Effective VM sizing in virtualized data centers

Abstract

Abstract-In this paper, we undertake the problem of server consolidation in virtualized data centers from the perspective of approximation algorithms. We formulate server consolidation as a stochastic bin packing problem, where the server capacity and an allowed server overflow probability p are given, and the objective is to assign VMs to as few physical servers as possible, and the probability that the aggregated load of a physical server exceeds the server capacity is at most p.We propose a new VM sizing approach called effective sizing, which simplifies the stochastic optimization problem by associating a VM's dynamic load with a fixed demand. Effective sizing decides a VM's resource demand through statistical multiplexing principles, which consider various factors impacting the aggregated resource demand of a host where the VM may be placed. Based on effective sizing, we design a suite of polynomial time VM placement algorithms for both VM migration cost-oblivious and migration cost-aware scenarios.Through analysis, we show that our algorithm is O(1)-approximation for the stochastic bin packing problem when the VM loads can be modeled as all Poisson or all normal distributions. Through evaluations driven by a real data center load trace, we show that our consolidation solution can achieve an order of reduction on physical server requirement compared to that before consolidation; the consolidation result is only 24% more than the optimal solution. With effective sizing, our server consolidation solution achieves 10% to 23% more energy savings than state-of-the-art approaches.

Power-Thermal Modeling and Control of Energy-Efficient Servers and Datacenters

“…Innovative DCIM support systems for datacenter management are thus needed. PMSM (i.e., Power Monitor System and Management) [44][45][46][47][48][49][50][51][52][53][54][55][56], developed at EPFL in cooperation with Credit Suisse [45], is an example of such an innovation.…”

“…In [51], Verma et al found out that workloads running on datacenters are strongly correlated one another. In order to achieve further power savings while maintaining quality of service (QoS) level, correlations among VMs' workload have been exploited in recent works [51][52][53][54]. In [51], Verma et al presented a clustering-based correlation-aware VM placement solution.…”

“…Pearson's correlation, is most widely used correlation measure to quantify the correlation of used CPU utilization among VMs [53]. It is calculated as the ratio of covariance of the two random variables to the product of their standard deviations.…”

SOC: Satisfaction-Oriented Virtual Machine Consolidation in Enterprise Data Centers

“…There are existing tools for monitoring and collecting hardware configuration information, and VM resources usage traces, such as IBM Tivoli Monitoring (ITM) [1], which is currently used by our industrial partner. VMs' resource demands can also be forecasted by several techniques such as the ones presented in [9,12]. Filling the DB is, however, out of the scope of this paper.…”

Comparing VM-Placement Algorithms for On-Demand Clouds

“…In on-demand clouds, there are potentially two types of VM-placement decisions to be made: (1) initial placement [13][14][15][16][17][18][19][20][21][22][23] and (2) migration (and/or resizing) of VMs over time [24][25][26][27][28][29][30], as PM availability changes, as consolidation is needed to conserve power and in response to the degree to which service-level agreements (SLAs) are being achieved. Most previous research on initial VM placement considered only PMs within a single cloud, but in one case [22] placement decisions considered which of several clouds to choose.…”