1 / 18

Eco4Cloud – HP Rest Report November 2013

Eco4Cloud – HP Rest Report November 2013. Eco4Cloud in a nutshell. An innovative solution for efficiency optimization and energy reduction in virtualized data centers Resources utilization up to 90% 30‑60% energy reduction

aimon
Télécharger la présentation

Eco4Cloud – HP Rest Report November 2013

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Eco4Cloud – HPRest ReportNovember 2013

  2. Eco4Cloud in a nutshell • An innovative solution for efficiency optimization and energy reduction in virtualized data centers • Resources utilization up to 90% • 30‑60% energy reduction • Pay-back times ~6 months and 3­yrs ROI >120% in typical scenarios

  3. Eco4CloudImprove your data center efficiency and save energy What we do We help our Customers improve the economics of their virtualized data centers by using an intelligent software platform that reduces the energy consumption and increases efficiency. Who we are E4C is a spin-off from ICAR-CNR, the internationally renowned excellence center on high-performance computing and networking, and Italy’s most relevant research center on Cloud Computing.

  4. Role of Eco4Cloud in a DCIM Beyond physical efficiency In the last years important results have been achieved by improving the efficiency of the physical infrastructure(e.g., supplying and cooling components), as testified by the very good values of PUE (Physical Usage Effectiveness) obtained in modern data centers. • Physical efficiency Computational efficiency via VM consolidation However there is still much room for improving the computational efficiency: very often computing resources (CPU, RAM, bandwidth) are under-utilized. The objective of consolidation is to use the minimum number of servers to sustain the VMs workload. Here is where Eco4Cloud plays! Using a novel distributed/self-organizing approach. • Computational efficiency

  5. Role of Eco4Cloud in a data center • Positioning in a DCIM Model • E4C Architecture Eco4Cloud acts as a Virtualization Infrastructure Management tool, in a broader DCIM strategy Eco4Cloud has an unobtrusive architecture, as it works as a plugin to the virtualization platform

  6. VALUE PROPOSITION - Business Risk Management of business-critical IT infrastructures Better informed IT Budget Planning Regulatory Requirements EPA2007, E-Server2007, ISO14000, EMAS, EnergyStar, CEEDA, EU CoC Environmental/Green Corporate Reputation 20-60% reduced CO2 emission Company Profitabiliity CapEx & OpEx Reduction Compelling ROI

  7. VALUE PROPOSITION - Technical Better informed Capacity Planning Energy, power, cooling, I/O Meet DC SLA’s Reliliability, availability, performance Ease of deployment& management Hypervisor agnostic Consolidation and efficiency Up to 90+% server utilization Reduction ofDC energy bill 30-60% Scalability Adaptive/self-organizeddistributed algorithm

  8. Eco4Cloud Technology in a Nutshell • The data center manager assigns and migrates VMs to servers based on local probabilistic trials: • Lightly and highly loaded servers tend to reject VMs • Servers with intermediate load tend to accept VMs • Eventually, the workload is distributed to a low number of highly utilized servers 1 0.8 0.6 ----- CPU utilization ----- 0.4 0.2 0 0 5 10 15 20 25 30 ----- Time (hours) -----

  9. Eco4Cloud solution key elements Efficient Use of Resources • E4C increases resources utilization by 2X • E4C consolidates VMs on the minimum number of servers reducing costs and consumed energy • E4C unloads up to 60% of servers, which become ready to accommodate more load Multi-disciplinaryEffort Eco4Cloud aligns the following cross-functional/roles priorities: CEO, COO, CFO: Profitability improvement, CAPEX/OPEX reduction, regulatory requirements, risk management, green reputation CIO, CTO: ROI of IT investments, resource release for innovation (“save to invest”), cloud architectural evolution of IT DC Manager: SLAs, Capacity Planning, Automation Technical Staff: Operational efficiency and control Corporate wide deployment Eco4Cloud scales across DCs of any size, as well as geographically distributed DCs

  10. Eco4Cloud solution key elements Innovation • Eco4Cloud adopts an advanced solution derived from research at the Italian National Research Center on High Performance Computing and Networks (ICAR-CNR) • An innovative self-organizing algorithm allows to solve the very hard problem of multi-resource consolidation, i.e., consolidating VMs on the minimum number of servers respecting the limits of all the hardware resources (CPU, RAM) • The algorithm also prevents any possible server overload by timely performing migrations when needed • The algorithm’s validity was proven through the analysis of real deployment results, simulations, mathematical models Transferability Eco4Cloud can be adopted in any virtualized data center Eco4Cloud is currently integrated with VMWare and Hyper-V, and will be soon integrated with KVM Eco4Cloud is ready to be used in Software Defined Networks, with even improved benefits, due to the possibility of migrating VMs on more servers and remote DCs

  11. Features • Consolidationof max number of VMs on min number of physical servers, and dynamicadaptation based on the data center workload • Support of most common virtualizationplatforms: VMWare vSphere, Microsoft Hyper-V, KVM • Easily downloadable and installable as a virtual appliance with console access from any web browser • Continuous monitoring of the data center to improve energysaving and prevent SLAviolations

  12. HP-Eco4Cloud deployment Methodology • Identification of typical workloadsand resource allocation trends of real datacenters • Setting up of tools which allow to define repeatable, synthetic loads, resembling the identified workloads on a different scale • Execution and repetition of such workloads under different conditions (e.g. by activating/deactivating optimization strategies) • Collections of measurements for the key performance indicators

  13. HP-Eco4Cloud Deployment • Deployment Scenario • Datacenter with typical resources utilization. • CPU/RAM usage in working hours: about 50% • CPU/RAM usage in non working hours : about 25%

  14. HP-Eco4Cloud Deployment Performance Results

  15. HP-Eco4Cloud Deployment Deployment Phase 2: Eco4Cloud assessment/tuning

  16. HP-Eco4Cloud Deployment Deployment Phase 3: Eco4Cloud activated # of active servers and power usage with Eco4Cloud • Consumed energy: • without Eco4Cloud: 53.089 kWh / 186.1 kWh • with Eco4Cloud: 27.098 kWh / 96.6 kWh Energy saving: 48.1%

  17. HP-Eco4Cloud Deployment Quality of Service • Quality of Service is measured through: • CPU ready time– % of time that the virtual machine was ready, but could not get scheduled to run on the physical CPU. • Ballooned memory – Ballooning occurs when the VMware Host recognizes that there is a shortage of machine memory and must be replenished using page replacement. • No ballooning observed • CPU ready-time well below the warning threshold (5%)

  18. Useful Links FAQ www.eco4cloud.com/download/FAQ-eco4cloud.pdf Architecture and requirements www.eco4cloud.com/download/Eco4Cloud-Architecture-and-Requirements.pdf User manual www.eco4cloud.com/download/Eco4Cloud-User-Manual.pdf ValueProposition www.eco4cloud.com/download/Eco4Cloud-Value-Proposition-CUSTOMER.pdf White Paper www.eco4cloud.com/download/Eco4Cloud-white-paper.pdf Eco4Cloud Demo Video http://www.youtube.com/watch?v=FK6n8wh7pZw

More Related