Upgrade virtual appliance deployment based on peak utilization
By collecting and analyzing usage data of virtual devices through the virtual service upgrade system, and dynamically adjusting resource allocation, the problem of uneven resource allocation in cloud computing systems is solved, thereby improving resource utilization efficiency and reducing costs.
Patent Information
- Application Number
- CN202280054854.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2021-12-06
- Filing Date
- 2022-07-26
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2042-07-26
AI Technical Summary
The uneven allocation of virtual devices in cloud computing systems leads to high computing costs, resource waste, and user dissatisfaction, making it difficult to efficiently adjust resource allocation dynamically according to individual needs.
The virtual service upgrade system collects and analyzes usage data of virtual devices to determine the current usage score, and dynamically upgrades or downgrades the resource configuration of virtual devices based on the score, thereby achieving flexible resource adjustment.
It improves the utilization efficiency of cloud computing resources, reduces data center power consumption and tenant costs, dynamically adapts to individual usage needs, and reduces resource waste.
Smart Images

Figure CN117795480B_ABST
Abstract
Description
Background Technology
[0001] In recent years, the use of computing devices (such as mobile devices, personal computers, and server devices) to perform various applications and tasks has increased dramatically. For example, many individuals and businesses have begun to use cloud computing resources to run a wide variety of applications and tasks. In fact, in order to take advantage of the increased computing, storage, and memory capacity offered by the cloud, many enterprises are transitioning from using expensive on-premises machines and services to virtual machines deployed in the cloud.
[0002] As the demand for cloud computing resources (e.g., storage, memory, and compute resources) continues to grow in scale and complexity, cloud computing systems have grown in size and complexity to meet that demand. However, providing sufficient resources efficiently presents a significant challenge because different tenants and individual users have different individual needs. For example, many deployments suffer from over-allocation of compute resources due to some individuals being associated with light or infrequent use of cloud resources, resulting in significant power consumption and unused storage capacity. This allocation of unused compute resources leads to higher computing costs for both the cloud provider and the tenant. Conversely, other individuals associated with heavy or frequent use of cloud resources often experience frustration due to insufficient resource allocation. Furthermore, enterprises or tenants with hundreds or thousands of individuals, each with unique usage profiles, are often associated with deployments that inefficiently allocate cloud resources to them.
[0003] Regarding the deployment of virtual devices and other services on cloud computing systems, as well as the scaling of such deployments, there are these and other issues. Attached Figure Description
[0004] Figure 1 An example environment including a virtual service upgrade system according to one or more embodiments is shown.
[0005] Figure 2 An example workflow according to one or more embodiments is shown, in which a virtual service upgrade system collects and evaluates usage data of virtual devices.
[0006] Figure 3A An example series of actions related to collecting and aggregating usage data of virtual devices, according to one or more embodiments, are shown.
[0007] Figure 3B An example series of actions related to analyzing and characterizing usage data of virtual devices, according to one or more embodiments, are shown.
[0008] Figure 3C An example series of actions related to determining the current usage score of a virtual device, according to one or more embodiments, are shown.
[0009] Figure 4 An example implementation of a bucket used by a virtual service upgrade system according to one or more embodiments is shown.
[0010] Figure 5 An example workflow for calculating the current usage score of a virtual device, according to one or more embodiments, is shown.
[0011] Figure 6 An example series of actions is shown, according to one or more embodiments, for facilitating the upgrade of virtual devices based on current usage scores.
[0012] Figure 7 This shows some components that can be included within a computer system. Detailed Implementation
[0013] This disclosure relates to a virtual service upgrade system configured to collect or otherwise receive usage data from virtual devices (or other virtual services), assess the unique characteristics of the usage data, and determine a current usage score that provides actionable information to facilitate the upgrade (or downgrade) of the virtual device. Specifically, as will be discussed in further detail below, the virtual service upgrade system may employ an agent on the virtual device to collect usage data at multiple time intervals (e.g., one second, five seconds, ten seconds, etc.) spanning a predetermined time period (e.g., two weeks). The virtual service upgrade system may additionally categorize virtual devices into corresponding buckets or categories and determine a current usage score based on this categorization, reflecting the extent to which the virtual device has been used in a recent period.
[0014] This determination can be used to implement or otherwise initiate the upgrading or downgrading of a virtual device to a virtual device with a different set of resource specifications. As discussed below, such upgrades / downgrading can be performed dynamically without data loss or otherwise interrupting services provided to users via the virtual device(s). In fact, in one or more embodiments, the upgrading or downgrading of a virtual device can be performed in a manner unknown to the end user or otherwise opaque.
[0015] As an illustrative example, and according to one or more embodiments, a virtual service upgrade system may receive usage data of a virtual device that has already been sampled by an agent on the virtual device or otherwise obtained. The virtual service upgrade system may aggregate the data, determine various usage metrics, and classify the virtual devices into usage buckets associated with a range of usage scores. In one or more embodiments, the virtual service upgrade system classifies the virtual devices by determining a spike metric based on the number of time intervals (e.g., zero, one, or more time intervals) in which usage data exceeds a threshold within a predetermined time period. The virtual service upgrade system may further classify the virtual devices by determining usage buckets associated with a range of usage scores. The virtual service upgrade system may additionally generate a current usage score within the predetermined time period, partly based on the assigned buckets. In one or more embodiments, the virtual service upgrade system may upgrade a virtual device based on a current usage score within the predetermined time period that is less than a threshold usage score.
[0016] This disclosure includes many practical applications that provide benefits and / or efficiently address issues related to evaluating the use and deployment of virtual devices (and other virtual services). Examples of some of these benefits will be discussed in further detail below.
[0017] For example, in one or more embodiments described herein, a virtual service upgrade system provides flexibility in the deployment of virtual devices, enabling modifications to virtual devices after deployment and based on observed changes and trends in virtual device usage over time. This is particularly helpful for deployments involving hundreds or thousands of individual virtual devices associated with individuals within an organization that have unique usage profiles within a user group. For instance, a virtual service upgrade system can facilitate the deployment of a virtual device associated with a first virtual machine type (e.g., a first VM family, VM stock keeping unit (SKU)) on the cloud for a first individual associated with heavy use of the virtual device. Alternatively, the virtual service upgrade system can deploy a virtual device associated with a second virtual machine type on the cloud for a second individual associated with light or infrequent use of the virtual device. These deployments can be implemented dynamically, with virtual devices being modified (e.g., upgraded and / or downgraded) over the air based on the current usage profile.
[0018] Beyond providing dynamic and flexible solutions for virtual machines with different usage profiles, virtual service upgrade systems can further deliver more efficient utilization of cloud computing resources than conventional deployments. For example, by dynamically upgrading and / or downgrading virtual appliances for various users deployed across multi-user tenants, virtual service upgrade systems can reduce power consumption caused by the over-allocation of cloud computing resources to individuals and organizations with peak capacity that is not utilizing those resources. As mentioned above, this can significantly reduce power consumption at the cloud data center and further free up substantial cloud computing resources for other deployments. It can also provide capacity at the data center to accommodate sudden or gradual increases in resource usage by one or more tenants.
[0019] Beyond the flexibility and resilience typically offered by scaling cloud computing resources, the virtual service upgrade system described in this paper features and functionalities that enable computationally efficient collection and / or analysis of usage data compared to other data collection and analysis systems. For example, where conventional systems typically collect vast amounts of data for each computing device from a large set of devices, the system described provides an intelligent approach to selectively analyze usage data and determine usage scores for subsets of devices that perform in a particular manner. For instance, the virtual service upgrade system can selectively analyze the usage data and usage scores of deployed devices based on the fact that a subset of devices are above a minimum threshold. Furthermore, even when a specific device that meets a minimum usage threshold is identified as a machine of interest, the virtual service upgrade system can still quantify the usage profile based on the number of intervals in which that device meets one or more usage thresholds.
[0020] As a more specific example, and as will be discussed in further detail herein, a virtual service upgrade system can further provide usage scores without analyzing every bit of the usage data. For example, in one or more embodiments, the virtual service upgrade system can determine various metrics and corresponding usage scores based on a count of time intervals in which agents on virtual devices observe certain conditions. This incremental approach allows the virtual service upgrade system to discard data associated with inactive time intervals. This unique approach also enables the virtual service upgrade system to determine and provide actionable intelligence without consuming significant storage, memory, and processing resources of the management system (e.g., on-premises and / or cloud-based systems). This approach further enables the virtual service upgrade system to determine and update current usage scores at frequent intervals, providing up-to-date and current profiles of resource utilization for any number of virtual devices.
[0021] Each of the benefits mentioned above can significantly reduce costs for both cloud providers and tenants. For example, by modifying the deployment of virtual appliances based on different usage profiles, tenants can save substantial costs by downgrading subscriptions for one or more individuals within their tenant accounts. Furthermore, by downgrading and / or upgrading virtual appliances to align with an individual's unique usage profile, cloud providers can reduce the costs associated with delivering power to higher-capacity server nodes, which represent an over-allocation of cloud resources for a particular individual. For instance, virtual service upgrade systems enable cloud providers to prevent value loss due to the over-allocation of unused compute cores and their loss of value over time. Moreover, by intelligently collecting and analyzing usage data, virtual service upgrade systems deliver the above services without incurring the significant resources required to analyze usage data for each appliance over long periods.
[0022] As will be discussed in further detail below, the virtual service upgrade system can additionally provide actionable intelligence to the administrators of the deployment, enabling them to upgrade and / or downgrade deployments based on the usage score identified herein. For example, as will be discussed here, the virtual service upgrade system can determine a current usage score representing CPU usage, memory usage, I / O usage, or other resource metrics over a predetermined period and provide indications for one or more virtual devices that can be further analyzed before an upgrade or downgrade decision is made. As will be discussed in further detail below, this usage score can be based on a combination of peak usage and usage time.
[0023] As explained in the foregoing discussion, this disclosure uses various terms to describe the features and advantages of the virtual service upgrade system. Additional details regarding the meaning of some of these terms will now be provided.
[0024] For example, as used herein, "cloud computing system" refers to a network of connected computing devices that provide various services to client devices (e.g., client equipment, network equipment). For instance, a distributed computing system may include a collection of physical server devices (e.g., server nodes) organized in a hierarchical structure, including clusters, computing zones, virtual local area networks (VLANs), racks, fault-tolerant domains, etc. Cloud computing system can refer to either a private or public cloud computing system.
[0025] As used herein, "virtual device" refers to an emulation of a computer system on a server node that provides functionality for one or more applications on a cloud computing system. In one or more embodiments, a virtual device refers to an emulation of a computing device associated with an individual account or individual user. A virtual device may refer to one of multiple virtual devices or other services deployed by a tenant. It should be understood that while one or more specific examples and implementations described herein specifically relate to virtual devices, the characteristics and functions described in conjunction with virtual devices can similarly refer to other types of virtual machines or computing containers configured to provide functionality for a wide variety of cloud computing applications.
[0026] As used herein, “usage data” or “utilization data” can be used interchangeably to refer to any information associated with the utilization of cloud computing resources on a cloud computing system. For example, usage data can refer to the use of storage resources, processing resources, and / or input / output (I / O) resources on one or more server nodes. In one or more embodiments, usage data specifically refers to the utilization of one or more cloud computing resources for a particular virtual appliance. In one or more embodiments described herein, usage data can refer to telemetry sampled, captured, or otherwise obtained by an agent of the virtual appliance, which has been configured to sample usage data at periodic time intervals. For example, usage data can refer to sampled snapshots showing the utilization of various cloud computing resources at periodic points in time. Usage data can refer to raw and / or refined or extrapolated data that provides a description or status of resource utilization for a relevant time period.
[0027] As used herein, “usage score” can refer to a value associated with a virtual device based on usage data. As will be discussed in further detail below, a usage score can be determined based on various usage metrics calculated from usage data over a period of time (e.g., two weeks). For example, a usage score can be determined based on a peak-level metric associated with the ratio of resource usage to resource capacity. In one or more embodiments, a usage score is determined based on a time-based usage metric associated with the frequency of use over a period of time.
[0028] In one or more embodiments described herein, usage data can be acquired over multiple time intervals and within a predetermined time period. As used herein, "time interval" refers to a period of time during which a virtual device collects or samples usage data. For example, a time interval can refer to a one-second interval, a five-second interval, a ten-second interval, or other relatively short duration for acquiring one or more types of usage data. In one or more embodiments, a time interval refers to a sampling interval during which an agent on the virtual device samples or accumulates data within a time interval. For example, where a time interval refers to a ten-second duration, the agent can capture the state of one or more types of usage data within a ten-second interval, or in some cases, can accumulate data captured at more granular moments within a ten-second interval.
[0029] As used herein, a “predetermined time period” can refer to any length of time that includes multiple time intervals. For example, in one or more embodiments described herein, a predetermined time period can refer to a day, a week, two weeks, a month, or other duration. As will be discussed in further detail below, usage data collected, received, or otherwise obtained over multiple time intervals can be accumulated over a predetermined time period, such as a two-week period starting from the time when a peak level metric (or other usage metric) is calculated for a virtual device.
[0030] Additional details regarding the virtual service upgrade system, along with illustrative figures depicting an example implementation, will now be provided. For example, Figure 1 This illustrates an example environment showcasing a virtual service upgrade system implemented on a cloud computing system. For example... Figure 1 As shown, environment 100 includes a cloud computing system 102 having at least one device thereon. For example, cloud computing system 102 may include one or more server devices 104 having a virtual service upgrade system 106 implemented thereon. Virtual service upgrade system 106 may include a data collection manager 108, a classification manager 110, a scoring engine 112, and an upgrade manager 114. Additional information relating to these components will be discussed below.
[0031] In addition to server device 104, the cloud computing system 102 may also include any number of server nodes 116a to 116n. For example... Figure 1 As shown, server nodes 116a to 116n may include virtual devices 118a to 118n (or other virtual machines) deployed thereon. Each virtual device among virtual devices 118a to 118n may include agents 120a to 120n and operating systems (OS) 122a to 122n. For example, as Figure 1As shown, the cloud computing system 102 may include a first server node 116a hosting one or more first virtual devices 118a. As further shown, the cloud computing system 102 may include additional server nodes 116b to 116n, each server node hosting one or more virtual devices 118b to 118n.
[0032] like Figure 1 As shown, environment 100 may include multiple client devices 124 communicating with cloud computing system 102 via network 126 (e.g., communicating with different server nodes). Client devices 124 may refer to various types of computing devices, including mobile devices, desktop computers, server devices, or other types of computing devices. Network 126 may include one or more networks using one or more communication platforms or technologies for transmitting data. For example, network 126 may include the Internet or other data links that enable the transmission of electronic data between the respective client devices 124 and devices in cloud computing system 102.
[0033] like Figure 1 As shown and as will be discussed in further detail herein, the virtual service upgrade system 106 may be a component that provides the various features and functions described herein. Although Figure 1 Examples of each component of the Virtual Service Upgrade System 106 implemented on a single server device are shown; however, it should be understood that one or more of the corresponding components may be implemented on additional server devices. Furthermore, it should be understood that the discrete features and functions described herein with respect to one or more of the corresponding components may be similarly implemented on the same server device or across multiple server devices. Moreover, while one or more features described herein may be discussed in conjunction with specific components, it should be understood that describing features and functions in conjunction with specific components does not preclude the performance of such features or functions by different components of the Virtual Service Upgrade System 106.
[0034] Figure 2 An example workflow 200 illustrating an example implementation of a virtual service upgrade system 106 according to one or more embodiments of this document is shown. Specifically, Figure 2 A workflow 200 according to one or more embodiments is illustrated, which includes components of an agent 204 on an example virtual appliance 202 and a virtual service upgrade system 106 related to collecting usage data, analyzing the usage data to generate a usage score, and generating update instructions for the virtual appliance 202. Similar workflows can be applied to any number of virtual appliances, such as a collection of hundreds or thousands of virtual appliances associated with a particular organization (e.g., a tenant deployment).
[0035] like Figure 2As shown, agent 204 of virtual device 202 can sample usage data of virtual device 202. In one or more embodiments, agent 204 samples usage data at regular intervals. For example, agent 204 can refer to a software agent configured to sample usage data of virtual device 202 at fixed intervals of several seconds. In one or more embodiments, agent 204 is configured to sample usage data every ten seconds (or other fixed time intervals).
[0036] According to one or more embodiments, agent 204 can be configured to sample different types of usage data. For example, in one or more embodiments, agent 204 samples CPU data indicating metrics of processing usage at regular time intervals. As another example, in one or more embodiments, agent 204 samples memory data indicating the amount of memory used by virtual device 202 at regular time intervals. As another example, agent 204 may sample I / O metrics indicating the I / O rate of virtual device 202 at regular time intervals. In one or more embodiments, agent 204 captures or otherwise samples usage data that includes combinations of different types of usage data.
[0037] like Figure 2 As shown, agent 204 can provide usage data to data collection manager 108 for further processing. Upon receiving the usage data, data collection manager 108 can process the usage data in various ways. For example, in one or more embodiments, data collection manager 108 can aggregate or compile the usage data over various time intervals greater than the sampling interval of agent 204. In this way, data collection manager 108 can generate interval snapshots and / or summaries of usage data that provide useful data over various time periods. For example, in one or more embodiments, data collection manager 108 aggregates data that has been sampled at ten-second intervals into a fifteen-minute summary. In one or more embodiments, data collection manager 108 generates hourly, daily, and / or weekly summaries that further provide a depiction of the utilization of virtual devices(s)102.
[0038] like Figure 2 As shown, the data collection manager 108 can provide usage summary data to the classification manager 110 for further processing. Specifically, as will be discussed in further detail below, the classification manager 110 can process the received usage data (e.g., using summary data) to determine various usage metrics. These usage metrics can provide usage characteristics based on the usage of the (multiple) virtual devices 202 over time.
[0039] As described above, the classification manager 110 can determine various usage metrics associated with different usage characteristics of the virtual device 202. For example, in one or more embodiments, the classification manager 110 can determine a peak level metric for the virtual device 202. As used herein, a peak level metric can refer to a usage profile in which the virtual device 202 utilizes cloud computing resources exceeding a threshold metric within a time interval. For example, in one or more embodiments, the virtual device 202 may have a peak level metric indicating that the virtual device 202 utilizes more than 75% of one or more cloud computing resources within a relevant time period.
[0040] Other implementations may include different (or additional) threshold percentages. For example, in one or more embodiments, the classification manager 110 may determine a first peak level metric indicating whether the virtual device 202 uses more than 50%, 75%, and / or 90% of cloud computing resources within a specific time interval. In one or more embodiments, the classification manager 110 determines peak metrics associated with different percentages and / or thresholds based on the unique application to which the virtual device 202 is configured to run. Furthermore, as described above, the classification manager 110 may determine a peak level metric for each type of usage data (e.g., usage data corresponding to different types of cloud computing resources).
[0041] As described above, the classification manager 110 can receive usage profile data that includes information associated with different time intervals. In one or more embodiments, the classification manager 110 can determine a peak level metric (or other usage metric) within a time interval in which the data collection manager 108 has aggregated data samples acquired within that time interval. As an illustrative example, the classification manager 108 can determine a peak level metric for each time interval (e.g., every fifteen-minute interval) that indicates that the virtual device 202 consumed at least a threshold amount of cloud computing resources within a specific time interval (e.g., on average).
[0042] In addition to peak-level metrics, in one or more embodiments, the classification manager 110 may determine a time usage metric for the virtual device 202. As used herein, a time usage metric may refer to a usage profile in which the virtual device 202 utilizes a minimum threshold of cloud computing resources within a time interval. For example, in one or more embodiments, the virtual device 202 may have a time usage metric indicating that the virtual device 202 utilizes a non-zero amount of cloud computing resources within a relevant time period (e.g., beyond simply running an OS or other basic-level applications).
[0043] Similar to a peak-level metric, the classification manager 110 can determine a time usage metric for each of a plurality of time intervals. For example, the classification manager 110 can simply determine a binary one (1) or zero (0) within each fifteen-minute time interval. In one or more embodiments, the classification manager 110 can determine a more specific usage metric indicating any number of sampling intervals (e.g., corresponding to the rate at which the agent samples usage data) within which the user of the virtual device 202 is utilizing cloud computing resources.
[0044] In one or more embodiments, the category manager 110 can implement a workflow in which time usage metrics and peak level metrics are determined as part of a combined process. For example, in one or more embodiments, the category manager 110 can first determine whether the time usage metric indicates that the virtual device 202 has been used more than a minimum amount (e.g., a non-zero amount or other minimum amount of use) within a relevant time interval. If the time usage metric indicates that the virtual device 202 has not been used within that time interval, the category manager 110 can automatically determine that the peak level metric is also zero within the same time interval. Alternatively, if the time usage metric indicates that the virtual device 202 has been used by a certain amount during the relevant time interval, the category manager 110 can continue to determine the peak level metric for that time interval. The following will be combined with... Figure 3B Discuss additional information related to this process.
[0045] As described above, the category manager 110 can determine usage metrics for each time interval within a predetermined time period. For example, in one or more embodiments, the category manager 110 can determine usage metrics for fifteen-second time intervals (or other time interval lengths) over multiple days and / or weeks. In one or more implementations described herein, the category manager 110 determines usage metrics over a two-week time period.
[0046] like Figure 2 As shown, the category manager 110 can provide usage metric data to the rating engine 112 for further processing. Upon receiving the usage metrics(s), the rating engine 112 can determine the usage score of the virtual device 202 based at least on the peak level metric and one or more additional usage metrics (e.g., time usage metric). As will be discussed in further detail below, the rating engine 112 can determine the usage score of the virtual device 202 within a predetermined time period in various ways.
[0047] For example, in one or more embodiments, the rating engine 112 may assign or otherwise associate the virtual device 202 with a usage bucket based on one or more usage metrics. As used herein, a usage bucket may refer to a category or classification of use corresponding to a usage level of a given virtual device. For example, the rating engine 112 may associate the virtual device 202 with one of a plurality of usage buckets associated with increasing usage levels (e.g., high usage, medium usage, low usage). As will be discussed in one or more embodiments herein, the rating engine may assign the virtual device 202 to a specific usage bucket from any number of possible usage buckets corresponding to different usage levels.
[0048] In one or more embodiments, the rating engine 112 assigns the virtual device 202 to a corresponding usage bucket based on a combination of peak usage metrics and time usage metrics. For example, the rating engine 112 may provide peak usage metrics for a predetermined time period and time usage metrics for a predetermined time period as input to a model that is trained to classify or otherwise categorize the combined usage metrics into the corresponding usage buckets.
[0049] In one or more embodiments, the scoring engine 112 determines usage buckets based on a piecewise linear model in which corresponding values are assigned to peak level metrics and / or time usage metrics based on a discrete range of usage metrics provided by the category manager 110. For example, the scoring engine 112 may determine a first peak usage metric, a second peak usage metric, or a third peak usage metric based on a comparison of the number of time intervals in which the virtual device exceeds a threshold (e.g., zero, one, or more time intervals). (E.g., where the usage summary of the corresponding time interval indicates a resource usage level above the peak threshold.) As another example, the scoring engine 112 may determine a first time usage metric, a second time usage metric, or a third time usage metric based on a comparison of the number of time intervals in which the virtual device 202 is used for any amount of time (e.g., a non-zero amount of time or greater than a certain minimum threshold).
[0050] For example, rather than restricting it, the scoring engine 112 can implement a piecewise linear model based on the following equation, where the peak level is determined according to the following threshold:
[0051]
[0052] Here, "x" represents the percentage of time intervals that meet the threshold resource utilization. As an explanation of the equation above, scoring engine 112 can determine that if virtual device 202 consumes cloud computing resources at a level above the threshold for less than 10% of time intervals (e.g., in 15-minute increments) within a predetermined time period (e.g., 2 weeks), scoring engine 112 can assign a "low" score or classification to virtual device 202 within the predetermined time period. Similarly, scoring engine 112 can assign a "medium" score or classification if it finds that virtual device 202 consumes cloud computing resources for between 10% and 25% of time intervals within the predetermined time period. Scoring engine 112 can assign a "high" score or classification if it finds that virtual device 202 consumes cloud computing resources for 25% or more of the time intervals within the predetermined time period.
[0053] Similarly, the rating engine 112 can determine the corresponding time usage metric based on the following thresholds:
[0054]
[0055] Here, "x" represents the number of minutes the virtual device was used exceeding a minimum threshold (regardless of whether a peak usage threshold was met). As an explanation of the equation above, the scoring engine 112 can assign a "low" score or category if the virtual device is used for less than 120 minutes in a predetermined time period (e.g., two weeks). The scoring engine can assign a "medium" score or category if the virtual device is used for between 120 and 600 minutes in a predetermined time period. The scoring engine 112 can assign a "high" score or category if the virtual device 202 is used for 600 minutes or more in a predetermined time period.
[0056] It should be understood that other implementations may include different thresholds for the corresponding usage metrics. Furthermore, other implementations may incorporate different predetermined time periods and / or different time intervals into determining the corresponding usage metric values.
[0057] As will be discussed in further detail below, each usage bucket in a usage bucket can have a corresponding range of values. To illustrate, in the example discussed above of assigning virtual devices to one of eight different categories (e.g., a range of x = 0 to x = 8), the range of values could be as follows:
[0058]
[0059]
[0060] It should be understood that other implementations may have different value ranges. Furthermore, the following will combine... Figure 4Additional details relating to one or more embodiments are discussed.
[0061] As described above, the scoring engine 112 can assign the virtual device 202 to one of a plurality of usage buckets based on one or more usage metrics. For example, in one or more embodiments, the scoring engine 112 assigns the virtual device 202 to a corresponding usage bucket based on a determined score or category assigned to the virtual device 202 according to the piecewise linear function described above. As a non-limiting example considering peak rating metrics and time usage metrics, the scoring engine 112 can assign the virtual device to one of eight categories based on a combination of the corresponding metrics. The category within the corresponding usage bucket can be determined as follows:
[0062]
[0063] Where “x” refers to a low, medium, or high category based on a peak ranking metric, and “y” refers to a low, medium, or high category based on a time usage metric.
[0064] In addition to determining the peak usage metrics(s), the scoring engine 112 can also determine the upper and lower bound functions for the corresponding usage metrics. For example, the upper bound of peak usage can be determined using the following equation:
[0065]
[0066] Conversely, the following equation can be used to determine the lower limit of peak usage:
[0067]
[0068] Each of the formulas above can be used to calculate the lower and upper limits of the peak percentage level, respectively. The following will combine... Figure 4 Let's discuss additional details.
[0069] Following a similar approach, the upper limit of time usage can be determined using the following equation:
[0070]
[0071] Conversely, the lower limit of time usage can be determined using the following equation:
[0072]
[0073] Each of the formulas above can be used to calculate the lower and upper limits of the time usage level. The following will combine... Figure 4 Let's discuss additional details.
[0074] After determining various bucket limits and categorizing virtual devices into one of multiple usage buckets, the scoring engine 112 can perform score adjustment to determine a more precise usage score. This adjusted score can be determined based on the score intervals of the identified usage buckets. Furthermore, the scoring engine 112 can consider some of the finer-grained usage data aggregated by the data collection manager 108 to determine a more specific value within the bucket value range. For example, the scoring engine 112 can use the following formula:
[0075]
[0076] b=CeilBucket(Bucket(SpikeLevel(x),UsageLevel(y)))
[0077] a=FloorBucket(Bucket(SpikeLevel(x),UsageLevel(y)))
[0078] Where “a” and “b” represent the upper and lower limits of the allocated bucket, “x” refers to the specific peak usage metric for the predetermined time period, and “y” refers to the specific time usage metric for the predetermined time period.
[0079] like Figure 2 As shown, the scoring engine 112 can output a usage score based on one or more of peak usage metrics, time usage metrics, and the associated usage buckets to which the virtual device 202 has been assigned. According to one or more embodiments described herein, the scoring engine 112 can output a higher score for healthier virtual devices that are not approaching peak usage capacity (e.g., indicating low or medium usage). Conversely, the scoring engine 112 can output a lower score for less healthy virtual devices that are approaching peak usage capacity (e.g., indicating medium or high usage). Figure 2 As shown, the rating engine 112 can provide the upgrade manager 114 with the current usage score for further processing.
[0080] As used herein, the current usage score may include a usage score determined based on the assigned usage bucket and / or various usage metrics dating back to the most recent period from the start of a predetermined time period. Therefore, where the predetermined time period refers to a two-week period, the current usage score may refer to the most recent two-week period. It should be understood that the scoring engine 112 can dynamically update the usage score over time as additional usage data is sampled, aggregated, and further analyzed to determine the updated usage score.
[0081] like Figure 2As shown, the upgrade manager 114 can generate an update instruction based on the received usage score. For example, the upgrade manager 114 can generate an instruction to upgrade the virtual device 202 if the usage score is less than a threshold usage score indicating that the virtual device 202 is operating at near-peak capacity. In one or more embodiments, the upgrade manager 114 can determine to upgrade the virtual device 202 based on the usage score being less than a threshold usage score within a predefined time period. Conversely, if the usage score is equal to or higher than the threshold, the upgrade manager 114 may not take any further action regarding upgrading the virtual device 202.
[0082] While one or more embodiments described herein relate to causing virtual device 202 to initiate an upgrade process, in one or more embodiments, upgrade manager 114 may determine that virtual device 202 should be downgraded. For example, if a usage score is higher than a threshold score indicating infrequent or light use of virtual device 202, upgrade manager 114 may generate an update instruction to downgrade virtual device 202 to a cheaper version or virtual device type with fewer cloud computing resources allocated to it. For example, upgrade manager 114 may downgrade virtual device 202 to a version of virtual device 202 with fewer compute cores, less memory, and / or fewer I / O ports allocated to it.
[0083] Upgrading and / or downgrading virtual appliance 202 can refer to various processes. For example, in one or more embodiments, upgrade manager 114 causes a first type of virtual appliance to be upgraded to a second type of virtual appliance with a higher number of cloud computing resources. For example, upgrade manager 114 may cause a virtual appliance with a first set of resource specifications (e.g., memory resources, processing resources) to be upgraded (or downgraded) to a virtual appliance with a second set of resource specifications. This may involve assigning or otherwise reconfiguring virtual appliances to different virtual machine families pre-configured for hosting by a cloud computing system (e.g., cloud computing system 102) (e.g., assigning virtual appliances with a first set of discrete or pre-configured resource specifications to different virtual appliance types with a second set of discrete or pre-configured resource specifications). In one or more embodiments, upgrade manager 114 simply adds additional memory and / or processors to the deployment of virtual appliance 202 (e.g., selectively modifying the memory capacity and / or processing capacity of the virtual appliance).
[0084] As used herein, “device type” can refer to any combination of virtual devices having associated cloud computing resources. For example, a device type can be associated with a specific virtual machine family in which the deployment of virtual devices of that device type is allocated a corresponding number of compute resources, memory resources, I / O resources, storage resources, or other cloud computing resources. In one or more embodiments, the cloud computing system may provide a list of dozens or hundreds of device types corresponding to different combinations of allocable cloud computing resources, which may be offered to tenants or otherwise included within the deployment of VMs.
[0085] In each of the examples above, the upgrade manager 114 can initiate upgrades, downgrades, or other modifications to the virtual device deployment while maintaining the virtual device's data. This may involve migrating data from one virtual device type to another in real time, or migrating data from one server node to another (e.g., on the same or different clusters). In one or more embodiments, this may involve initiating virtual device migrations during periods when the virtual device is not in use or during a scheduled update process.
[0086] In one or more embodiments, performing an upgrade involves automatically initiating an upgrade or downgrade based on a deployment or upgrade policy (e.g., one established by the administrator of the tenant deployment). For example, in one or more embodiments, the upgrade manager 114 may automatically upgrade virtual appliance 202 based on its usage score being lower than the usage score for a predetermined time period. Alternatively, the upgrade manager 114 may provide instructions to the deployment or tenant administrator to allow the administrator to perform additional analysis of usage data before making a determination about whether to modify the deployment of virtual appliance 202. This can be used to limit the number of updates to the virtual appliance deployment and allow the administrator to have more insight into the tenant deployment.
[0087] Now we will give information about the above combination. Figure 2 Additional details about each stage of the discussed workflow. For example, Figure 3A An example series of actions 302 related to sampling and collecting usage data is shown according to one or more embodiments. For example, Figure 3A A series of actions 302 are illustrated, including a first action 304 of sampling data at periodic sampling intervals. In this example, an agent on a virtual device can be configured to sample usage data every ten seconds. Other implementations may involve sampling data more frequently or less frequently.
[0088] like Figure 3AAs shown, the series of actions 302 also includes action 306, which aggregates sampled usage data at predetermined time intervals (e.g., every 15 minutes). For example, in one or more embodiments, the virtual service upgrade system 106 may aggregate one or more types of usage data sampled by an agent on a virtual device and received over time. The aggregated data may include a summary or average of one or more usage metrics over time, such as the amount or percentage of cloud computing resources used by the virtual device within the relevant time interval.
[0089] like Figure 3A As shown, this series of actions 302 may also include action 308, which generates usage summary data within a certain time period (e.g., one day, one week, two weeks). Usage summary data may include any information associated with the consumption of various cloud computing resources during the relevant time period. For example, usage summary data may include an indication or summary of CPU resources, memory resources, I / O resources, or any other cloud computing resources consumed by the virtual appliance during the relevant time period. Although Figure 3A An example of a virtual service upgrade system 106 generating a daily usage summary is shown, but other implementations may include weekly or bi-weekly summaries associated with a predetermined time period in which a current usage score is calculated.
[0090] Figure 3B Example series of actions 310 are shown for determining various usage metrics over time. For example, such as... Figure 3B As shown, the virtual service upgrade system 106 can perform action 312 to initialize usage metrics for a predetermined time period. In the example of providing a daily usage summary, the virtual service upgrade system 106 can initialize the usage metrics to zero and continue analyzing usage data over multiple time intervals.
[0091] like Figure 3B As shown, the series of actions 310 may include actions 314 that iterate over usage metrics and loop to the next interval. For example, the virtual service upgrade system 106 may loop or otherwise continue analyzing data from the first interval (e.g., after initializing the usage metric to zero). As another example, the virtual service upgrade system 106 may loop from the first interval to the second interval. The virtual service upgrade system 106 may iterate over any number of intervals included within a predetermined time period corresponding to the usage profile. In one or more embodiments, the virtual service upgrade system 106 may iterate over each interval of the predetermined time period from beginning to end, calculating a usage score for the virtual device within that predetermined time period.
[0092] like Figure 3BAs shown, the virtual service upgrade system 106 can perform action 316 to determine daily usage metrics. Specifically, after iterating through multiple intervals constituting a daily period (or other predetermined period) from beginning to end, the virtual service upgrade system 106 can determine a set of daily usage metrics reflecting the consumption of cloud computing resources within the daily period. This can include a count of intervals that satisfy one or more conditions, such as the number of intervals in which non-zero measures of resource usage are observed and / or the number of intervals in which virtual devices use a threshold percentage of cloud computing resources.
[0093] The virtual service upgrade system 106 can additionally perform a series of actions for each interval before determining (multiple) daily usage metrics. For example, such as Figure 3B As shown, the virtual service upgrade system 106 can perform action 318 to determine whether a virtual device is used for a minimum threshold amount of time interval. If the virtual device is not used at all or if a minimum amount of resource usage is observed within the relevant time interval, the virtual service upgrade system 106 can return to action 314, which involves looping to the next interval without performing any additional processing within that interval.
[0094] Conversely, if the usage data for that time interval indicates cloud resource usage above a minimum threshold, the virtual service upgrade system 106 can perform action 320 to calculate a time usage metric for that time interval. In one or more embodiments, this may involve determining the amount of cloud resources available for that time interval. In one or more embodiments, this may simply involve identifying whether a minimum threshold metric is met within that time interval.
[0095] like Figure 3B As shown, based on the observed minimum resource consumption threshold and the determination of the time usage metric for that time interval, the virtual service upgrade system 106 can perform action 322 to determine whether a peak usage threshold is met within that time interval. As mentioned above, this may involve determining the percentage of cloud computing resources consumed within a specific time interval. As an illustrative example, the virtual service upgrade system can determine whether a virtual device consumes more than 75% of cloud computing resources within that time interval.
[0096] If virtual devices are not used at or above a threshold percentage, the virtual service upgrade system 106 can return to action 314, which involves looping to the next interval without increasing the peak metric. In this example, the virtual service upgrade system 106 can update the time usage interval without increasing the peak metric value. Conversely, if the usage data for that interval indicates a peak percentage higher than the threshold percentage, the virtual service upgrade system 106 can perform action 324 to determine the peak metric for that interval and return to iterating over the usage metric value determined for that interval. In this example, the virtual service upgrade system 106 can update both the time usage metric and the peak usage metric for the relevant time period.
[0097] Figure 3C Example series of actions 330 is shown for determining the usage score of virtual devices over a predetermined period of time (e.g., two weeks). Specifically, Figure 3C An example series of actions 330 is shown, which can be executed by the virtual service upgrade system 106 to determine a current usage score based on a combination of peak usage metrics and time usage metrics determined for time intervals constituting a predetermined fourteen-day period. These actions 330 can be executed by the scoring engine 112 of the virtual service upgrade system according to one or more examples discussed above.
[0098] like Figure 3C As shown, the virtual service upgrade system 106 can perform action 332 of aggregating usage metrics over a predetermined time period. In one or more embodiments, the virtual service upgrade system accumulates usage information based on information received for each time interval. In one or more embodiments, the virtual service upgrade system 106 accumulates usage data from daily summaries obtained or otherwise generated during that time period. The aggregation can be updated at incremental intervals. For example, in one or more embodiments, the virtual service upgrade system 106 updates the aggregated data daily. In one or more embodiments, the virtual service upgrade system 106 updates the aggregated data for each time interval when updating and receiving usage metrics.
[0099] Based on the aggregated data, the virtual service upgrade system 106 can perform a series of actions according to the various calculations described above in conjunction with the scoring engine. For example, Figure 3C As shown, the virtual service upgrade system can perform action 334 to calculate peak percentage levels and usage time levels. As further shown, the virtual service upgrade system 106 can perform action 336 to calculate peak percentage boundaries and usage time boundaries. As mentioned above, these boundaries can refer to the boundaries of the values constituting the usage bucket.
[0100] As further shown, the virtual service upgrade system 106 can perform action 338 to calculate the usage buckets of virtual devices (e.g., within a relevant time period) based on peak percentage levels and usage time levels. In one or more embodiments, the virtual service upgrade system 106 can perform action 340 to calculate bucket boundaries. As further shown, and as will be combined below Figure 5 In further detail, the virtual service upgrade system 106 can perform action 342 to calculate the current usage score based on usage metrics.
[0101] In one or more embodiments, the virtual service upgrade system 106 determines a usage score based on a composite of usage metrics determined for multiple types of usage data. For example, in one or more embodiments, the usage score is based on usage metrics determined for a corresponding usage data type (such as CPU usage, memory usage, etc.). Alternatively, the usage score may be based on a specific type of usage data. In one or more embodiments, the virtual service upgrade system 106 determines multiple usage scores for each type of usage data and determines a combined usage score based on the usage scores calculated individually for each corresponding usage data type.
[0102] As discussed above, the virtual service upgrade system 106 can allocate virtual devices to one of multiple usage buckets based on associated usage data tracked over a previous time period. As an illustrative example, Figure 4 An example illustration of multiple usage buckets mapped onto a bucket grid 402, with axes corresponding to usage metrics determined for the virtual device, is provided. Specifically, Figure 4 Example bucket grid 402 is shown, which has an x-axis corresponding to a time usage metric in minutes and a y-axis corresponding to a peak usage metric in percentages. Figure 4 As shown, each of the nine regions in bucket grid 402 can be assigned a bucket value between 0 and 8. It should be understood that... Figure 4 An example implementation of bucket values corresponding to the respective regions of bucket grid 402 is shown. Other implementations may assign different bucket values to different regions. Furthermore, it should be understood that the size of the regions may be uniform, or alternatively, may be weighted differently based on any number of factors (e.g., the relative health of virtual devices using a metric for a specific range, the percentage of virtual devices assigned to the corresponding bucket, etc.).
[0103] like Figure 4As further illustrated, each of the buckets or discrete portions of the bucket grid 402 can be associated with a specific score range, as shown in the example bucket score table 404. These score ranges can refer to the range of usage scores that can be assigned to a virtual device when assigned to the corresponding usage bucket. Similar to the specific orientation and size of the bucket regions within the bucket grid, the ranges of usage scores are provided by example and can vary depending on various factors. In this example, a higher usage score could correspond to a healthier virtual device that is not operating at or near peak capacity, while a lower usage score could correspond to a less healthy virtual device that is operating at or near peak capacity and may be experiencing intermittent outages of various applications or services.
[0104] Figure 5 An example workflow is shown, illustrating one implementation where a usage score can be calculated for a virtual device. Specifically, Figure 5 A series of actions 502 that can be performed to determine specific values and metrics that can be used to determine a virtual device usage score over a predetermined time period are illustrated. Each of these metrics is discussed in further detail above with example formulas and equations. In particular, such as Figure 5 As shown, the virtual service upgrade system can determine a peak upper limit, a peak lower limit, a usage lower limit, and a usage upper limit. The virtual service upgrade system can further determine the peak level and usage level (e.g., usage metric) to identify the corresponding usage bucket for the virtual device. Based on the bucket identifier, the virtual service upgrade system can determine the lower limit bucket value and the upper limit bucket value, which can be used to determine the current usage score of the virtual device.
[0105] Now go to Figure 6 The diagram illustrates an example flowchart that includes a series of actions for collecting usage data, evaluating the usage data to determine various usage metrics, and determining a virtual machine's usage score based on one or more of these metrics. While Figure 6 Actions according to one or more embodiments are shown, but alternative embodiments may omit, add, reorder, and / or modify. Figure 6 Any action shown. Figure 6 The action can be performed as part of a method. Alternatively, a non-transitory computer-readable medium can include instructions that, when executed by one or more processors, cause a computing device to perform... Figure 6 The action. In a further embodiment, the system can perform... Figure 6 The action.
[0106] like Figure 6As shown, this series of actions 600 may include action 610 of collecting usage data of the virtual device. For example, in one or more embodiments, action 610 may involve receiving usage data associated with the virtual device. In one or more embodiments, the usage data is obtained by an agent on the virtual device.
[0107] As further illustrated, the series of actions 600 may include action 620 of classifying virtual devices into usage buckets associated with a usage score range based on a peak level metric for the virtual device. In one or more embodiments, action 620 involves determining a peak level metric for the virtual device based on the number of time intervals in which usage data exceeds a threshold within a predetermined time period. Determining the peak level metric based on the number of time intervals can significantly reduce processing costs, as this is the result of accumulating data according to the number of time intervals, rather than accumulating specific usage data over multiple days or weeks corresponding to a predetermined time period. Furthermore, in one or more embodiments, action 620 involves determining a usage bucket associated with the virtual device based on the peak level metric, wherein the usage bucket is associated with a range of usage scores. Similar to the concept of determining a peak level metric based on the number of time intervals, determining the usage bucket for a virtual device can similarly enable the system described herein to save processing costs in determining the usage category or classification of a device. In fact, by classifying virtual devices into buckets, it is not necessary to further analyze many virtual devices to determine whether upgrades and / or downgrades are necessary or otherwise helpful.
[0108] As further illustrated, the series of actions 600 may include action 630 of generating a current usage score for the virtual device based on the assigned usage bucket. For example, in one or more embodiments, action 630 involves determining the usage bucket associated with the virtual device based on a peak rank metric, wherein the usage bucket is associated with a range of usage scores. Similar to the above, generating the current usage score based on the category within the usage bucket and determining the usage metric based on the number of time intervals (rather than more robust usage data) enables the system of this paper to determine usage in a computationally efficient manner.
[0109] As further illustrated, the series of actions 600 may include action 640 of upgrading a virtual device based on its current usage score. For example, in one or more embodiments, action 640 involves upgrading the virtual device based on the current usage score being less than a threshold usage score over a predetermined time period. Upgrading virtual devices in this manner allows for resource elasticity for tenant deployments. Furthermore, by selectively identifying virtual devices for upgrades in this way, the system described herein can provide upgrades / downgrades in a selective manner unique to a specific deployment of the device.
[0110] In one or more embodiments, the virtual device is implemented on a server node of a cloud computing system and associated with a user in the cloud computing system. In one or more embodiments, the virtual device is associated with a first device type having a first number of computing resources and a first number of storage resources. In one or more implementations, upgrading the virtual device includes upgrading the virtual device from the first device type to a second device type having a second number of computing resources and a second number of storage resources. In one or more embodiments, upgrading the virtual device includes increasing the number of computing resources and / or increasing the number of storage resources of the virtual device. Furthermore, in one or more embodiments, upgrading the virtual device includes migrating the virtual device's data from a first server node to a second server node, wherein the second server node has greater storage capacity or computing capacity availability than the first server node. As discussed above, facilitating upgrades by upgrading from a first device type to a second device type, increasing computing resources, or simply migrating data from one node to another provides significant computing resilience and efficient utilization of computing resources for tenant deployments.
[0111] In one or more embodiments, determining a usage bucket associated with a virtual device includes comparing a peak level metric to a plurality of thresholds associated with different usage buckets. Determining a usage bucket may also include determining the usage bucket based on comparing the peak level metric to the plurality of thresholds. In one or more embodiments, a usage bucket is associated with a range of usage scores, wherein the current usage score is generated based on the percentage of time during which the peak level metric exceeds at least one of a plurality of thresholds within a predetermined time period. As described above, determining usage buckets and utilizing them to determine the current usage score provides a computationally efficient way to determine the usage score.
[0112] In one or more embodiments, the series of actions 600 includes determining a time usage metric for the virtual device based on a second number of time intervals indicating the use of the virtual device within a predetermined time period, indicated by usage data associated with the virtual device. As discussed herein, by determining the second number of time intervals associated with the use of the virtual device within the predetermined time period, a current usage score can be determined for a selective subset of the virtual devices. In one or more implementations, determining usage buckets is further based on the time usage metric for the virtual device. Furthermore, in one or more embodiments, determining usage buckets includes mapping the number of time intervals to a first coordinate in a grid space, mapping the second number of time intervals to a second coordinate in a grid space, and identifying usage buckets based on regions of the grid space corresponding to the first and second coordinates. Similar to one or more embodiments described above, determining the current usage score based on usage buckets provides a computationally efficient method for classifying the use of virtual devices.
[0113] In one or more embodiments, the series of actions 600 includes performing a set of actions for each time interval within a predetermined time period. For example, the series of actions 600 may include determining whether usage data indicates the use of a virtual device during the time interval. Furthermore, the series of actions 600 may include determining whether the usage data exceeds a threshold during the time interval. If the usage data indicates the use of a virtual machine during the time interval (e.g., a non-zero or higher than a minimum threshold amount), determining whether the usage data exceeds the threshold may include analyzing the usage data to determine whether the usage data exceeds the threshold during the time interval. Determining whether the usage data exceeds the threshold may further involve inferring that the usage data does not exceed the threshold without further analysis of the usage data if the usage data does not indicate the use of a virtual machine during the time interval. As described above, by comparing usage data to a threshold, this method enables selective analysis of a subset of virtual devices in a manner that saves significant processing costs in identifying which virtual devices among those used for deployment should be considered for upgrade (or downgrade).
[0114] In one or more embodiments, the data used includes at least one of telemetry for one or more central processing units (CPUs) associated with the virtual device, telemetry for memory associated with the virtual device, and / or telemetry for I / O activity associated with the virtual device. In one or more embodiments, the virtual device is one of a plurality of virtual machines associated with a tenant of a cloud computing system. Furthermore, in one or more embodiments, the series of actions 600 may include providing an upgrade interface to an administrative user associated with an account for the virtual device, wherein the upgrade interface includes optional options associated with upgrading the virtual device. The optional options may be presented based on a current usage score being less than a threshold usage score over a predetermined time period. Presenting an upgrade interface as discussed herein allows users to easily view which virtual devices in the virtual device have current scores that may require further review before determining whether to upgrade or downgrade the machine. According to the implementation described herein, this allows deployment administrators to view a small subset of devices for which the deployment is underutilized or overutilized, rather than resolving a set of scores for hundreds or thousands of virtual devices for a given deployment.
[0115] Figure 7 Certain components that may be included within computer system 700 are shown. One or more computer systems 700 may be used to implement the various devices, components, and systems described herein.
[0116] Computer system 700 includes processor 701. Processor 701 can be a general-purpose single-chip or multi-chip microprocessor (e.g., an advanced RISC (Reduced Instruction Set Computer) machine (ARM)), a special-purpose microprocessor (e.g., a digital signal processor (DSP)), a microcontroller, a programmable gate array, etc. Processor 701 can be referred to as a central processing unit (CPU). Although in Figure 7 The computer system 700 shows only a single processor 701, but in alternative configurations, a combination of processors (e.g., ARM and DSP) can be used.
[0117] The computer system 700 also includes a memory 703 that is in electronic communication with the processor 701. The memory 703 can be any electronic component capable of storing electronic information. For example, the memory 703 can be embodied as random access memory (RAM), read-only memory (ROM), disk storage media, optical storage media, flash memory devices in RAM, onboard memory included in the processor, erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, and combinations thereof.
[0118] Instruction 705 and data 707 may be stored in memory 703. Instruction 705 may be executable by processor 701 to implement some or all of the functions disclosed herein. Executing instruction 705 may involve using data 707 stored in memory 703. Any various examples of the modules and components described herein may be implemented in part or in whole as instruction 705 stored in memory 703 and executed by processor 701. Any various examples of the data described herein may be in data 707 stored in memory 703 and used during the execution of instruction 705 by processor 701.
[0119] The computer system 700 may also include one or more communication interfaces 709 for communicating with other electronic devices. The communication interface 709 may be based on wired communication technology, wireless communication technology, or both. Some examples of the communication interface 709 include Universal Serial Bus (USB), Ethernet adapters, and wireless adapters operating according to the Institute of Electrical and Electronics Engineers (IEEE) 802.11 wireless communication protocol. Wireless communication adapter and infrared (IR) communication port.
[0120] Computer system 700 may also include one or more input devices 711 and one or more output devices 713. Some examples of input devices 711 include keyboards, mice, microphones, remote control devices, buttons, joysticks, trackballs, touchpads, and light pens. Some examples of output devices 713 include speakers and printers. A specific type of output device typically included in computer system 700 is a display device 715. Display devices 715 used with the embodiments disclosed herein can utilize any suitable image projection technology, such as liquid crystal displays (LCDs), light-emitting diodes (LEDs), gas plasma, electroluminescence, etc. A display controller 717 may also be provided for converting data 707 stored in memory 703 into text, graphics, and / or moving images (as applicable) displayed on display device 715. Various components of computer system 700 may be interconnected via one or more buses, which may include power buses, control signal buses, status signal buses, data buses, etc. For clarity, the various buses are referred to as... Figure 7 It is shown as bus system 719.
[0121] The techniques described herein can be implemented in hardware, software, firmware, or any combination thereof, unless specifically described as being implemented in a particular manner. Any feature described as a module, component, etc., can also be implemented together as an integrated logic device or separately as a discrete but interoperable logic device. If implemented in software, the technique can be implemented at least in part through a non-transitory processor-readable storage medium comprising instructions that, when executed by at least one processor, perform one or more of the methods described herein. The instructions can be organized into routines, programs, objects, components, data structures, etc., which can perform specific tasks and / or implement specific data types, and can be combined or distributed as needed in various embodiments.
[0122] Computer-readable media can be any available medium that can be accessed by a general-purpose computer system or a special-purpose computer system. A computer-readable medium storing computer-executable instructions is a non-transitory computer-readable storage medium (device). A computer-readable medium carrying computer-executable instructions is a transmission medium. Therefore, embodiments of this disclosure may include at least two distinct types of computer-readable media, such as, and not by limitation: a non-transitory computer-readable storage medium (device) and a transmission medium.
[0123] As used herein, a non-transitory computer-readable storage medium (device) may include RAM, ROM, EEPROM, CD-ROM, solid-state drive (“SSD”) (e.g., RAM-based), flash memory, phase-change memory (“PCM”), other types of memory, other optical disc storage devices, disk storage devices or other magnetic storage devices, or any other medium that may be used to store desired program code in the form of computer-executable instructions or data structures and that may be accessed by a general-purpose computer or a special-purpose computer.
[0124] The steps and / or actions of the methods described herein may be interchanged without departing from the scope of the claims. In other words, unless the correct operation of the described method requires a specific order of steps or actions, the order and / or use of specific steps and / or actions may be modified without departing from the scope of the claims.
[0125] The term "determine" encompasses a wide variety of actions, and therefore, "determine" can include calculation, obtaining, processing, deriving, investigating, searching (e.g., looking in a table, database, or other data structure), ascertaining, etc. Furthermore, "determine" can include receiving (e.g., receiving information), accessing (e.g., accessing data in memory), etc. Additionally, "determine" can include resolving, selecting, picking, establishing, etc.
[0126] The terms “comprising,” “including,” and “having” are intended to be inclusive and mean that additional elements besides those listed may be present. Furthermore, it should be understood that references to “one embodiment” or “embodiment” in this disclosure are not intended to exclude the existence of additional embodiments also incorporating the described features. For example, where compatible, any element or feature described with respect to embodiments herein may be combined with any element or feature of any other embodiment described herein.
[0127] This disclosure may be embodied in other specific forms without departing from the spirit or characteristics of this disclosure. The described embodiments are to be considered illustrative rather than restrictive. Therefore, the scope of this disclosure is indicated by the appended claims rather than by the foregoing description. Changes within the meaning and equivalence of the claims should be included within the scope of the claims.
Claims
1. A computer-implemented method, comprising: Receive usage data instructing virtual devices to utilize one or more cloud computing resources; The peak level metric of the virtual device is determined based on the number of time intervals in which the data usage exceeds a threshold within a predetermined time period; The usage buckets associated with the virtual device are determined based on the peak level metric, wherein the usage buckets are associated with a range of usage scores; The current usage score of the virtual device is generated based on its association with the usage bucket. as well as The virtual device is upgraded based on the current usage score being less than a threshold usage score for the predetermined time period, wherein upgrading the virtual device includes: migrating the data of the virtual device from a first server node to a second server node, the second server node having greater availability of storage capacity or computing capacity compared to the first server node.
2. The computer-implemented method according to claim 1, wherein the virtual device is associated with a user of the cloud computing system.
3. The computer-implemented method according to claim 1, wherein the virtual device is associated with a first device type having a first number of computing resources and a first number of memory resources.
4. The computer-implemented method according to claim 3, wherein upgrading the virtual device further includes: The virtual device is upgraded from the first device type to the second device type, the second device type having a second number of computing resources and a second number of memory resources.
5. The computer-implemented method of claim 3, wherein migrating data of the virtual device from the first server node to the second server node results in at least one of the following: Increase the amount of computing resources for the virtual device; Increase the amount of memory resources for the virtual device; or Increase the amount of computing and storage resources of the virtual device.
6. The computer-implemented method of claim 1, wherein determining the usage bucket associated with the virtual device comprises: The peak level metric is compared with multiple thresholds associated with different usage buckets; as well as The usage bucket associated with the virtual device is determined by comparing the peak level metric with the plurality of thresholds.
7. The computer-implemented method of claim 6, wherein the usage bucket is associated with a range of usage scores, and wherein the current usage score is generated based on the percentage of time during which the peak level metric exceeds at least one of the plurality of thresholds within the predetermined time period.
8. The computer-implemented method according to claim 1, further comprising: A time usage metric for the virtual device is determined based on a second number of time intervals during which the virtual device was used within the predetermined time period, indicated by the usage data associated with the virtual device; and The determination of the usage bucket is also based on the time usage metric of the virtual device.
9. The computer-implemented method of claim 8, wherein determining the bucket usage comprises: Map the number of time intervals to a first coordinate in the grid space; Map the second number of time intervals to a second coordinate on the grid space; as well as The usage bucket is identified based on the region of the grid space corresponding to the first coordinate and the second coordinate.
10. The computer-implemented method of claim 1, further comprising, for each time interval within the predetermined time period: Determine whether the usage data indicates that the virtual device was used during the time interval; as well as Determining whether the usage data exceeds the threshold within the time interval includes: If the usage data indicates that the virtual device was used during the time interval, the usage data is analyzed to determine whether the usage data for the time interval exceeds the threshold; as well as If the usage data does not indicate that the virtual device was used during the time interval, it is inferred that the usage data did not exceed the threshold and the usage data is not analyzed further.
11. The computer-implemented method of claim 1, wherein the data used includes at least one of the following: Telemetry for at least one central processing unit (CPU) associated with the virtual device; Telemetry for memory associated with the virtual device; or Telemetry for input / output I / O activities associated with the virtual device.
12. The computer-implemented method of claim 1, wherein the virtual device is one of a plurality of virtual machines associated with a tenant of a cloud computing system.
13. The computer-implemented method according to claim 1, further comprising: An upgrade interface for managing users is provided, associated with an account for the virtual device, wherein the upgrade interface includes optional options associated with upgrading the virtual device, the optional options being presented based on the current usage score being less than the threshold usage score for the predetermined time period.
14. A system comprising: At least one processor; The memory communicates electronically with the at least one processor; as well as Instructions stored in the memory, which are executable by the at least one processor to: Receive usage data instructing virtual devices to utilize one or more cloud computing resources; The peak level metric of the virtual device is determined based on the number of time intervals in which the data usage exceeds a threshold within a predetermined time period; The usage buckets associated with the virtual device are determined based on the peak level metric, wherein the usage buckets are associated with a range of usage scores; The current usage score of the virtual device is generated based on its association with the usage bucket. as well as The virtual device is upgraded based on the current usage score being less than a threshold usage score for the predetermined time period, wherein upgrading the virtual device includes: migrating the data of the virtual device from a first server node to a second server node, the second server node having greater availability of storage capacity or computing capacity compared to the first server node.
15. The system of claim 14, wherein the virtual device is associated with a first device type having a first number of computing resources and a first number of memory resources, and wherein migrating data of the virtual device from the first server node to the second server node results in at least one of the following: Increase the amount of computing resources for the virtual device; Increase the amount of memory resources for the virtual device; or Increase the amount of computing and storage resources of the virtual device.
16. The system of claim 14, wherein determining the usage bucket associated with the virtual device comprises: The peak level metric is compared with multiple thresholds associated with different usage buckets; as well as The usage bucket associated with the virtual device is determined by comparing the peak level metric with the plurality of thresholds, wherein the current usage score is generated based on the percentage of time during which the peak level metric exceeds at least one of the plurality of thresholds within the predetermined time period.
17. The system of claim 14, further comprising instructions executable by the at least one processor to: A time usage metric for the virtual device is determined based on a second number of time intervals during which the virtual device was used within the predetermined time period, indicated by the usage data associated with the virtual device; and The determination of the usage bucket is also based on the time usage metric of the virtual device.
18. A method comprising: Receive usage data instructing virtual devices to utilize one or more cloud computing resources; A time usage metric for the virtual device, indicating that the virtual device was used within the predetermined time period, is determined based on a first number of time intervals in which the usage data exceeds a minimum threshold within a predetermined time period. The peak level metric of the virtual device is determined based on a second number of time intervals in which the data usage exceeds a maximum threshold within a predetermined time period; The usage buckets associated with the virtual device are determined based on the time usage metric and the peak level metric, wherein the usage buckets are associated with a range of usage scores; The current usage score of the virtual device is generated based on its association with the usage bucket. as well as Upgrading the virtual device based on its association with the usage bucket, wherein upgrading the virtual device includes migrating the data of the virtual device from a first server node to a second server node, the second server node having greater availability of storage or computing capacity compared to the first server node.
19. The method of claim 18, wherein migrating data of the virtual device from the first server node to the second server node results in at least one of the following: Increase the amount of computing resources for the virtual device; Increase the amount of memory resources for the virtual device; or Increase the amount of computing and storage resources of the virtual device.
20. The method of claim 19, wherein upgrading the virtual device further comprises: The virtual device is upgraded from a first device type with a first number of computing resources and memory resources to a second device type with a second number of computing resources and memory resources.
Citation Information
Patent Citations
Scaling a virtual machine instance
US20140058871A1
