Vm Management For Green It Data Centers
3 likes1,239 views
The document introduces OpenNebula, an open-source tool for managing virtual machines across distributed infrastructures. It describes OpenNebula's model for managing VM lifecycles and scheduling VM workloads. It also outlines a prototype scheduler that aims to distribute VM workloads based on energy requirements to improve efficiency in data centers.
Vm Management For Green It Data Centers
- 1. VM Management for Green Data Centres with the OpenNebula Virtual Infrastructure Engine Ignacio M. Llorente OGF-EU: Using IT to reduce Carbon Emissions and Delivering the Potential of Energy Efficient Computing OGF25, Catania, Italy 5 March 2009
- 2. Objectives Workshop : Development of a reference model for the management of energy efficiency in virtualized distributed environments Presentation : Introduce the OpenNebula Virtual Infrastructure Engine Describe its management model for virtualized distributed environments Describe a first prototype of scheduler for distribution of VM workloads based on energy requirements VM Management for Green Data Centres with OpenNebula
- 3. What is OpenNebula? VM Management for Green Data Centres with OpenNebula Dynamic deployment and re-placement of virtual machines on a pool of physical resources Transform a rigid distributed physical infrastructure into a flexible and agile virtual infrastructure Extending the Benefits of Virtualization to Clusters Private Cloud : Virtualization of cluster or data-center for internal users Backend of Public Cloud : Internal management of the infrastructure Cloud Interoperation : On-demand access to public clouds ONE Core Scheduler XEN / KVM Plugins Cloud Plugins OpenNebula Local Infrastructure Public Cloud Interface Internal users External users Public Cloud
- 4. Virtual Machine Management Model VM Management for Green Data Centres with OpenNebula Service structure Service components run in VMs Inter-connection relationship Placement constraints The VM Manager is service agnostic Provide infrastructure context Service as Management Entity Distributed V M Management System The three pillars of a Distributed VM Manager Distributed VM Management Model Networking Image Management Virtualization
- 5. Benefits Centralized management of VM workload and distributed infrastructures Support for VM placement policies : balance of workload, server consolidation… Dynamic resizing of the infrastructure Dynamic partition and isolation of clusters Support for heterogeneous workload Dynamic scaling of private infrastructure to meet fluctuating demands VM Management for Green Data Centres with OpenNebula System Manager Service Manager On-demand provision of virtual machines System Integrators Open and flexible architecture and interfaces, open source software Integration with any component in the virtualization/cloud ecosystem, such as cloud providers, hypervisors, cloud-like interfaces, virtual image managers, service managers, schedulers…
- 6. Features VM Management for Green Data Centres with OpenNebula Feature Function User Interface Unix-like CLI to manage VM life-cycle and physical boxes XML-RPC API and libvirt interface Scheduler Requirement/rank matchmaker Generic framework to build any scheduler Virtualization Management Xen, KVM and libvirt connectors Amazon EC2 Image Management General mechanisms to transfer and clone VM images Network Management Definition of virtual networks to interconnect VMs Fault Tolerance Persistent database backend to store host and VM information Scalability Tested in the management of hundreds of VMs Installation Installation on a UNIX cluster front-end without requiring new services in the remote resources Distributed in Ubuntu 9.04 (Jaunty Jackalope), due to be released in April 2009
- 7. Open and Flexible Architecture VM Management for Green Data Centres with OpenNebula Scheduler CLI libvirt sqlite Persistent Database XML-RPC OpenNebula core Information Manager Image Manager VM Manager Network Manager Host, Net & VM Pools OpenNebula API Access Drivers XEN KVM EC2 libvirt
- 8. Use Cases On-demand Scaling of Computing Clusters On-demand Scaling of Web Servers Elastic execution of a SGE computing cluster Dynamic growth of the number of worker nodes to meet demands using EC2 Private network with NIS and NFS EC2 worker nodes connect via VPN VM Management for Green Data Centres with OpenNebula Elastic execution of the NGinx web server The capacity of the elastic web application can be dynamically increased or decreased by adding or removing NGinx instances
- 9. Ecosystem Schedulers VM Management for Green Data Centres with OpenNebula Haizea: Open-source VM-based lease management architecture (allows AR of capacity). Interfaces Libvirt: Provides an abstraction of a whole cluster of resources as one host, hiding specific hypervisor details. Nimbus: C an be used as a WSRF or EC2 front-end. Plug-Ins ElasticHosts: Enables the dynamically increase capacity of your virtualized infrastructure to meet fluctuating peak demands using a cloud provider.
- 10. Model for VM Management in Distributed Infrastructures VM Management for Green Data Centres with OpenNebula VM Virtual Infrastructure Manager VM VM VM Scheduler Monitoring & Accounting Controlling (CLI/API) Centralized management of VMs and resources VM life-cycle management VM image management Virtual network management Fault tolerance Automation of VM placement: Balance of workload Server consolidation Placement constraints and affinity Advance reservation of capacity SLA commitment Energy Policy Enactment Energy Policy Definition Physical Box VM Manager Physical Box VM Manager Physical Box VM Manager Physical Box VM Manager
- 11. Monitoring and Accounting VM Management for Green Data Centres with OpenNebula VM Life-cycle and Monitoring Attribute Description ID VM identifier NAME Name of the VM STAT Status CPU CPU percentage used by the VM MEM Memory used by the VM HOSTNAME Host where the VM is running TIME Time since submission
- 12. Monitoring and Accounting VM Management for Green Data Centres with OpenNebula Resource Monitoring Flexible Architecture Easy addition of new VM and resource metrics (energy efficiency metrics) Experimenting with Advanced Configuration and Power Interface (ACPI) for power monitoring Accounting Statistics of host usage, VM image transfer times…. Attribute Description HID Host ID NAME Host name RVM Number of running VMs TCPU Total CPU (percentage) FCPU Free CPU (percentage) ACPU Available CPU (not allocated by VMs) TMEM Total memory FMEM Free memory STAT Host status
- 13. Controlling VM Management for Green Data Centres with OpenNebula VM Controlling Resource Controlling Action Description create Submits a new virtual machine, adding it to the VM pool deploy Starts a previously submitted VM on a specific host shutdown Shutdown an already deployed VM livemigrate Migrates a running VM to another host without downtime migrate Saves a running VM and starts it again in the specified host stop Stops a running VM Release/hold Releases/hold a VM from/to hold state Suspend/resume Saves/resume a VM Action Description Create/delete Adds/removes a machine to/from the pool Enable/disable Enables/disables host
- 14. Resource Schedulers VM Management for Green Data Centres with OpenNebula Advance reservation of capacity and queuing of best effort requests Haizea Lease Manager (University of Chicago) Policy-driven probabilistic admission control and dynamic placement optimization to satisfy site level management policies RESERVOIR Policy Engine (IBM Haifa/Elsa Datamat) Requirement/rank Matchmaker (default) First placement of pending VMs to resources meeting the “requirements” and sorted using a “rank“ expression Periodic re-placement of VMs for server consolidation and suspension/resume of physical resources Experimental for supporting research on distribution of VM workload based on energy requirements and policies VM Consolidation Scheduler
- 15. VM Consolidation Scheduler VM Management for Green Data Centres with OpenNebula Control Flow Complements the work done by the requirement/rank matchmaker for first deployment of VMs in pending state VM pending? Enable host Disable free hosts Consolidation? Packing algorithm Migration yes yes Period: 20 minutes Physical Resource Management Virtual Machine Management
- 16. VM Consolidation Scheduler VM Management for Green Data Centres with OpenNebula A Simple Example Initial situation Exec. 1: Packing algorithm in the VM Consolidation Scheduler requests a migration Exec. 2: VM Consolidation Scheduler “disables” a physical resource Exec. 3: VM consolidation scheduler “enables” a physical resource because there are pending VMs The requirement/rank matchmaker schedules the pending VMs
- 17. VM Consolidation Scheduler VM Management for Green Data Centres with OpenNebula A Simple Example
- 18. VM Consolidation Scheduler VM Management for Green Data Centres with OpenNebula Future Work: Energy Policy-driven Scheduler Monitoring Power and heat metrics (ACPI specification) Past history (trends) of resources and workload Enactment of policies Management of power-state of the resources (ACPI specification) Policies Support for “generic energy policies” Selection of the VMs to migrate and the target host for migration …
- 19. Conclusions VM Management for Green Data Centres with OpenNebula Workshop : Development of a reference model for the management of energy efficiency in virtualized distributed environments Presentation : Introduce the OpenNebula Virtual Infrastructure Engine Describe its management model for virtualized distributed environments Describe a first prototype of scheduler for distribution of VM workloads based on energy requirements
- 20. THANK YOU FOR YOUR ATTENTION!!! More info, downloads, mailing lists at www.OpenNebula.org Real demo at booth number 4 Acknowledgements Rubén S. Montero Javier Fontan Raúl Sampedro Tino Vazquez Rafael Moreno VM Management for Green Data Centres with OpenNebula