Cloud-Assisted VR.

Slides:



Advertisements
Similar presentations
Ravi Sankar Technology Evangelist | Microsoft
Advertisements

The Who, What, Why and How of High Performance Computing Applications in the Cloud Soheila Abrishami 1.
Green Cloud Computing Hadi Salimi Distributed Systems Lab, School of Computer Engineering, Iran University of Science and Technology,
Efficient Autoscaling in the Cloud using Predictive Models for Workload Forecasting Roy, N., A. Dubey, and A. Gokhale 4th IEEE International Conference.
COMMA: Coordinating the Migration of Multi-tier applications 1 Jie Zheng* T.S Eugene Ng* Kunwadee Sripanidkulchai† Zhaolei Liu* *Rice University, USA †NECTEC,
Gueyoung Jung, Nathan Gnanasambandam, and Tridib Mukherjee International Conference on Cloud Computing 2012.
A Congestion Pricing User Study Using a a Wireless LAN Jimmy Shih, Randy Katz, Anthony Joseph.
By- Jaideep Moses, Ravi Iyer , Ramesh Illikkal and
CSE598C Project: Dynamic virtual server placement Yoojin Hong.
Cloud Computing All Copyrights reserved to Talal Abu-Ghazaleh Organization
Utility Computing Casey Rathbone 1http://cyberaide.org.edu.
A User Experience-based Cloud Service Redeployment Mechanism KANG Yu.
Dynamic and Decentralized Approaches for Optimal Allocation of Multiple Resources in Virtualized Data Centers Wei Chen, Samuel Hargrove, Heh Miao, Liang.
Virtual Machine Hosting for Networked Clusters: Building the Foundations for “Autonomic” Orchestration Based on paper by Laura Grit, David Irwin, Aydan.
Bargaining Towards Maximized Resource Utilization in Video Streaming Datacenters Yuan Feng 1, Baochun Li 1, and Bo Li 2 1 Department of Electrical and.
A Cloud is a type of parallel and distributed system consisting of a collection of inter- connected and virtualized computers that are dynamically provisioned.
Virtual Machine Course Rofideh Hadighi University of Science and Technology of Mazandaran, 31 Dec 2009.
Virtualization. Virtualization  In computing, virtualization is a broad term that refers to the abstraction of computer resources  It is "a technique.
OPTIMAL SERVER PROVISIONING AND FREQUENCY ADJUSTMENT IN SERVER CLUSTERS Presented by: Xinying Zheng 09/13/ XINYING ZHENG, YU CAI MICHIGAN TECHNOLOGICAL.
Network Aware Resource Allocation in Distributed Clouds.
Cloud Computing Energy efficient cloud computing Keke Chen.
Storage Management in Virtualized Cloud Environments Sankaran Sivathanu, Ling Liu, Mei Yiduo and Xing Pu Student Workshop on Frontiers of Cloud Computing,
IISWC 2007 Panel Benchmarking in the Web 2.0 Era Prashant Shenoy UMass Amherst.
LOGO Service and network administration Storage Virtualization.
Cloud Resource Scheduling for Online and Batch Applications Kick-off meeting.
BOF: Megajobs Gracie: Grid Resource Virtualization and Customization Infrastructure How to execute hundreds of thousands tasks concurrently on distributed.
The Only Constant is Change: Incorporating Time-Varying Bandwidth Reservations in Data Centers Di Xie, Ning Ding, Y. Charlie Hu, Ramana Kompella 1.
Performance Analysis of Preemption-aware Scheduling in Multi-Cluster Grid Environments Mohsen Amini Salehi, Bahman Javadi, Rajkumar Buyya Cloud Computing.
Steven Adler Enterprise Technology Strategist Microsoft EMEA.
Capacity Planning in a Virtual Environment Chris Chesley, Sr. Systems Engineer
Architecture for Resource Allocation Services Supporting Interactive Remote Desktop Sessions in Utility Grids Vanish Talwar, HP Labs Bikash Agarwalla,
Unit 2 VIRTUALISATION. Unit 2 - Syllabus Basics of Virtualization Types of Virtualization Implementation Levels of Virtualization Virtualization Structures.
© 2012 Eucalyptus Systems, Inc. Cloud Computing Introduction Eucalyptus Education Services 2.
Extreme Scale Infrastructure
NFV Group Report --Network Functions Virtualization LIU XU →
Md Baitul Al Sadi, Isaac J. Cushman, Lei Chen, Rami J. Haddad
VPN Extension Requirements for Private Clouds
LIGHTWEIGHT CLOUD COMPUTING FOR FAULT-TOLERANT DATA STORAGE MANAGEMENT
SPLA Licensing Overview
Examples based on draft-cheng-supa-applicability-00.txt
Memshare: a Dynamic Multi-tenant Key-value Cache
CHT Project Progress Report
Dedicated Servers vs Cloud Hosting
ATLAS Cloud Operations
6WIND MWC IPsec Demo Scalable Virtual IPsec Aggregation with DPDK for Road Warriors and Branch Offices Changed original subtitle. Original subtitle:
Building a Virtual Infrastructure
Job Scheduling in a Grid Computing Environment
CS 425 / ECE 428 Distributed Systems Fall 2016 Nov 10, 2016
AWS Solutions Architect Presentation
Virtual media? A new reality
VIDIZMO Deployment Options
CS 425 / ECE 428 Distributed Systems Fall 2017 Nov 16, 2017
Computing Resource Allocation and Scheduling in A Data Center
CHAPTER 1 INTRODUCTION:
Cloud-Assisted VR.
Memory Management for Scalable Web Data Servers
NPAR Dell - QLogic October 2011.
Buy Exact IBM C Exam Questions With Answers - C Dumps PDF Dumps4Download
3.2 Virtualisation.
Vlad Nae, Radu Prodan, Thomas Fahringer Institute of Computer Science
Scheduling Algorithms in Broad-Band Wireless Networks
Cloud computing mechanisms
AWS Cloud Computing Masaki.
Technical Capabilities
Cloud Computing Architecture
Developing for Windows Azure
Use Case #1: Mobile Virtual Desktop
Cloud Resource Scheduling for Online and Batch Applications
Requirements of Computing in Network
Progress Report 04/27 Simon.
Presentation transcript:

Cloud-Assisted VR

Introduction Most of the virtual reality applications are designed for single user. Require high-end desktop or mobile phone in order to provide good QoS. Recently, social VR is emerging as a popular social media. Support multiple users. Cloud servers are required to coordinate the states and interactions between users.

Social VR Users interact with each others with their avatars in a virtual “chat room”. With voice and (mostly) the movements of head and hands. Examples: vTime: https://www.youtube.com/watch?v=cUI6C_0T5_U Facebook social VR: https://www.youtube.com/watch?v=W71o4RbckNA

Social VR in the Cloud There are multiple VR chat rooms in a data center. Each VR chat room is independent and should be isolated with the others. VR chat rooms may have different resource requirements according to the number of connected users and virtual objects created. We apply container technique to host VR chat rooms in the data center.

Container Technique A light-weight operating-system-level virtualization method. Each container is an isolated user-space instance. Compared to virtual machine Lower booting overhead. More flexible in resource allocation. VM:根據使用者設定的規格分配資源,沒用到的資源無法讓別人使用。(獨佔) Container:使用者設定的規格為資源使用“上限”,根據實際使用資源量來分配。還沒用到的資源可以分給別人使用。

Scenario Given containers hosting VR chat rooms, we need to dynamically deploy these containers according to their resource requirements, in order to maintain the performance / QoS. The resource required by a container depends on the number of connected users and virtual objects created in the VR chat rooms.

Scenario(Cont.) 1. “Create” Request from User Application manager VR Chat room Application manager 1. “Create” Request from User Resource manager 2. Prepare a container hosting VR chat room 3. Deploy to server and start the container Deploy strategy

Objective Guarantee the performance / QoS while maximizing the system resource utilization. Use penalty to estimate the performance. Minimize the total cost. Allocating insufficient resources to a VR chat room leads to performance degradation, which generates penalty. Low resource utilization increases the operating cost of a data center.

Container Deployment Strategies We propose five strategies for deploying a container to a server. A. Best-Fit B. Next-Fit (Round-Robin + First-Fit) C. Worst-Fit D. Spread E. Age-Based 將 container 指派到可利用資源與目前需求最接近的 server 上 將 container 指派到 server list 中下一個可利用資源大於目前需求的 server 上 將 container 指派到可利用資源大於目前需求,且差距最大的 server 上 將 container 指派到已運行 container 數量最少且可利用資源足夠的 server 上。若平手則隨意挑 將 container 指派到資源足夠的 server中,已運行 container 開啟時間最久的 server 上

Resource Allocation for Containers Since the resource requirement of a container varies through time, we also consider different resource requirements during the deployment of a container. I. According to container specification II. According to actual usage III. According to actual usage + reservation 依照 container 規格要求資源。 Server 會依照 container 規格,保留資源給該 container,即使目前實際需求還不到這麼多。 依照目前實際需求要求資源 要求略大於目前實際需求的資源

Simulations We conduct some simulation to compare the performance and resource utilization of the above mentioned strategies. Setting: Server specification: (Core, Memory, Network Bandwidth) : (16, 10GB, 1GBits/s) Container specification: (Core, Memory, Network Bandwidth) : (4, 1GB, 256MBit/s)

Workloads We generate 5 sets of workload traces. Each trace simulates the changes in the number of users and virtual objects of a VR chat room. A newly created container will choose 1 of these 5 traces as its workload trace before being deployed to server.

Simulation Results

Summary We apply social VR as our targeting scenario in cloud-assisted VR. Each container hosts a virtual chat room. The resource requirement of each container is related to the number of users and virtual objects, which varies through time. We propose five deployment strategies in order to optimizing the performance of applications and resource utilization of data center.

Summary(Cont.) The simulation results show that … We will apply different social VR applications to enrich the workloads in the data center in the next year. Also, we will consider the heterogeneity of servers.