Auto-scaling Mechanism Research Articles

Energy-efficient auto-scaling of virtualized network function instances based on resource execution pattern

To scale virtualized network functions (VNF), information about computing resource utilization plays an important role. With the usage of both virtualized and physical resources, introduced new challenges that need to be solved by virtual network functions manager (VNFM), which are responsible for managing the lifecycle of VNFs. These challenges demand an auto-scaling mechanism that relies on proper factors. The proposed system considers the execution time, thresholds of a VNF, and weight factor reflecting virtualization overhead as the practical factors. An auto-scaling application, along with the monitoring application, is developed. These two applications communicate with each other via application programming interfaces (API) to generate the auto-scaling configurations of resources. In the light of mentioned factors, mechanism, and later provided results, the proposed system proves to be useful in solving the challenges related to energy and allocation of resources for both virtual and physical infrastructures.

Bottleneck Detection in Cloud Computing Performance and Dependability: Sensitivity Rankings for Hierarchical Models

Cloud computing became widespread on IT industry, saving costs of acquisition and maintenance for companies of all sizes, and enabling fair management of resources according to the demand. Stochastic models can enable performance and dependability evaluation of cloud computing systems efficiently, what is needed for proper capacity planning. Distinct models may be combined in a hierarchy to address the huge number of components and levels of interaction among the system parts. Identification of bottlenecks in such composite models might be hard yet, due to the huge amount of input factors and variables which may interfere with the results. This paper proposes a method for bottleneck detection of computational systems represented with hierarchical models, that is remarkably applied in cloud computing systems. This is achieved through the composition of indices computed from lower level models in equations and solution methods of the top level model, for computing the sensitivity indices of all parameters with respect to a global system measure. A unified sensitivity ranking, comprising the composite indices, indicates the parameters with highest impact on output metrics. A case study supports the demonstration of accuracy and utility of our methodology. The study addresses a web service running on a private cloud with auto scaling mechanisms. The methods and algorithms presented here are helpful for decision-making when designing and managing cloud computing infrastructures, regarding incremental and architectural improvements.

Horizontal Pod Autoscaling in Kubernetes for Elastic Container Orchestration.

Kubernetes, an open-source container orchestration platform, enables high availability and scalability through diverse autoscaling mechanisms such as Horizontal Pod Autoscaler (HPA), Vertical Pod Autoscaler and Cluster Autoscaler. Amongst them, HPA helps provide seamless service by dynamically scaling up and down the number of resource units, called pods, without having to restart the whole system. Kubernetes monitors default Resource Metrics including CPU and memory usage of host machines and their pods. On the other hand, Custom Metrics, provided by external software such as Prometheus, are customizable to monitor a wide collection of metrics. In this paper, we investigate HPA through diverse experiments to provide critical knowledge on its operational behaviors. We also discuss the essential difference between Kubernetes Resource Metrics (KRM) and Prometheus Custom Metrics (PCM) and how they affect HPA’s performance. Lastly, we provide deeper insights and lessons on how to optimize the performance of HPA for researchers, developers, and system administrators working with Kubernetes in the future.

Razor: Scaling Backend Capacity for Mobile Applications

The dramatic growth of mobile application usage has posed great pressure on application developers to better manage their backend capacity. Rule-based or schedule-based auto-scaling mechanisms have been proposed, but it is difficult or expensive to frequently adjust the backend capacity to track the burstiness of mobile traffic. In this paper, we explore a fundamentally different approach. Instead of scaling the backend in line with the mobile traffic, we smooth out traffic profiles to reduce the required backend capacity and increase its utilization. Our proposed solution, called Razor, is inspired by two key insights on mobile traffic. First, mobile traffic exhibits high short-term fluctuations but steady long-term trend, so that we may temporarily delay user requests and periodically adapt backend capacity based on the predicted traffic volume. Second, user requests have different priorities: while some requests are urgent (e.g., sending a message), some are delay-tolerant (e.g., changing the profile photo) and can be postponed without much influence on the user experience. Based on these observations, our design features a two-tier architecture: on a long timescale, Razor predicts future traffic using machine learning algorithms and plans the optimal backend capacity to minimize the budget with performance guarantee; on a short timescale, Razor schedules which requests to delay and by how much time to delay according to their delay tolerance. We implement a fully-functional prototype of Razor, and evaluate its performance with both real and synthetic traces. Extensive experimental results show that Razor can effectively help mobile application developers reduce their backend cost while guaranteeing the user experience.

Towards Yo-Yo attack mitigation in cloud auto-scaling mechanism

Cloud platforms could automatically scale underlying network resources up and down in response to changes in the traffic load. Such an auto-scaling mechanism can largely enhance the elasticity and scalability of cloud platforms. However, it may introduce new security threats. For example, the Yo-Yo attack is a newly disclosed attack against the cloud auto-scaling mechanism. Attackers periodically send bursts of traffic to cause the auto-scaling mechanism to oscillate between the scale-up process and the scale-down process, which may result in significant performance degradation and economic loss. None of the prior work addressed the problem of mitigating such an attack. In this paper, we propose a Trust-based Adversarial Scanner Delaying (TASD) approach to effectively and proactively mitigate the Yo-Yo attack on the cloud auto-scaling mechanism. In TASD, we first propose to use the trust-based scheme to establish trust values for users, which is leveraged to identify adversarial requests. Trust values are updated by jointly considering the request mode and the auto-scaling status. Then, we aim to disable the condition under which the Yo-Yo attack takes effect by injecting certain delay, under the QoS constraints, to manipulate the response time of suspicious requests and deceive the attackers. Our extensive evaluation demonstrates that our approach achieves promising results, e.g., it can detect at least 80% Yo-Yo adversarial users and reduce more than 41% malicious scale-ups.

Efficient resource provisioning for elastic Cloud services based on machine learning techniques

Automated resource provisioning techniques enable the implementation of elastic services, by adapting the available resources to the service demand. This is essential for reducing power consumption and guaranteeing QoS and SLA fulfillment, especially for those services with strict QoS requirements in terms of latency or response time, such as web servers with high traffic load, data stream processing, or real-time big data analytics. Elasticity is often implemented in cloud platforms and virtualized data-centers by means of auto-scaling mechanisms. These make automated resource provisioning decisions based on the value of specific infrastructure and/or service performance metrics. This paper presents and evaluates a novel predictive auto-scaling mechanism based on machine learning techniques for time series forecasting and queuing theory. The new mechanism aims to accurately predict the processing load of a distributed server and estimate the appropriate number of resources that must be provisioned in order to optimize the service response time and fulfill the SLA contracted by the user, while attenuating resource over-provisioning in order to reduce energy consumption and infrastructure costs. The results show that the proposed model obtains a better forecasting accuracy than other classical models, and makes a resource allocation closer to the optimal case.

Chameleon: A Hybrid, Proactive Auto-Scaling Mechanism on a Level-Playing Field

Auto-scalers for clouds promise stable service quality at low costs when facing changing workload intensity. The major public cloud providers provide trigger-based auto-scalers based on thresholds. However, trigger-based auto-scaling has reaction times in the order of minutes. Novel auto-scalers from literature try to overcome the limitations of reactive mechanisms by employing proactive prediction methods. However, the adoption of proactive auto-scalers in production is still very low due to the high risk of relying on a single proactive method. This paper tackles the challenge of reducing this risk by proposing a new hybrid auto-scaling mechanism, called Chameleon, combining multiple different proactive methods coupled with a reactive fallback mechanism. Chameleon employs on-demand, automated time series-based forecasting methods to predict the arriving load intensity in combination with run-time service demand estimation to calculate the required resource consumption per work unit without the need for application instrumentation. We benchmark Chameleon against five different state-of-the-art proactive and reactive auto-scalers one in three different private and public cloud environments. We generate five different representative workloads each taken from different real-world system traces. Overall, Chameleon achieves the best scaling behavior based on user and elasticity performance metrics, analyzing the results from 400 hours aggregated experiment time.

A Lightweight Autoscaling Mechanism for Fog Computing in Industrial Applications

Fog computing provides a more flexible service environment than cloud computing. The lightweight fog environment is suitable for industrial applications. In order to strengthen service scalability, container virtualization has been proposed and studied in recent years. It is vital to explore the tradeoff between service scalability and operating expenses. This paper integrates the hypervisor technique with container virtualization, and constructs an integrated virtualization (IV) fog platform for deploying industrial applications based on the virtual network function. This paper presents a fuzzy-based real-time autoscaling (FRAS) mechanism and implements it in the IV fog platform. The FRAS mechanism provides a dynamic, rapid, lightweight, and low-cost solution to the service autoscaling problem. Experimental results showed that the proposed FRAS mechanism yields a better service scale with lower average delay, error rate, and operating expenses compared to other autoscaling schemes.

HAS: Hybrid auto-scaler for resource scaling in cloud environment

Auto-scaling is a crucial mechanism that supports autonomic provisioning and de-provisioning of computing resources in accordance with fluctuating demands in a cloud environment. The success factor of autonomic provisioning depends on efficient resource utilization and response time performance metrics. Existing literature focuses on reactive or predictive auto-scaling mechanism where the computing system is unable to scale proportionally with the Slashdot effect or abrupt traffic bursts while these mechanisms are employed in a discrete fashion. Predictive methods strive to predict the future computational needs and subsequently obtain or release the resources in advance; however it could be directed to under-utilization. Hence, a Hybrid Auto-Scaler (HAS) is proposed to adjust the required resources automatically to the application in demand. HAS forecasts the future behaviour of the system using a time series method and deploys the anticipated resources by computing the required capacity through a queuing model. Further, it uses a reactive approach to scale out the resources in accordance as the provisioned resources are insufficient to deal with the current needs. HAS also balances the load efficiently by employing Continuous Time Markov Model (CTMM). The proposed HAS is validated with several benchmark workloads to achieve significant improvement in CPU utilization and response time.

Slice Orchestration for Multi-Service Disaggregated Ultra-Dense RANs

Ultra-dense networks (UDNs) are a natural deployment evolution for handling the tremendous traffic increase related to the emerging 5G services, especially in urban environments. However, the associated infrastructure cost may become prohibitive. The evolving paradigm of network slicing can tackle such a challenge while optimizing the network resource usage, enabling multi-tenancy and facilitating resource sharing and efficient service-oriented communications. Indeed, network slicing in UDN deployments can offer the desired degree of customization in not only vanilla radio access network (RAN) designs, but also in the case of disaggregated multi-service RANs. In this article, we devise a novel multi-service RAN environment, RAN runtime, capable of supporting slice orchestration procedures and enabling flexible customization of slices as per tenant needs. Each network slice can exploit a number of services, which can be either dedicated or shared between multiple slices over a common RAN. The novel architecture we present concentrates on the orchestration and management systems. It interacts with the RAN modules, through the RAN runtime, via a number of new interfaces enabling a customized dedicated orchestration logic for each slice. We present results for a disaggregated UDN deployment where the RAN runtime is used to support slice-based multi-service chain creation and chain placement, with an auto-scaling mechanism to increase the performance.

Modeling and Extracting Load Intensity Profiles

Today’s system developers and operators face the challenge of creating software systems that make efficient use of dynamically allocated resources under highly variable and dynamic load profiles, while at the same time delivering reliable performance. Autonomic controllers, for example, an advanced autoscaling mechanism in a cloud computing context, can benefit from an abstracted load model as knowledge to reconfigure on time and precisely. Existing workload characterization approaches have limited support to capture variations in the interarrival times of incoming work units over time (i.e., a variable load profile). For example, industrial and scientific benchmarks support constant or stepwise increasing load, or interarrival times defined by statistical distributions or recorded traces. These options show shortcomings either in representative character of load variation patterns or in abstraction and flexibility of their format. In this article, we present the Descartes Load Intensity Model (DLIM) approach addressing these issues. DLIM provides a modeling formalism for describing load intensity variations over time. A DLIM instance is a compact formal description of a load intensity trace. DLIM-based tools provide features for benchmarking, performance, and recorded load intensity trace analysis. As manually obtaining and maintaining DLIM instances becomes time consuming, we contribute three automated extraction methods and devised metrics for comparison and method selection. We discuss how these features are used to enhance system management approaches for adaptations during runtime, and how they are integrated into simulation contexts and enable benchmarking of elastic or adaptive behavior. We show that automatically extracted DLIM instances exhibit an average modeling error of 15.2% over 10 different real-world traces that cover between 2 weeks and 7 months. These results underline DLIM model expressiveness. In terms of accuracy and processing speed, our proposed extraction methods for the descriptive models are comparable to existing time series decomposition methods. Additionally, we illustrate DLIM applicability by outlining approaches of workload modeling in systems engineering that employ or rely on our proposed load intensity modeling formalism.

Cloud Resource Management With Turnaround Time Driven Auto-Scaling

Cloud resource management research and techniques have received relevant attention in the last years. In particular, recently numerous studies have focused on determining the relationship between server-side system information and performance experience for reducing resource wastage. However, the genuine experiences of clients cannot be readily understood only by using the collected server-side information. In this paper, a cloud resource management framework with two novel turnaround time driven auto-scaling mechanisms is proposed for ensuring the stability of service performance. In the first mechanism, turnaround time monitors are deployed in the client-side instead of the more traditional server-side, and the information collected outside the server is used for driving a dynamic auto-scaling operation. In the second mechanism, a schedule-based auto scaling preconfiguration maker is designed to test and identify the amount of resources required in the cloud. The reported experimental results demonstrate that using our original framework for cloud resource management, stable service quality can be ensured and, moreover, a certain amount of quality variation can be handled in order to allow the stability of the service performance to be increased.

Efficient Resource Allocation Mechanism for Federated Clouds

This study proposes a novel efficient resource allocation mechanism for federated clouds, which takes the communication overhead into consideration, to improve system throughput and reduce resource repacking overhead in the auto-scaling mechanism. In general, when the amount of service requests increases, more and more resources are allocated to satisfy these requests. However, single cloud cannot provide unlimited services with limited physical resources; therefore, the federation of multiple clouds may be one possible solution. In the federated cloud environment, when the workload changes, the resource allocation mechanism could adopt vertical/horizontal scaling fashions to repack the required resource into virtual machines. In the vertical scaling approach, the resource allocation mechanism allocates more resources into virtual machines for improving virtual machine's capability. In the horizontal scaling approach, the resource allocation mechanism allocates more virtual machines for enhancing the virtual cluster's capability. However, frequent resource repacking may reduce the system performance. Therefore, in order to improve system throughput and reduce repacking overhead, the proposed mechanism captures the execution pattern of applications by the profiling system and the resource status by the monitoring system, and then allocates resources for configuring the virtual cluster. Performance for NAS Parallel Benchmarks is evaluated. Experimental results show that the authors' approach could reduce repacking overhead and improve system throughput by comparing two previous works.

Auto-Scaling, Load Balancing and Monitoring As service in public cloud

Cloud computing(1)(2)(3) word has changed the classical computing environment in IT industries. it is most emerging & popular technology in IT & research field because of its great feature such as virtualization and on-demand resource allocating (dynamic) . Now - a - days because of increased use of internet the associate resource are increasing rapidly resulting generation of high work load. To provide the reliable service to client with QOS(5 ) the load balancing mechanism is necessary in cloud environment, to prevent system from overloading and crash an autoscaling mechanism must also be provided according to the application and incoming user traffic. load balancing mechanism provides the distribution of load among one or more nodes of cloud system, for efficient service model autoscaling feature also enabled with the load balancer to handle the excess load. Auto scaling scaled-up and scaled down the platform dynamically according to the clients incoming traffic this save money and physical resources. Latency based routing is the new concept in cloud computing which provide the load balancing based on DNS latency to global client by mapping domain name system (DNS) (10 )through the different hosted zone .provide the load balancing based on geographical service region. To achieve above mentioned we use the public cloud services such as amazons'EC2. ELB. This research is divided in four part such as: i) load balancing ii) auto scaling iii)latency based routing iv) resource monitoring . while discussing each topic in detailed we will implement the individual service and test while providing load from external software tool putty we will produce result for efficient load balancing.

An auto-scaling mechanism for virtual resources to support mobile, pervasive, real-time healthcare applications in cloud computing

Cloud computing with virtualization technologies has become an important trend in the information technology industry. Due to its salient features of reliability and cost effectiveness, cloud computing has changed the paradigms of development for mobile pervasive services, effectively permeating the market. While most types of best effort mobile pervasive applications can be seamlessly migrated to cloud computing infrastructures, we need to consider specialized elements to make cloud computing infrastructures more effective in real-time healthcare applications. The client side of those applications dramatically increases its transmission rate whenever it detects an abnormal event. However, the existing server side mechanisms have limitations in adaptively allocating necessary computing resources in order to handle these various data volumes over time. In this article, we propose a novel server-side auto-scaling mechanism to autonomously allocate virtual resources on an on-demand basis. The mechanism is tested in an Amazon EC2, and the results show how the proposed mechanism can efficiently scale up and down the virtual resources, depending on the volume of requested real-time tasks.