operation;
  • the input return ratio is determined according to the number of VMs used by the system after performing the second operation and the performance index corresponding to each of the performance indicators, wherein the number of performance indicators that meet the preset performance index requirements of each performance indicator is larger. The smaller the total number of VMs used by the system at each historical moment, the higher the return on investment ratio;
  • an embodiment of the present invention provides an apparatus and method for determining a VM number adjustment operation. If the number of VMs is adjusted according to the load index of the system and the number of VMs used by the system, it is not only determined whether to increase, decrease, or keep the number of VMs, but also accurately determine the value of the VM number adjustment.
  • the number of VM adjustments can be determined in real time according to the load change of the system and the number of VMs currently used, and the real-time performance of the VM adjustment is improved.
  • the threshold of the VM number adjustment can be flexibly set according to the load of the system, and the different thresholds are evaluated to select the optimal threshold, thereby improving the system VM utilization. , to ensure the performance requirements of the system.
  • the performance indicators of the system meet the preset performance index requirements, and the number of VMs used by the system is the minimum. It can balance the performance of the system and the number of VMs of the system, and improve the VM utilization while ensuring the performance requirements of the system.
  • the method of reinforcement learning is adopted, and when the adjusted value is determined, the accumulated return value of the historical moment before the current moment is considered. Moreover, the possible return value at the next moment is also considered, so that it is no longer necessary to set the cooling time in the method shown in FIG. 1, thereby improving the real-time performance of the VM number adjustment.
  • unnecessary expansion and shrinkage operations can be reduced, and the ping-pong effect is slowed down.
  • the load indicator upper limit threshold and the load indicator lower limit threshold are adjusted according to the historical load of the system, after the VM number is adjusted, the performance index of the system meets the preset performance index requirement, and the number of VMs used is the least, and both can be taken into consideration.
  • the performance metrics of the system and the number of VMs in the system are adjusted according to the historical load of the system.
  • the load indicator upper limit threshold and the load indicator lower limit threshold are adjusted according to the historical load of the system, the available thresholds are respectively evaluated for different system load conditions, and the optimal threshold is selected, so that the adjustment system is used.
  • the number of VMs is adjusted according to the selected optimal threshold, the utilization efficiency of the VM is improved.
  • embodiments of the present invention can be provided as a method, system, or computer program product. Accordingly, the present invention may employ an entirely hardware embodiment, an entirely software embodiment, Or in the form of an embodiment of the software and hardware aspects. Moreover, the invention can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) including computer usable program code.
  • computer-usable storage media including but not limited to disk storage, CD-ROM, optical storage, etc.
  • the computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture comprising the instruction device.
  • the apparatus implements the functions specified in one or more blocks of a flow or a flow and/or block diagram of the flowchart.
  • These computer program instructions can also be loaded onto a computer or other programmable data processing device such that a series of operational steps are performed on a computer or other programmable device to produce computer-implemented processing for execution on a computer or other programmable device.
  • the instructions provide steps for implementing the functions specified in one or more of the flow or in a block or blocks of a flow diagram.

Landscapes

  • Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Supply And Distribution Of Alternating Current (AREA)
  • Debugging And Monitoring (AREA)

Abstract

La présente invention concerne une technologie de l'informatique en nuage, en particulier un dispositif et un procédé pour déterminer une opération pour régler le nombre de machines virtuelles, et est utilisée pour fournir une solution pour configurer de manière souple des machines virtuelles (VM). Dans le dispositif pour déterminer une opération pour régler le nombre de VM fournies par les modes de réalisation de la présente invention, un module de collecte de données acquiert un indice de charge d'un système et le nombre de VM utilisées par le système à l'instant présent ; et un premier module de détermination détermine, selon l'indice de charge du système et le nombre de VM utilisées par le système à l'instant présent, une première opération devant être exécutée sur le système, la première opération consistant : à augmenter le nombre de VM à m, à réduire le nombre de VM à n, ou à garder le nombre de VM inchangé, m et n étant des nombres entiers positifs et étant déterminés par le premier module de détermination selon l'indice de charge du système et le nombre de VM utilisées par le système à l'instant présent. Puisque le nombre accru ou réduit de VM peut être déterminé selon un indice de charge d'un système et le nombre de VM utilisées par le système à l'instant présent, et n'est plus une valeur réglée fixe, sa mise en œuvre est plus souple.
PCT/CN2014/094226 2014-12-18 2014-12-18 Dispositif et procédé pour déterminer une opération pour régler un nombre de machines virtuelles WO2016095162A1 (fr)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/CN2014/094226 WO2016095162A1 (fr) 2014-12-18 2014-12-18 Dispositif et procédé pour déterminer une opération pour régler un nombre de machines virtuelles
CN201480028886.9A CN106170767B (zh) 2014-12-18 2014-12-18 一种确定虚拟机数量调整操作的装置和方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2014/094226 WO2016095162A1 (fr) 2014-12-18 2014-12-18 Dispositif et procédé pour déterminer une opération pour régler un nombre de machines virtuelles

Publications (1)

Publication Number Publication Date
WO2016095162A1 true WO2016095162A1 (fr) 2016-06-23

Family

ID=56125624

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2014/094226 WO2016095162A1 (fr) 2014-12-18 2014-12-18 Dispositif et procédé pour déterminer une opération pour régler un nombre de machines virtuelles

Country Status (2)

Country Link
CN (1) CN106170767B (fr)
WO (1) WO2016095162A1 (fr)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109783221A (zh) * 2017-11-14 2019-05-21 中国移动通信有限公司研究院 一种虚拟机资源分配方法、装置及资源服务器
CN111381959A (zh) * 2018-12-29 2020-07-07 中兴通讯股份有限公司 扩容方法及装置
CN113381873A (zh) * 2020-03-10 2021-09-10 华为技术有限公司 网络设备的扩容方法和相关装置

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102214117A (zh) * 2010-04-07 2011-10-12 中兴通讯股份有限公司 虚拟机管理方法、系统及虚拟机管理服务器
CN103793278A (zh) * 2013-09-30 2014-05-14 中国电子设备系统工程公司研究所 一种基于虚拟器件运维规则的资源自动调整方法
CN103870341A (zh) * 2014-03-12 2014-06-18 汉柏科技有限公司 一种调整虚拟机资源的方法和系统

Family Cites Families (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8671294B2 (en) * 2008-03-07 2014-03-11 Raritan Americas, Inc. Environmentally cognizant power management
CN101694633A (zh) * 2009-09-30 2010-04-14 曙光信息产业(北京)有限公司 计算机作业调度设备、方法及系统
US8276139B2 (en) * 2009-09-30 2012-09-25 International Business Machines Corporation Provisioning virtual machine placement
CN101894050B (zh) * 2010-07-28 2014-04-16 山东中创软件工程股份有限公司 云资源池的jee应用资源弹性调度方法、装置及系统
CN102291390B (zh) * 2011-07-14 2014-06-04 南京邮电大学 一种基于云计算平台的防御拒绝服务攻击的方法
CN102508693A (zh) * 2011-09-29 2012-06-20 华中科技大学 基于虚拟机的Web服务器扩容系统
CN103294552A (zh) * 2012-02-22 2013-09-11 鸿富锦精密工业（深圳）有限公司 增加服务器的虚拟机配置数量的方法
WO2014101010A1 (fr) * 2012-12-26 2014-07-03 华为技术有限公司 Procédé de gestion de ressources de système de machine virtuelle, système de machine virtuelle et appareil
US9665154B2 (en) * 2013-05-31 2017-05-30 Lenovo Enterprise Solutions (Singapore) Pte. Ltd. Subsystem-level power management in a multi-node virtual machine environment
CN103533063A (zh) * 2013-10-18 2014-01-22 北京华胜天成科技股份有限公司 一种可实现web应用资源动态扩展的方法及装置
CN103595780B (zh) * 2013-11-08 2017-01-04 中国人民解放军理工大学 基于消重的云计算资源调度方法
CN103984602A (zh) * 2014-05-20 2014-08-13 华为技术有限公司 一种vm资源调度方法、装置及系统

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102214117A (zh) * 2010-04-07 2011-10-12 中兴通讯股份有限公司 虚拟机管理方法、系统及虚拟机管理服务器
CN103793278A (zh) * 2013-09-30 2014-05-14 中国电子设备系统工程公司研究所 一种基于虚拟器件运维规则的资源自动调整方法
CN103870341A (zh) * 2014-03-12 2014-06-18 汉柏科技有限公司 一种调整虚拟机资源的方法和系统

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109783221A (zh) * 2017-11-14 2019-05-21 中国移动通信有限公司研究院 一种虚拟机资源分配方法、装置及资源服务器
CN109783221B (zh) * 2017-11-14 2020-12-29 中国移动通信有限公司研究院 一种虚拟机资源分配方法、装置及资源服务器
CN111381959A (zh) * 2018-12-29 2020-07-07 中兴通讯股份有限公司 扩容方法及装置
CN113381873A (zh) * 2020-03-10 2021-09-10 华为技术有限公司 网络设备的扩容方法和相关装置
CN113381873B (zh) * 2020-03-10 2023-02-03 华为技术有限公司 网络设备的扩容方法和相关装置

Also Published As

Publication number Publication date
CN106170767B (zh) 2020-05-15
CN106170767A (zh) 2016-11-30

Similar Documents

Publication Publication Date Title
WO2020258920A1 (fr) Procédé et appareil de gestion de ressources de tranches de réseau
US7552152B2 (en) Risk-modulated proactive data migration for maximizing utility in storage systems
US10200446B2 (en) Method and system to estimate user desired delay for resource allocation for mobile cloud applications
LaCurts et al. Cicada: Introducing predictive guarantees for cloud networks
JP6493400B2 (ja) サービスチェーン管理装置、サービスチェーン管理システム、サービスチェーン管理方法、及び、プログラム
US20080104605A1 (en) Methods and apparatus for dynamic placement of heterogeneous workloads
US20120221730A1 (en) Resource control system and resource control method
US20200012602A1 (en) Cache allocation method, and apparatus
CN103916396A (zh) 一种基于负载自适应的云平台应用实例自动伸缩方法
JP2008519322A (ja) 電気通信サービスおよび／またはネットワーク管理用のプラットフォームにおけるリソースの管理方法、対応プラットフォーム、およびそのコンピュータ・プログラム生成物
CN108366082A (zh) 扩容方法及扩容装置
WO2016095162A1 (fr) Dispositif et procédé pour déterminer une opération pour régler un nombre de machines virtuelles
CN103812911B (zh) 一种控制使用PaaS云计算平台服务资源的方法和系统
Phung-Duc et al. Design and analysis of deadline and budget constrained autoscaling (DBCA) algorithm for 5G mobile networks
EP4013087A1 (fr) Système et procédé de gestion de réseau mobile 5g
KR101448413B1 (ko) Atca-기반 장비에서 통신 트래픽을 스케줄링하기 위한 방법 및 장치
Ali et al. Proactive vnf scaling and placement in 5g o-ran using ml
Rachkidi et al. Towards efficient automatic scaling and adaptive cost-optimized ehealth services in cloud
CN113254191A (zh) 用于运行应用的方法、电子设备和计算机程序产品
CN118056185A (zh) 接收外部信息并且包含短期和长期资源规划的异常感知云资源管理系统
CN108268363B (zh) 用于业务容量管理的方法和设备
CN106533730B (zh) Hadoop集群组件指标的采集方法及装置
Tolosana-Calasanz et al. Enforcing quality of service on opennebula-based shared clouds
EP3360287A1 (fr) Analyse automatique d'utilisation de bande passante par le client pour promouvoir une capacité dynamique
Kumar et al. A QoS-based reactive auto scaler for cloud environment

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14908197

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 14908197

Country of ref document: EP

Kind code of ref document: A1