CN114629909A - Cloud resource cost analysis method - Google Patents
Cloud resource cost analysis method Download PDFInfo
- Publication number
- CN114629909A CN114629909A CN202210315810.0A CN202210315810A CN114629909A CN 114629909 A CN114629909 A CN 114629909A CN 202210315810 A CN202210315810 A CN 202210315810A CN 114629909 A CN114629909 A CN 114629909A
- Authority
- CN
- China
- Prior art keywords
- index
- cost
- cloud resource
- evaluation
- utilization
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000007728 cost analysis Methods 0.000 title claims abstract description 18
- 238000011156 evaluation Methods 0.000 claims abstract description 148
- 238000005457 optimization Methods 0.000 claims abstract description 56
- 239000011159 matrix material Substances 0.000 claims description 83
- 238000003860 storage Methods 0.000 claims description 52
- 238000013139 quantization Methods 0.000 claims description 18
- 238000004364 calculation method Methods 0.000 claims description 15
- 230000006870 function Effects 0.000 claims description 13
- 238000012360 testing method Methods 0.000 claims description 4
- 230000001174 ascending effect Effects 0.000 claims description 3
- 230000008901 benefit Effects 0.000 abstract description 7
- 238000004891 communication Methods 0.000 abstract description 3
- 238000009826 distribution Methods 0.000 description 28
- 238000007726 management method Methods 0.000 description 17
- 230000009467 reduction Effects 0.000 description 8
- 230000007704 transition Effects 0.000 description 8
- 238000012544 monitoring process Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000006243 chemical reaction Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000012795 verification Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000009897 systematic effect Effects 0.000 description 2
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000019771 cognition Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000013468 resource allocation Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
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
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/3003—Monitoring arrangements specially adapted to the computing system or computing system component being monitored
- G06F11/3006—Monitoring arrangements specially adapted to the computing system or computing system component being monitored where the computing system is distributed, e.g. networked systems, clusters, multiprocessor systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
- H04L43/0876—Network utilisation, e.g. volume of load or congestion level
- H04L43/0894—Packet rate
-
- 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
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
- H04L67/1004—Server selection for load balancing
-
- 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
- H04L67/1097—Protocols in which an application is distributed across nodes in the network for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Environmental & Geological Engineering (AREA)
- Computing Systems (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Quality & Reliability (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of communication, in particular to a cloud resource cost analysis method. According to the cloud resource cost condition evaluation method and the cloud resource cost condition evaluation system, practical and effective optimization suggestions are provided for the cloud resource cost through the cloud resource cost condition evaluation step, cost expenditure evaluation and cloud resource use benefit evaluation of all current cloud products are provided for a user, and the user can visually know the current cloud resource cost condition from the score level. In addition, the cloud resource utilization condition is accurately evaluated through the cloud resource utilization condition evaluation step, the quality of the cloud resource utilization mode is effectively and objectively evaluated, and a useful basis is provided for future cloud resource optimization.
Description
Technical Field
The invention relates to the technical field of communication, in particular to a cloud resource cost analysis method.
Background
Cloud resources are a generic term for cloud computing-related services, including but not limited to cloud services such as cloud servers, block storage, virtual private networks, and elastic IP. The characteristics of the cloud resources include: 1. the resources are wide in geographical distribution, and the types and the quantity of the resources are huge. 2. Resources can be dynamically added to and removed from the cloud computing platform, and physical resources can be converted into scalable virtual shared resources. 3. And virtualization, namely virtualizing physical resources into a resource pool to perform uniform and flexible on-demand distribution. 4. Very large scale, a cloud typically has thousands of servers, which can give users more computing power than ever before. 5. And the universality can construct diversified applications in the cloud environment.
A Cloud Management Platform (CMP) is a unified Management Platform for data center resources, and can manage multiple open-source or heterogeneous Cloud computing technologies or products, such as cloudstock, OpenStack, VMware, Docker, and the like. A Cloud Management Platform (CMP) provides unified cloud management, supports organizations to quickly transform their existing virtual infrastructure into a highly extensible private cloud, while fully utilizing public cloud resources. The multi-cloud management platform refers to a platform capable of managing the cloud resources. The multi-cloud management platform can be simultaneously accessed and managed to resources from different cloud manufacturers, so that the resource management and operation and maintenance efficiency of enterprises is improved, the resource allocation of the cloud manufacturers is realized through the cross-cloud management capability, and the enterprise cost is reduced. In the cloud computing era, a multi-cloud management platform is used as a tool for improving the utilization rate of cloud resources, and a richer application scene and a wider development space are bound to be met under the value drive of social demands.
The multi-cloud management platform is continuously updated in an iterative mode, the existing multi-cloud management platform is basically complete, and the function of managing a plurality of cloud resources can be well achieved. However, as the applications of the multiple cloud management platforms are gradually increased, the defects of the multiple cloud management platforms are also seen. Firstly, the existing multi-cloud management platform on the market basically has the function of cost operation analysis, and the general functional module has the functions of metering and charging, bill and report management, cost visualization, cost optimization and the like. The above functions can meet the requirement of operators for deep cost analysis, but cannot provide overall analysis result display of cloud resource cost. Secondly, although the current cost optimization function board block on the market contains a specific optimization suggestion for cloud resource usage, the overall evaluation on whether the cloud resource is effectively used is lacked, so that a user cannot effectively understand the specific problems of cloud resource cost expenditure and effective cloud resource usage. The prior art cannot provide current cloud resource cost condition assessment and cloud resource utilization condition assessment for a user, so that the user cannot master the cost expenditure and the use benefit of a cloud product, the accuracy of cloud resource optimization is affected, and the method is one of the problems to be solved urgently in the technical field of communication. Therefore, a method for analyzing the cost of cloud resources is needed to provide a user with a current evaluation of the cost and utilization of cloud resources.
Disclosure of Invention
The invention aims to overcome at least one defect of the prior art, and provides an evaluation method of cloud resource cost and cloud resource utilization condition, which is used for solving the problem that a user cannot evaluate the cloud resource cost condition and the cloud resource utilization condition.
The technical scheme adopted by the invention is as follows:
a cloud resource cost analysis method includes: a cloud resource cost condition evaluation step and a cloud resource utilization condition evaluation step;
the cloud resource cost evaluation step comprises:
acquiring index data, and determining a cost utilization strategy to be implemented according to the index data; the index data is index data of an evaluation product; the cost utilization strategy is used for optimizing the cost of the evaluation product;
calculating an optimized cost rate according to the cloud resource total cost before the cost utilization strategy is implemented and the cloud resource total cost after the cost utilization strategy is implemented;
inputting the cost rate capable of being optimized into an evaluation function, and calculating the cloud resource cost condition score;
the cloud resource utilization condition evaluation step includes:
setting an evaluation criterion, wherein the evaluation criterion comprises a plurality of quantization indexes;
constructing a judgment matrix according to the evaluation criterion; the judgment matrix determines the weight value of each element of the judgment matrix through pairwise comparison between the quantitative indexes;
carrying out consistency check on the judgment matrix;
if the test fails, the judgment matrix needs to be reconstructed;
if the check is passed, calculating a cloud resource utilization condition score according to the weight vector of the judgment matrix and the initial score of the quantitative index;
and carrying out cloud resource cost evaluation according to the cloud resource cost condition score and the cloud resource utilization condition score.
Specifically, the existing technology cannot provide current cloud resource cost condition evaluation and cloud resource utilization condition evaluation for a user, so that the user lacks knowledge of cloud resource cost expenditure and use benefit, the user is slightly influenced in recognizing cloud resource cost, and the user is seriously caused to take wrong optimization measures for cloud resources, thereby causing economic loss. In order to avoid the error of cloud resource cost evaluation and more reasonably utilize cloud resources, the scheme adopts a cloud resource cost analysis method.
Evaluating the cost condition of the cloud resources: firstly, the multi-cloud management platform acquires monitoring data, namely index data, by butting a monitoring API of a cloud service provider. The cloud service provider refers to a manufacturer providing cloud services, such as Ali cloud, Tencent cloud, Huayun, and the like. The index data is data for measuring the operation performance and the requirement of cloud resources/cloud services/cloud products. And selecting corresponding suggestions according to the index data to optimize the cloud resource cost, so that the cloud resources can better match with the user requirements. The cloud resource cost optimization may be to reduce the cost by reducing the device configuration, or may be to increase the cost by increasing the device configuration. And then, calculating the current cloud resource total cost and the cloud resource total cost after implementing the cost utilization strategy. The cost and the cost rate which can be optimized are calculated according to the current cloud resource total cost and the cloud resource total cost after the cost utilization strategy is implemented, so that a user can intuitively know the optimization effect of the cost utilization strategy. And finally, inputting the optimized cost rate into an evaluation function to obtain the cloud resource cost condition score.
Evaluating the utilization condition of the cloud resources: first, an evaluation criterion is set, namely, a hierarchical structure model is established based on an analytic hierarchy process. The Analytic Hierarchy Process (AHP) is a systematic method that takes a complex multi-objective decision problem as a system, decomposes a target into multiple targets or criteria, further decomposes the targets into multiple levels of multiple indexes (or criteria, constraints), and calculates the level order (weight) and total order by a qualitative index fuzzy quantization method to be used as a target (multiple indexes) and multi-scheme optimization decision. The hierarchical structure model includes: the system comprises a target layer, a standard layer and a scheme layer, wherein the target layer is set to evaluate the utilization condition of resources, the standard layer is set to a plurality of quantitative indexes, and the scheme layer is set to be a user. Then, the relative importance degree between the quantization indexes is measured through pairwise comparison between the quantization indexes, the weight value of each quantization index is determined, namely the weight of each element of the judgment matrix is determined, the comparison scale of the hierarchy analysis method 1-9 is adopted for the weight value, and the judgment matrix is constructed according to the hierarchy model and the weight value of the quantization indexes. And then, carrying out consistency check on the judgment matrix, reconstructing the judgment matrix if the judgment matrix does not pass the judgment matrix, otherwise determining the weight vector of the matrix, and calculating the cloud resource utilization condition score according to the weight vector of the matrix and the initial score of the quantization index. The initial score of the quantitative index is the average score of multiple scores of the quantitative index in a specified time period, and the score of each score is the average value of scores of the quantitative indexes given to multiple professionals. According to the scheme, through the cloud resource cost condition evaluation step, a practical and effective optimization suggestion is provided for the cloud resource cost, cost expenditure evaluation and cloud resource use benefit evaluation of all current cloud products are provided for a user, and the user can visually know the current cloud resource cost condition from the score. In addition, the cloud resource utilization condition is accurately evaluated through the cloud resource utilization condition evaluation step, the quality of the cloud resource utilization mode is effectively and objectively evaluated, and a useful basis is provided for future cloud resource optimization.
Further, the index data includes: calculating performance indexes, storing performance indexes, network performance indexes and database performance indexes.
Specifically, calculating the performance indicator includes: CPU utilization rate, memory utilization rate, total network flow, total IO amount of a disk and resource use duration; the storage performance indicators include: binding to the ECS host and the API request times; the network performance indexes are as follows: bandwidth utilization; the database performance indicators include: connection number utilization rate, CPU utilization rate, memory utilization rate and resource use duration.
Further, determining a cost utilization policy from the metric data comprises:
judging the index data according to a preset judgment condition to obtain a judgment result;
determining the cost utilization strategy according to a judgment result;
the determining includes: the method comprises the following steps of (1) predicting and judging idle resources, predicting and judging over-allocated resources, predicting and judging resource expansion and rationality of a payment mode;
the cost utilization strategy comprises the following steps: a descending allocation optimization strategy, an ascending allocation optimization strategy, a resource release optimization strategy and a payment mode optimization strategy.
Specifically, the determination of idle resources for calculating performance indicators: and judging whether the CPU utilization rate max value, the memory utilization rate max value, the total network flow and the total disk IO amount are smaller than an idle threshold, if an index smaller than the idle threshold exists, judging that the resource corresponding to the index is an idle resource, and providing a resource release optimization strategy. The idle threshold of the CPU utilization rate max value is 5%, the idle threshold of the memory utilization rate max value is 15%, the idle threshold of the network total flow is 30M, and the idle threshold of the disk IO total amount is 30M.
And (3) calculating the over-allocation resource judgment of the performance index: and judging whether the max value of the CPU utilization rate and the max value of the memory utilization rate are smaller than the super-distribution threshold value, if so, judging that the resource corresponding to the index is the super-distribution resource, and providing a distribution reduction optimization strategy. The super-distribution threshold value of the CPU utilization rate max value is 40%, and the super-distribution threshold value of the memory utilization rate max value is 40%.
And (3) calculating capacity expansion judgment of the performance index: and judging whether the max value of the CPU utilization rate and the max value of the memory utilization rate are larger than a capacity expansion threshold, if so, judging that the resource corresponding to the index is a capacity expansion resource, and providing a distribution-increasing optimization strategy. The expansion threshold of the CPU utilization rate max value is 80%, and the expansion threshold of the memory utilization rate max value is 40%.
And (3) calculating the reasonability judgment of the payment mode of the performance index: and judging whether the using time is less than or greater than the transition threshold, and providing a payment mode optimization strategy according to the judgment result. The transition threshold of the payment mode is 11 days, if the using time is longer than 11 days, the payment mode is suitable for covering a month and a year, and if the using time is shorter than 11 days, the payment mode is suitable for paying according to the amount.
Judging idle resources of the storage performance indexes: the decision block stores whether to bind to the ECS host, and if not, the block stores as an idle resource, providing a resource release optimization strategy.
And (3) judging the over-allocated resources of the storage performance indexes: and judging whether the number of object storage API requests reaches a super-allocation threshold, if so, providing a resource corresponding to the index as a super-allocation resource, and providing a distribution reduction optimization strategy. The over-matching threshold of the storage performance index is the frequency of 1 visit of a single file per month and the frequency of 1 visit of a single file for 90 days.
And (3) judging the reasonability of the payment mode of the storage performance index: comparing which of the costs incurred by using the storage packet and using the traffic packet is higher, and providing a payment pattern optimization strategy based on the comparison result.
And (3) judging idle resources of the network performance indexes: and judging whether the elastic public network IP is bound to the ECS host, if not, the elastic public network IP is idle resources, and providing a resource release optimization strategy.
And (3) judging the over-allocated resources of the network performance indexes: and judging whether the max value of the bandwidth utilization rate is smaller than the super-distribution threshold value, and if so, providing a distribution reduction optimization strategy for the elastic public network IP which is a super-distribution resource. The over-provisioning threshold for the network performance indicator is 40%.
Judging idle resources of the database performance indexes: and judging whether the average value of the utilization rate of the RDS connection number is smaller than an idle threshold value or not, if so, determining the RDS as an idle resource, and providing a resource release optimization strategy. The idle threshold for the database performance index is 0.1%.
Judging the over-allocated resources of the database performance indexes: and judging whether the max value of the CPU utilization rate and the max value of the memory utilization rate are smaller than the super-distribution threshold value, if so, judging that the resource corresponding to the index is the super-distribution resource, and providing a distribution reduction optimization strategy. The over-provisioning threshold for the CPU utilization max value is 40%. The over-provisioning threshold for the memory usage max value is 40%.
And (3) judging the reasonability of the payment mode of the database performance index: and judging whether the service life is less than or greater than the transition threshold value, and providing a payment mode optimization strategy according to the judgment result. The transition threshold of the payment mode is 20 days, if the using time is longer than 20 days, the payment mode is suitable for covering a month and a year, and if the using time is shorter than 20 days, the payment mode is suitable for paying according to the amount.
Further, the optimizable cost ratio is:
the optimizable cost rate (optimizable cost/total cost of cloud resources before implementing the cost utilization policy) x 100%;
wherein the optimizable cost is a difference between a total cost of the cloud resource before implementing the cost utilization policy and a total cost of the cloud resource after implementing the cost utilization policy.
Further, the evaluation function is:
score ═ (1-optimizable cost rate) × 100;
wherein, Score is cloud resource cost status Score.
Specifically, the higher the optimizable cost rate, the greater the degree to which the representation needs to be optimized, and the lower the cloud resource cost status score.
Further, the quantization index includes: evaluating field indexes and evaluating product indexes, wherein each evaluating field index corresponds to a plurality of evaluating product indexes; the evaluation field index comprises a calculation field index, a storage field index, a network field index and a database field index; the calculating of the evaluation product index corresponding to the field index comprises: cloud host indexes; the evaluation product index corresponding to the storage field index comprises: an object storage index and a file storage index; the evaluation product index corresponding to the network field index comprises: load balancing indexes and elastic public network IP indexes; the evaluation product indexes corresponding to the database field indexes comprise: database instance indices.
Further, the cloud host metrics include: CPU utilization rate, memory utilization rate, public network outgoing bandwidth utilization rate and disk utilization rate; the object storage metrics include: an effective request rate; the file storage index includes: a capacity usage rate; the load balancing indexes include: QPS rate and concurrent connectivity rate; the elastic public network IP indexes comprise: network ingress bandwidth utilization and network egress bandwidth utilization; the database instance metrics include: connection number usage, CPU usage, and memory usage.
Further, constructing a judgment matrix according to the evaluation criterion, comprising:
constructing a judgment matrix among the computing field, the storage field, the network field and the database field:
further, constructing a judgment matrix according to the evaluation criterion, further comprising:
constructing a judgment matrix of the evaluation product index corresponding to the calculation field index:
constructing a judgment matrix of the evaluation product index corresponding to the storage field index:
constructing a judgment matrix of the evaluation product index corresponding to the network field index:
constructing a judgment matrix of the evaluation product index corresponding to the database field index:
further, according to the weight vector of the judgment matrix, calculating a cloud resource utilization state score:
obtaining an initial score of the evaluation product index;
according to the weight vector (a, b, c, d) of the judgment matrixTCalculating the cloud resource utilization condition score by adopting the following formula:
S＝a×A+b×B+c×C+d×D；
wherein, in order to calculate the weight of the domain index, a is the initial score of the calculated domain index, B is the weight of the stored domain index, B is the initial score of the stored domain index, C is the weight of the network domain index, C is the initial score of the network domain index, and d is the weight of the database domain index. D is the initial score of the database index.
Specifically, the initial score of the evaluation product is the average score of the evaluation product over 24 hours. The process of obtaining the initial score is as follows: and acquiring the evaluation product index data every 30 minutes within 24 hours, and scoring based on forward processing of the evaluation product index data.
The indexes selected by the scheme are all interval indexes, namely the index data is closer to a certain interval, and the index score is higher. For the convenience of subsequent calculation analysis, all indexes need to be converted into forward indexes. Taking the utilization rate of the cloud host CPU as an example: { xiThe CPU utilization rate sequence (i monitoring data in total) in one hour is defined as the interval with the best CPU utilization rate of the cloud host as default [ a, b ]]The score of the initial score is calculated by the formula
The best interval range for index scoring is shown in FIG. 9, and the interval range is set by multiple experts.
Calculating an initial score A ═ a of the domain index1×A1+a2×A2+a3×A3+a4×A4(ii) a Wherein A is1，A2，A3，A4Calculating an initial score of an evaluation product index corresponding to the field index; a is1，a2，a3，a4The weight of the evaluation product index corresponding to the field index is calculated.
The initial score B ═ B of the storage area index1×B1+b2×B2+b3×B3+b4×B4(ii) a Wherein, B1，B2，B3，B4Storing an initial score of an evaluation product index corresponding to the field index; b1，b2，b3，b4And storing the weight of the evaluation product index corresponding to the field index.
The initial score C of the network domain index is C1×C1+c2×C2+c3×C3+c4×C4(ii) a Wherein, C1，C2，C3，C4An initial score of an evaluation product index corresponding to the network field index; c. C1，c2，c3，c4And the weight of the evaluation product index corresponding to the network field index.
An initial score D ═ D of the database domain index1×D1+d2×D2+d3×D3+d4×D4(ii) a Wherein D is1，D2，D3，D4The initial scores of the evaluation product indexes corresponding to the database field indexes are obtained; d1，d2，d3，d4And the weight of the evaluation product index corresponding to the database field index.
Compared with the prior art, the invention has the beneficial effects that: according to the scheme, through the cloud resource cost condition evaluation step, a practical and effective optimization suggestion is provided for the cloud resource cost, cost expenditure evaluation and cloud resource use benefit evaluation of all current cloud products are provided for a user, and the user can visually know the current cloud resource cost condition from the score level. In addition, the cloud resource utilization condition is accurately evaluated through the cloud resource utilization condition evaluation step, the quality of the cloud resource utilization mode is effectively and objectively evaluated, and a useful basis is provided for future cloud resource optimization.
Drawings
FIG. 1 is a schematic diagram of the cost utilization strategy of the present invention;
FIG. 2 is a diagram of a cloud resource utilization index hierarchy according to the present invention;
FIG. 3 is a table of an assessment domain matrix of the present invention;
FIG. 4 is a table of a domain index matrix for the present invention;
FIG. 5 is a table of a storage domain indicator matrix of the present invention;
FIG. 6 is a table of network domain indicator matrices of the present invention;
FIG. 7 is a database domain index matrix table of the present invention;
FIG. 8 is a table of random consistency indicators according to the present invention;
FIG. 9 is a schematic diagram of an optimal interval for index utilization according to the present invention;
FIG. 10 is a table of low utilization status scores for cloud host resources in accordance with the present invention;
FIG. 11 is a table of scores for the fair use of all resources in accordance with the present invention;
FIG. 12 is a partial resource overload utilization scoring table in accordance with the present invention;
FIG. 13 is a table of low utilization index data derived from the index data of the present invention;
FIG. 14 is a table of data indicating the reasonable utilization of resources according to the index data of the present invention;
fig. 15 is a table of partial resource overload utilization data derived from the index data of the present invention.
Detailed Description
The drawings are only for purposes of illustration and are not to be construed as limiting the invention. For a better understanding of the following embodiments, certain features of the drawings may be omitted, enlarged or reduced, and do not represent the size of an actual product; it will be understood by those skilled in the art that certain well-known structures in the drawings and descriptions thereof may be omitted.
Example 1
The embodiment provides a cloud resource cost analysis method, which comprises the following steps: a cloud resource cost condition evaluation step and a cloud resource utilization condition evaluation step;
the cloud resource cost evaluation step comprises:
acquiring index data, and determining a cost utilization strategy to be implemented according to the index data; the index data is index data of an evaluation product; the cost utilization strategy is used for optimizing the cost of the evaluation product;
calculating an optimized cost rate according to the cloud resource total cost before the cost utilization strategy is implemented and the cloud resource total cost after the cost utilization strategy is implemented;
inputting the cost rate capable of being optimized into an evaluation function, and calculating the cloud resource cost condition score;
the cloud resource utilization condition evaluation step includes:
setting an evaluation criterion, wherein the evaluation criterion comprises a plurality of quantization indexes;
constructing a judgment matrix according to the evaluation criterion; the judgment matrix determines the weight value of each element of the judgment matrix through pairwise comparison between the quantitative indexes;
carrying out consistency check on the judgment matrix;
if the test fails, the judgment matrix needs to be reconstructed;
if the check is passed, calculating a cloud resource utilization condition score according to the weight vector of the judgment matrix and the initial score of the quantitative index;
and carrying out cloud resource cost evaluation according to the cloud resource cost condition score and the cloud resource utilization condition score.
Specifically, the existing technology cannot provide current cloud resource cost condition evaluation and cloud resource utilization condition evaluation for a user, so that the user lacks understanding of cloud resource cost expenditure and use benefit, the user is influenced on cloud resource cost cognition in a light case, and the user takes wrong optimization measures on cloud resources in a severe case, so that economic loss is caused. In order to avoid the error of cloud resource cost evaluation and more reasonably utilize cloud resources, the scheme adopts a cloud resource cost analysis method.
Evaluating the cost condition of the cloud resources: firstly, the multi-cloud management platform acquires monitoring data, namely index data, by butting a monitoring API of a cloud service provider. The cloud service provider refers to a manufacturer providing cloud services, such as Ali cloud, Tencent cloud, Huayun, and the like. The index data is data for measuring the operation performance and the requirement of cloud resources/cloud services/cloud products. And selecting corresponding suggestions according to the index data to optimize the cloud resource cost, so that the cloud resources can better match with the user requirements. The cloud resource cost optimization may be to reduce the cost by reducing the device configuration, or may be to increase the cost by increasing the device configuration. And then, calculating the current cloud resource total cost and the cloud resource total cost after implementing the cost utilization strategy. The cost and the cost rate which can be optimized are calculated according to the current cloud resource total cost and the cloud resource total cost after the cost utilization strategy is implemented, so that a user can intuitively know the optimization effect of the cost utilization strategy. And finally, inputting the optimized cost rate into an evaluation function to obtain the cloud resource cost condition score.
Evaluating the utilization condition of the cloud resources: first, an evaluation criterion is set, namely, a hierarchical structure model is established based on an analytic hierarchy process. The Analytic Hierarchy Process (AHP) is a systematic method that takes a complex multi-objective decision problem as a system, decomposes a target into multiple targets or criteria, further decomposes the targets into multiple levels of multiple indexes (or criteria, constraints), and calculates the level order (weight) and total order by a qualitative index fuzzy quantization method to be used as a target (multiple indexes) and multi-scheme optimization decision. The hierarchical structure model includes: the system comprises a target layer, a standard layer and a scheme layer, wherein the target layer is set to evaluate the utilization condition of resources, the standard layer is set to a plurality of quantitative indexes, and the scheme layer is set to be a user. Then, the relative importance degree between the quantization indexes is measured through pairwise comparison between the quantization indexes, the weight value of each quantization index is determined, namely the weight of each element of the judgment matrix is determined, the comparison scale of the hierarchy analysis method 1-9 is adopted for the weight value, and the judgment matrix is constructed according to the hierarchy model and the weight value of the quantization indexes. And then, carrying out consistency check on the judgment matrix, reconstructing the judgment matrix if the judgment matrix does not pass the judgment matrix, otherwise determining the weight vector of the matrix, and calculating the cloud resource utilization condition score according to the weight vector of the matrix and the initial score of the quantization index. The initial score of the quantitative index is the average score of multiple scores of the quantitative index in a specified time period, and the score of each score is the average value of scores of the quantitative indexes given to multiple professionals. According to the scheme, through the cloud resource cost condition evaluation step, a practical and effective optimization suggestion is provided for the cloud resource cost, cost expenditure evaluation and cloud resource use benefit evaluation of all current cloud products are provided for a user, and the user can visually know the current cloud resource cost condition from the score level. In addition, the cloud resource utilization condition is accurately evaluated through the cloud resource utilization condition evaluation step, the quality of the cloud resource utilization mode is effectively and objectively evaluated, and a useful basis is provided for future cloud resource optimization.
Further, the metric data includes: calculating performance indexes, storing performance indexes, network performance indexes and database performance indexes.
Specifically, calculating the performance indicator includes: CPU utilization rate, memory utilization rate, total network flow, total IO amount of a disk and resource use duration; the storage performance indicators include: binding to the ECS host and the API request times; the network performance indexes are as follows: bandwidth utilization; the database performance indicators include: connection number utilization rate, CPU utilization rate, memory utilization rate and resource use duration.
Fig. 1 is a schematic diagram of the cost utilization policy of the present invention, and as shown in the figure, determining the cost utilization policy according to the index data includes:
judging the index data according to a preset judgment condition to obtain a judgment result;
determining the cost utilization strategy according to the judgment result;
the determining includes: the method comprises the following steps of (1) predicting and judging idle resources, predicting and judging over-allocated resources, predicting and judging resource expansion and rationality of a payment mode;
the cost utilization strategy comprises the following steps: a descending allocation optimization strategy, an ascending allocation optimization strategy, a resource release optimization strategy and a payment mode optimization strategy.
Specifically, the determination of idle resources for calculating performance indicators: and judging whether the CPU utilization rate max value, the memory utilization rate max value, the total network flow and the total disk IO amount are smaller than an idle threshold, if an index smaller than the idle threshold exists, judging that the resource corresponding to the index is an idle resource, and providing a resource release optimization strategy. The idle threshold of the CPU utilization rate max value is 5%, the idle threshold of the memory utilization rate max value is 15%, the idle threshold of the network total flow is 30M, and the idle threshold of the disk IO total amount is 30M.
And (3) calculating the over-allocation resource judgment of the performance index: and judging whether the max value of the CPU utilization rate and the max value of the memory utilization rate are smaller than the over-distribution threshold value, if the indexes smaller than the over-distribution threshold value exist, judging that the resources corresponding to the indexes are over-distribution resources, and providing a distribution reduction optimization strategy. The super-distribution threshold value of the CPU utilization rate max value is 40%, and the super-distribution threshold value of the memory utilization rate max value is 40%.
And (3) calculating capacity expansion judgment of the performance index: and judging whether the max value of the CPU utilization rate and the max value of the memory utilization rate are greater than a capacity expansion threshold, if the index greater than the capacity expansion threshold exists, judging the resource corresponding to the index as a capacity expansion resource, and providing a distribution optimization strategy. The expansion threshold of the CPU utilization rate max value is 80%, and the expansion threshold of the memory utilization rate max value is 40%.
And (3) calculating the reasonability judgment of the payment mode of the performance index: and judging whether the service life is less than or greater than the transition threshold value, and providing a payment mode optimization strategy according to the judgment result. The transition threshold of the payment mode is 11 days, if the using time is longer than 11 days, the payment mode is suitable for covering a month and a year, and if the using time is shorter than 11 days, the payment mode is suitable for paying according to the amount.
Judging idle resources of the storage performance indexes: the decision block stores whether to bind to the ECS host, and if not, the block stores as an idle resource, providing a resource release optimization strategy.
And (3) judging the over-allocated resources of the storage performance indexes: and judging whether the number of object storage API requests reaches a super-allocation threshold, if so, providing a resource corresponding to the index as a super-allocation resource, and providing a distribution reduction optimization strategy. The over-provisioning threshold for the storage performance indicator is the frequency of 1 access per month for a single file and 1 access for 90 days for a single file.
And (3) judging the reasonability of the payment mode of the storage performance index: comparing which of the costs incurred by using the storage packet and using the traffic packet is higher, and providing a payment pattern optimization strategy based on the comparison result.
And (3) judging idle resources of the network performance indexes: and judging whether the elastic public network IP is bound to the ECS host, if not, the elastic public network IP is idle resource and provides a resource release optimization strategy.
And (3) judging the over-allocated resources of the network performance indexes: and judging whether the max value of the bandwidth utilization rate is smaller than the super-distribution threshold value, and if so, providing a distribution reduction optimization strategy for the elastic public network IP which is a super-distribution resource. The over-provisioning threshold for the network performance indicator is 40%.
Judging idle resources of the database performance indexes: and judging whether the average value of the utilization rate of the RDS connection number is smaller than an idle threshold value or not, if so, determining the RDS as an idle resource, and providing a resource release optimization strategy. The idle threshold for the database performance index is 0.1%.
Judging the over-allocated resources of the database performance indexes: and judging whether the max value of the CPU utilization rate and the max value of the memory utilization rate are smaller than the super-distribution threshold value, if so, judging that the resource corresponding to the index is the super-distribution resource, and providing a distribution reduction optimization strategy. The over-provisioning threshold for the CPU utilization max value is 40%. The over-provisioning threshold for the memory usage max value is 40%.
And (3) judging the reasonability of the payment mode of the database performance index: and judging whether the using time is less than or greater than the transition threshold, and providing a payment mode optimization strategy according to the judgment result. The transition threshold of the payment mode is 20 days, if the using time is longer than 20 days, the payment mode is suitable for covering a month and a year, and if the using time is shorter than 20 days, the payment mode is suitable for paying according to the amount.
Further, the optimizable cost ratio is:
optimizable cost rate (optimizable cost/total cost of cloud resources before implementing the cost utilization policy) x 100%;
wherein the optimizable cost is a difference between a total cost of the cloud resource before implementing the cost utilization policy and a total cost of the cloud resource after implementing the cost utilization policy.
Further, the evaluation function is:
score ═ (1-optimizable cost rate) × 100;
wherein, Score is cloud resource cost status Score.
Specifically, the higher the optimizable cost rate, the greater the degree to which the representation needs to be optimized, and the lower the cloud resource cost status score.
Further, fig. 2 is a hierarchical diagram of cloud resource utilization indicators of the present invention, and as shown in the figure, the quantization indicators include: evaluating field indexes and evaluating product indexes, wherein each evaluating field index corresponds to a plurality of evaluating product indexes; the evaluation field index comprises a calculation field index, a storage field index, a network field index and a database field index; the calculating of the evaluation product index corresponding to the field index comprises: cloud host indexes; the evaluation product index corresponding to the storage field index comprises: an object storage index and a file storage index; the evaluation product index corresponding to the network field index comprises: load balancing indexes and elastic public network IP indexes; the evaluation product indexes corresponding to the database field indexes comprise: database instance indices.
Further, the cloud host metrics include: CPU utilization rate, memory utilization rate, public network outgoing bandwidth utilization rate and disk utilization rate; the object storage metrics include: an effective request rate; the file storage index includes: a capacity usage rate; the load balancing indexes include: QPS rate and concurrent connectivity rate; the elastic public network IP indexes comprise: network ingress bandwidth utilization and network egress bandwidth utilization; the database instance metrics include: connection number usage, CPU usage, and memory usage.
Further, constructing a judgment matrix according to the evaluation criterion includes:
constructing judgment matrixes among the calculation field, the storage field, the network field and the database field, and fig. 3 is an evaluation field matrix table of the invention, wherein the matrixes constructed according to the table are as follows:
further, constructing a judgment matrix according to the evaluation criterion, further comprising:
a judgment matrix of the evaluation product index corresponding to the calculation field index is constructed, fig. 4 is a calculation field index matrix table of the invention, and the matrix constructed according to the table is as follows:
a judgment matrix of the evaluation product index corresponding to the storage field index is constructed, and fig. 5 is a storage field index matrix table of the present invention, wherein the matrix constructed according to the table is as follows:
a judgment matrix of the evaluation product index corresponding to the network field index is constructed, fig. 6 is a network field index matrix table of the present invention, and the matrix constructed according to the table is as follows:
a judgment matrix of the evaluation product index corresponding to the database field index is constructed, fig. 7 is a database field index matrix table of the invention, and the matrix constructed according to the table is as follows:
further, according to the weight vector of the judgment matrix, calculating a cloud resource utilization state score:
obtaining an initial score of the evaluation product index;
according to the weight vector (a, b, c, d) of the judgment matrixTCalculating the cloud resource utilization condition score by adopting the following formula:
S＝a×A+b×B+c×C+d×D；
wherein, in order to calculate the weight of the field index, A is the initial score of the field index, B is the weight of the stored field index, B is the initial score of the stored field index, C is the weight of the network field index, C is the initial score of the network field index, and d is the weight of the database field index. D is the initial score of the database index.
Specifically, the initial score of the evaluation product is the average score of the evaluation product over 24 hours. The process of obtaining the initial score is as follows: and acquiring the evaluation product index data every 30 minutes within 24 hours, and scoring based on forward processing of the evaluation product index data.
The indexes selected by the scheme are all interval type indexes, namely the index data is closer to a certain interval, and the index score is higher. For the convenience of subsequent calculation analysis, all indexes need to be converted into forward indexes. Taking the utilization rate of the cloud host CPU as an example: { xiThe CPU utilization rate sequence (i monitoring data in total) in one hour is defined as the interval with the best CPU utilization rate of the cloud host as default [ a, b ]]The score of the initial score is calculated by the formula
M＝max{xi},
The best interval range for index scoring is shown in FIG. 9, and the interval range is set by multiple experts.
Calculating an initial score A ═ a of the domain index1×A1+a2×A2+a3×A3+a4×A4(ii) a Wherein A is1，A2，A3，A4Calculating an initial score of an evaluation product index corresponding to the field index; a is1，a2，a3，a4The weight of the evaluation product index corresponding to the field index is calculated.
The initial score B ═ B of the storage area index1×B1+b2×B2+b3×B3+b4×B4(ii) a Wherein, B1，B2，b3，B4Storing an initial score of an evaluation product index corresponding to the field index; b1，b2，b3，b4And storing the weight of the evaluation product index corresponding to the field index.
An initial score C ═ C of the network realm indicator1×C1+c2×C2+c3×C3+c4×C4(ii) a Wherein, C1，C2，C3，C4An initial score of an evaluation product index corresponding to the network field index; c. C1，c2，c3，c4And the weight of the evaluation product index corresponding to the network field index.
An initial score D ═ D of the database domain index1×D1+d2×D2+d3×D3+d4×D4(ii) a Wherein D is1，D2，D3，D4The initial scores of the evaluation product indexes corresponding to the database field indexes are obtained; d1，d2，d3，d4And the weight of the evaluation product index corresponding to the database field index.
Further, by
Obtain a feature vector AAssessment field；
The feature vector AField of evaluationNormalizing to obtain
By passing
Calculating weight vector w ═ 1.88311.05210.61810.4467]T；
Normalizing the weight vector w to obtain
From feature vector AField of evaluationAnd normalized weight vector
To obtain
Will be provided with
Substituting the maximum characteristic root formula
Calculate out
Will be lambdamaxSubstitution of 4.0550
As can be seen from fig. 8, RI is 0.9.
Substituting CI and RI into formula
For matrix AField of evaluationPerforming consistency verification according to
It can be seen that the matrix consistency check passes.
Similarly, the consistency verification is performed on the matrix A, the matrix B, the matrix C and the matrix D, and the following calculation results are obtained:
matrix A:
λmax＝4.1117，CI＝0.0372，RI＝0.9，CR＝0.0414；
matrix B:
λmax＝2，CI＝0，RI＝0，CR＝0；
matrix C:
λmax＝4.0258，CI＝0.0086，RI＝0.9，CR＝0.0096；
matrix D:
λmax＝3.0069，CI＝0.0035，RI＝0.58，CR＝0.0060；
the CR values of all the above matrixes are less than or equal to 0.1, namely all the matrixes pass the consistency verification.
Further, let A, B, C, D represent the single initial score of the assessment domain respectively; order (a, b, c, d)TA normalized weight vector representing an assessment domain;
let A1,A2,A3,A4Representing an initial score of an evaluation product index corresponding to the calculation field index; (a)1,a2,a3,a4)TRepresenting a normalized weight vector relative to a computational domain;
let B1,B2An initial score representing an evaluation product index corresponding to the storage field index; order (b)1,b2)TRepresenting a normalized weight vector relative to a storage domain;
let C1,C2,C3,C4Representing an initial score of an evaluation product index corresponding to the network field index; order (c)1,c2,c3,c4)TRepresenting a normalized weight vector relative to a network domain;
let D1,D2,D3Representing an initial score of an evaluation product index corresponding to the database index; order (d)1,d2,d3)TRepresenting a normalized weight vector relative to a database;
the cloud resource utilization score is thus calculated by the following formula:
S＝a×A+b×B+c×C+d×D。
example 2
In the present embodiment, there is a case where the user lacks an evaluation product index. For example, a user generally purchases cloud products such as object storage, file storage, load balancing and the like according to needs, and index data related to the products cannot be acquired on the basis that the user does not purchase the products. Based on this, the weight of the model described in the above embodiment needs to be redistributed to the user's existing evaluation product index.
The specific operation method comprises the following steps: (1) setting the weight of the lacking evaluation product index as 0, and redistributing the weight of the evaluation product index corresponding to the same evaluation field index with the lacking evaluation product index (2) if the evaluation field index has no other evaluation product index, setting the weight of the evaluation field as 0, and redistributing the weights of the rest evaluation fields. For example:
at present, a user A only purchases three products of a cloud host, object storage and load balancing. (1) User A lacks indexes of file storage, elastic public network IP and database instance (namely capacity utilization rate and network inflow bandwidth utilization rate c)3Network egress bandwidth utilization c4Database connection number usage rate D1Database CPU utilization rate D2Memory usage rate D3) The weight is 0, and the effective request rate B of object storage under the storage field1Weight of b1Becomes 1; data packet rate c of load balancing in network field1Weight of (2)
Concurrent connection rate c20.5050; since the database field cannot be obtained, the database field index d is 0, and the weight of the field index a is calculated
b＝0.2961，c＝0.1739。
Therefore, the final score formula of the resource utilization condition of the user A is as follows:
example 3
In this embodiment, the cloud resource utilization condition evaluation step is tested by using 3 sets of index data of different resource utilization states (a partial resource low utilization state, a all-resource reasonable utilization state, and a partial resource overload utilization state). Each group of index data acquires monitoring data of resources subscribed by a user (only comprising cloud hosts, object storage, elastic public network IP and RDS-MySQL examples) within 24 hours as a calculation basis, data is acquired every 30 minutes, and the data aggregation mode is average (namely, a single index has 48 recorded values).
The selected indexes are interval indexes, that is, the index data is closer to a certain interval, and the index score is higher. Therefore, before the formal model test is performed, a professional operation and maintenance or a developer needs to establish a default optimal utilization interval of each index as a scoring basis. Fig. 9 shows the optimal interval for utilization of the unified indexes after evaluation by the technician. (note: the index only considers the general situation of resource usage by using the optimal interval standard, and the actual situation should be established according to the specific service condition.)
(1) Model application-low utilization data
According to fig. 13, since the indexes are all interval indexes, the indexes need to be converted into forward indexes, and the index scoring table after conversion is shown in fig. 10;
the utilization condition final score formula is:
(2) model application-all resource equitable utilization state data
According to fig. 14, since the indexes are interval indexes, the indexes need to be converted into forward indexes, and the index score table after conversion is shown in fig. 11;
this state is finally divided into:
S＝a×A+b×B+c×C+d×D＝a×(a1×A1+a2×A2+a3×A3+a4×A4)+b×(b1×B1+b2×B2)+c×(c1×C1+c2×C2+c3×C3+c4×C4)+d×(d1×D1+d2×D2+d3×D3＝0.4708×0.4587×A1+0.2975×A2+0.1004×A3+0.1435×A4+0.263×B1+0.1545×0.6310×C3+0.3690×C4+0.1117×0.3144×D1+0.357×D2+0.3286×D3＝0.4708×0.4587×98+0.2975×97+0.1004×97+0.1435×98+0.263×98+0.1545×0.6310×96+0.3690×95+0.1117×90.3144×72+0.357×98+0.3286×67)＝95.52。
(3) model application-partial resource overload usage
According to fig. 15, since the indexes are all interval indexes, the indexes need to be converted into forward indexes, and the index scoring table after conversion is shown in fig. 12;
this state is finally divided into:
S＝a×A+b×B+c×C+d×D＝a×(a1×A1+a2×A2+a3×A3+a4×A4)+b×(b1×B1+b2×B2)+c×(c1×C1+c2×C2+c3×C3+c4×C4)+d×(d1×D1+d2×D2+d3×D3＝0.4708×0.4587×A1+0.2975×A2+0.1004×A3+0.1435×A4+0.263×B1+0.1545×0.6310×C3+0.3690×C4+0.1117×0.3144×D1+0.357×D2+0.3286×D3＝0.4708×0.4587×68+0.2975×79+0.1004×94+0.1435×86+0.263×93+0.1545×0.6310×98+0.3690×93+0.1117×(0.3144×80+0.357×73+0.3286×89)＝84.46。
substituting the index result into a calculation formula to obtain the cloud resource utilization state scores of three resources in different utilization states: a partial resource low utilization state (77.25), an all resources fair utilization state (95.52), and a partial resource overload utilization state (84.46). Comparing the calculation results, when the resource index is in a reasonable use state, the score is higher than the scores of the low utilization state and the overload state, so that the resource utilization evaluation system established in the method is reasonable and effective.
It should be understood that the above-mentioned embodiments of the present invention are only examples for clearly illustrating the technical solutions of the present invention, and are not intended to limit the specific embodiments of the present invention. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention claims should be included in the protection scope of the present invention claims.
Claims (10)
1. A cloud resource cost analysis method is characterized by comprising the following steps: a cloud resource cost condition evaluation step and a cloud resource utilization condition evaluation step;
the cloud resource cost condition evaluation step comprises the following steps:
acquiring index data, and determining a cost utilization strategy to be implemented according to the index data; the index data is index data of an evaluation product; the cost utilization strategy is used for optimizing the cost of the evaluation product;
calculating an optimized cost rate according to the cloud resource total cost before the cost utilization strategy is implemented and the cloud resource total cost after the cost utilization strategy is implemented;
inputting the cost rate capable of being optimized into an evaluation function, and calculating the cloud resource cost condition score;
the cloud resource utilization condition evaluation step includes:
setting an evaluation criterion, wherein the evaluation criterion comprises a plurality of quantization indexes;
constructing a judgment matrix according to the evaluation criterion; the judgment matrix determines the weight value of each element of the judgment matrix through pairwise comparison between the quantitative indexes;
carrying out consistency check on the judgment matrix;
if the test fails, the judgment matrix needs to be reconstructed;
if the check is passed, calculating a cloud resource utilization condition score according to the weight vector of the judgment matrix and the initial score of the quantitative index;
and carrying out cloud resource cost evaluation according to the cloud resource cost condition score and the cloud resource utilization condition score.
2. The cloud resource cost analysis method according to claim 1, wherein the index data includes: calculating performance indexes, storing performance indexes, network performance indexes and database performance indexes.
3. The cloud resource cost analysis method of claim 1, wherein determining a cost utilization policy from the metric data comprises:
judging the index data according to a preset judgment condition to obtain a judgment result;
determining the cost utilization strategy according to the judgment result;
the determining includes: the method comprises the following steps of (1) predicting and judging idle resources, predicting and judging over-allocated resources, predicting and judging resource expansion and rationality of a payment mode;
the cost utilization strategy comprises the following steps: a descending allocation optimization strategy, an ascending allocation optimization strategy, a resource release optimization strategy and a payment mode optimization strategy.
4. The cloud resource cost analysis method according to claim 1, wherein the optimizable cost rate is:
the optimizable cost rate (optimizable cost/total cost of cloud resources before implementing the cost utilization policy) x 100%;
wherein the optimizable cost is a difference between a total cost of the cloud resource before implementing the cost utilization policy and a total cost of the cloud resource after implementing the cost utilization policy.
5. The cloud resource cost analysis method of claim 1, wherein the evaluation function is:
score ═ (1-optimizable cost rate) × 100;
wherein, Score is cloud resource cost status Score.
6. The cloud resource cost analysis method of claim 1, wherein the quantitative indicators comprise: evaluating field indexes and evaluating product indexes, wherein each evaluating field index corresponds to a plurality of evaluating product indexes; the evaluation field index comprises a calculation field index, a storage field index, a network field index and a database field index; the calculating of the evaluation product index corresponding to the field index comprises: cloud host indexes; the evaluation product index corresponding to the storage field index comprises: an object storage index and a file storage index; the evaluation product index corresponding to the network field index comprises: load balancing indexes and elastic public network IP indexes; the evaluation product indexes corresponding to the database field indexes comprise: database instance indices.
7. The cloud resource cost analysis method of claim 6, wherein the cloud host metrics comprise: CPU utilization rate, memory utilization rate, public network outgoing bandwidth utilization rate and disk utilization rate; the object storage metrics include: an effective request rate; the file storage index includes: a capacity usage rate; the load balancing indexes include: QPS rate and concurrent connectivity rate; the elastic public network IP indexes comprise: network ingress bandwidth utilization and network egress bandwidth utilization; the database instance metrics include: connection number usage, CPU usage, and memory usage.
8. The cloud resource cost analysis method of claim 7, wherein constructing a judgment matrix according to the evaluation criteria comprises:
constructing a judgment matrix among the computing field, the storage field, the network field and the database field:
9. the cloud resource cost analysis method of claim 8, wherein constructing a judgment matrix according to the evaluation criteria further comprises:
constructing a judgment matrix of the evaluation product index corresponding to the calculation field index:
constructing a judgment matrix of the evaluation product index corresponding to the storage field index:
constructing a judgment matrix of the evaluation product index corresponding to the network field index:
constructing a judgment matrix of the evaluation product index corresponding to the database field index:
10. the method according to claim 6, wherein a cloud resource utilization score is calculated according to the weight vector of the determination matrix and the initial score of the quantization index:
obtaining an initial score of the evaluation product index;
according to the weight vector (a, b, c, d) of the judgment matrixTCalculating the cloud resource utilization condition score by adopting the following formula:
S＝a×A+b×B+c×C+d×D；
wherein, in order to calculate the weight of the field index, A is the initial score of the field index, B is the weight of the stored field index, B is the initial score of the stored field index, C is the weight of the network field index, C is the initial score of the network field index, and d is the weight of the database field index. D is the initial score of the database index.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210315810.0A CN114629909A (en) | 2022-03-28 | 2022-03-28 | Cloud resource cost analysis method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210315810.0A CN114629909A (en) | 2022-03-28 | 2022-03-28 | Cloud resource cost analysis method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114629909A true CN114629909A (en) | 2022-06-14 |
Family
ID=81904339
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210315810.0A Pending CN114629909A (en) | 2022-03-28 | 2022-03-28 | Cloud resource cost analysis method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114629909A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115118731A (en) * | 2022-06-24 | 2022-09-27 | 中国建设银行股份有限公司 | Cloud resource management method and device, storage medium and electronic equipment |
| CN115460082A (en) * | 2022-08-25 | 2022-12-09 | 浪潮云信息技术股份公司 | Cloud cost optimization method and system based on government affair cloud scene |
-
2022
- 2022-03-28 CN CN202210315810.0A patent/CN114629909A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115118731A (en) * | 2022-06-24 | 2022-09-27 | 中国建设银行股份有限公司 | Cloud resource management method and device, storage medium and electronic equipment |
| CN115118731B (en) * | 2022-06-24 | 2024-02-27 | 中国建设银行股份有限公司 | Cloud resource management method and device, storage medium and electronic equipment |
| CN115460082A (en) * | 2022-08-25 | 2022-12-09 | 浪潮云信息技术股份公司 | Cloud cost optimization method and system based on government affair cloud scene |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10896055B2 (en) | Capacity risk management for virtual machines | |
| US8745216B2 (en) | Systems and methods for monitoring and controlling a service level agreement | |
| CN106020715B (en) | Storage pool capacity management | |
| WO2020119051A1 (en) | Cloud platform resource usage prediction method and terminal device | |
| CN114629909A (en) | Cloud resource cost analysis method | |
| US20160098297A1 (en) | System and Method for Determining Capacity in Computer Environments Using Demand Profiles | |
| JP5162579B2 (en) | Deploy virtual machines to hosts based on workload characteristics | |
| CN106233276B (en) | The coordination admission control of network-accessible block storage device | |
| US8725741B2 (en) | Assessing application performance with an operational index | |
| Mokadem et al. | A data replication strategy with tenant performance and provider economic profit guarantees in Cloud data centers | |
| US20100125715A1 (en) | Storage System and Operation Method Thereof | |
| US10970123B1 (en) | Determining suitability of computing resources for workloads | |
| US11307885B1 (en) | Identifying optimized computing resources for running workloads | |
| CN111695830A (en) | Power resource allocation method, system and equipment | |
| US20170139754A1 (en) | A mechanism for controled server overallocation in a datacenter | |
| WO2021180056A1 (en) | Method for resource migration, system and device | |
| CN107277143A (en) | A kind of resource matched management method and device | |
| US20210035115A1 (en) | Method and system for provisioning software licenses | |
| US9305068B1 (en) | Methods and apparatus for database virtualization | |
| Alam et al. | Multi-objective interdependent VM placement model based on cloud reliability evaluation | |
| CN114741160A (en) | Dynamic virtual machine integration method and system based on balanced energy consumption and service quality | |
| CN108279968A (en) | A kind of dispatching method and device of resources of virtual machine | |
| Villalpando et al. | A three-dimensional performance measurement model for cloud computing | |
| Jia et al. | Perceptual Forecasting Model of Power Big Data Based on Improved Random Forest Algorithm | |
| US20230122363A1 (en) | Storage allocation based on application priority specific to a select entity |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20220614 |
|
| RJ01 | Rejection of invention patent application after publication |