ITRI Cloud OS Virtual Resource Management Patrick

Скачать презентацию ITRI Cloud OS Virtual Resource Management Patrick

083e2b0d3fd2e501441ad8eb0418e671.ppt

Количество слайдов: 26

What is Cloud OS ? • System Management Software layer – Physical Resource Provisioning – Virtual Resource Management • Improve manageability of massive Cloud Data Center • Enhance self-provisioning • Optimize physical resource utilization • High Availability for any single point of failure • Energy management – Highly Available Distributed Storage Management – Service Load Balancing – Security – High Speed Networking • Backup Virtual Cluster Mail Virtual Cluster HC Virtual Cluster App. X Virtual Cluster System Service daemons Compute Nodes Cloud OS agents Data Nodes Service Nodes What is it not? – It’s not Operating System – It’s not Virtualization Hypervisor Physical Node Copyright 2010 ITRI 業技術研究院 Physical Node Storage Storag Server e Server 3

Service/Technology Mapping Saa. S Paa. S Applications Automated Cloudification Technology Cloud Application Middleware Platform LAMP . NET Web. Sphere Web. Logic Google App Engine Cloud System Software Platform Hypervisor Virtualization Mgmt Storage Mgmt Security Backup/Replication Data Center Automation Energy Management Iaa. S Cloud Hardware Platform Scalable System Architecture Servers Storage Arrays System Management + Switches Copyright 2010 ITRI 業技術研究院 Cooling Power Distribution 4

Software Stack for Cloud OS Virtual Data. Center Mgmt Console Virtual Cluster Provisioning Physical Cluster Deployment Tool Power Management Physical Compute Servers Distributed Main/Secondary Storage System/Network Management Security All-layer-2 Network Intra-Virtual-Cluster Load Balancing Virtual Machine Management Copyright 2010 ITRI 業技術研究院 5

Cloud OS Service Model • Provisioning & Runtime monitoring of Virtual Resources – Virtual Instance Ø Hypervisor construct Ø An image of a guest OS – Virtual Cluster Ø A group of VM instances providing same service, front-ended by a network load balancer Ø Configuration - # of virtual machines and its configuration Storage space requirement External network bandwidth requirement Load balancing policy Firewall/IDS setting Network configuration, including DNS and DHCP OS image and application image – Virtual Data Center Ø One or more virtual cluster working in coordination (multi-tier web services, EMR’s, VDI’s, etc) Copyright 2010 ITRI 業技術研究院 6

Cloud. OS Virtualization Level VDC VCluster … OS OS Copyright 2010 ITRI 業技術研究院 … vm APs OS PM APs OS OS APs … vm vm vm OS vm APs OS OS PM APs … vm vm vm OS OS vm APs OS APs PM vm vm vm OS OS … OS APs APs vm vm vm VCluster PM 7 Cloud. OS

Resource Provisioning • To prepare VMs with appropriate resources and make them ready for user applications – Allocating resources to VMs to match the workloads • To prepare a virtual cluster with appropriate instances and make it ready for virtual cluster computation – Consolidating VMs onto physical servers • Goals: – High resource utilization – Energy efficiency – Low performance interference Copyright 2010 ITRI 業技術研究院 8

Provisioning Challenges • VM size estimation – Static SLA model/forecasting future use • Placement – Deploy a VM onto physical servers (initially) – Policy: immediate, best effort, advance reservation, etc. • Consolidation and Load Balancing – High consolidation ration and resource utilization – low cost of running data center – Statically • Heuristic based • Average resource utilization – Dynamic replacement • Measure-Forecast-Remap (MFR) • Live migration • Balancing overloaded and underloaded nodes – Constrained bin packing problem w/ SLA • Performance isolation – Cohosting VMs on a server creates performance interference – How to model and prevent the interference Copyright 2010 ITRI 業技術研究院 9

RPM Static Resource Provisioning • Statically provision from SLA • SLA w/ historical data? – No, conservatively allocation – Yes, forecasting joint-VM provisioning • Immediate provisioning model (before instantiation of virtual machine) • Placement policy – Proprietary – Virtual cluster affinity placement policy • Performance gain from locality • Place VMs from the same virtual cluster as possible • Need experiments to support Copyright 2010 ITRI 業技術研究院 10 Cloud. OS

Joint Provisioning via VM Multiplexing • Dataset from a commercial data center – 15, 897 VMs – 1325 physical hosts • 94% of the hosts have more than one VM • Joint provisioning averagely saves 40% of the capacity Copyright 2010 ITRI 業技術研究院 12 Meng, X. et al. Efficient Resource Provisioning in Compute Clouds via VM Multiplexing. ICAC ‘ 10

Load balancing and DVMM PM reconfiguration Reach reconfiguration point Resource Provisioning Manager (RPM) Utilization ratio Consolidation manager/ DVMM VM victim histogram Joint-VM histogram Over provisioning Under provisioning PM PM Copyright 2010 ITRI 業技術研究院 PM PM 14 Cloud. OS

Cloud OS RPM Software Components Adaptive physical resource provisioning Utilization rate reaches threshold, sending reallocation request Static joint-VM provisioning Reconfiguration map Placement RPM core VM monitoring New PM map Consolidation manager Victims Runtime joint-VM provisioning Performance interference Utilization change Power on/off PMs New PM map PRM DVMM Copyright 2010 ITRI 業技術研究院 15 Cloud OS

Adaptive Physical Resource Provisioning Power Management PM Pool Provisioned Utilization threshold Over provisioning Low utilization Under provisioning High utilization PM reallocation algorithm Utilization rate reaches threshold, sending reallocation request Reconfiguration map VM placement DVMM live migration Load balancer Consolidation manager New PM map Utilization monitor Power on/off PMs PRM Copyright 2010 ITRI 業技術研究院 Cloud. OS 20

Challenges • Triggering mechanism – – No workload consolidation “recently” (e. g. past hour) No physical machine load balancing going on No physical server was powered on “recently” (e. g. past hour) Avoid oscillation • Cost of migration – Network load – Cache effects – Domain in suspension • Multi-dimensional bin packing – CPU, Memory, Network, Disk I/O • Migration plan – Only 1 migration per Physical server @ a time – # of cores vs. # of VMs Copyright 2010 ITRI 業技術研究院 21

Procedure of Power management in Monitoring thread Receiving CCinstance data per 30 sec Calculate inst. Monitor. Thread. Data->cputotal inst. Monitor. Thread. Data->memorytotal inst. Monitor. Thread. Data->count Receive data from Dom 0 Calculate the data Workload Trigger? No Check inst. Monitor. Thread. Data->state inst. Monitor. Thread. Data->pwmtime inst. Monitor. Thread. Data->lbtime inst. Monitor. Thread. Data->cputotal inst. Monitor. Thread. Data->memorytotal inst. Monitor. Thread. Data->count Yes Perform Change instance state inst. Dvmm. Bloc->state=doinglb inst. Monitor. Thread. Data->state=doinglb inst. Dvmm. Bloc->dest. Host=resource->host. Name Change external state Call PRM to turn on Machine Consolidation Plan Stop receiving data from Dom 0 No Do Load balancing M*K < N ? If necessary Yes Do Consolidation Call PRM to shut down Copyright 2010 ITRI 業技術研究院 Done Machine Change instance state inst. Dvmm. Bloc->state=doingpwm inst. Monitor. Thread. Data->state=doingpwm inst. Dvmm. Bloc->dest. Host=resource->host. Name Change external state Change instance state inst. Dvmm. Bloc->pwmtime=now inst. Dvmm. Bloc->state=pwmdone 24 inst. Monitor. Thread. Data->state=pwmdone inst. Monitor. Thread. Data->pwmtime=now

Data Mining: VM Resource Usage Patterns of each VC • Find VM resource usage patterns for each VC CPU usage Monday High (or unpredictable) Medium Low • L Aid to predict the trend of resource usages Copyright 2010 ITRI 業技術研究院 Time 25