An adaptive technique to model virtual machine behavior for scalable cloud monitoring

Abstract

Abstract-Supporting the emerging digital society is creating new challenges for cloud computing infrastructures, exacerbating scalability issues regarding the processes of resource monitoring and management in large cloud data centers. Recent research studies show that automatically clustering similar virtual machines running the same software component may improve the scalability of the monitoring process in IaaS cloud systems. However, to avoid misclassifications, the clustering process must take into account long time series (up to weeks) of resource measurements, thus resulting in a mechanism that is slow and not suitable for a cloud computing model where virtual machines may be frequently added or removed in the data center. In this paper, we propose a novel methodology that dynamically adapts the length of the time series necessary to correctly cluster each VM depending on its behavior. This approach supports a clustering process that does not have to wait a long time before making decisions about the VM behavior. The proposed methodology exploits elements of fuzzy logic for the dynamic determination of time series length. To evaluate the viability of our solution, we apply the methodology to a case study considering different algorithms for VMs clustering. Our results confirm that after just 1 day of monitoring we can cluster without misclassifications up to 80% of the VMs, while for the remaining 20% of the VMs longer observations are needed.

AGATE: Adaptive Gray Area-Based TEchnique to Cluster Virtual Machines with Similar Behavior

“…A preliminary version of the gray area-based technique was proposed in [6], but this study improves our previous work in several ways. First, the original mechanism was based on a fixed threshold parameter, while in this proposal we adaptively compute the threshold depending on the results of the clustering operation: this greatly reduces the amount of data collected and the time required to determine the VMs behavior.…”

“…A preliminary version of the gray area-based technique was proposed in [6], but this previous work has many limitations. First, the size of the gray area was determined on the basis of a fixed value instead of being adaptively computed according to the results of clustering, resulting in a significant increase in the amount of collected data and in the time needed to take decisions on VMs behavior.…”

Advancing Cloud Network Management with Smart Monitoring Technique

“…While there have been several methods in use to cluster virtual machines, such mechanism does not reach the required reactive levels in situations where virtual machines are required to be added or removed based on dynamic nature of Cloud Computing dynamics involving consumers' needs. In a research [4], the authors propose an adaptive approach to clustering virtual machines based on time series and degrees of uncertainty that may result from clustering processes. The applicability of virtual network functions (VFNs) has a great deal of significance in the context of geographically distributed Cloud Computing systems and infrastructure.…”

Sieve: Actionable Insights from Monitored Metrics in Microservices

“…Meng et al [73] optimize monitoring scalability by choosing appropriate monitoring window lengths and adjusting the monitoring intensity at runtime. Canali et al [42] achieve scalability by clustering metric data. A fuzzy logic approach is used to speed up clustering, and thus obtain data for decision making within shorter periods.…”