aki

Document Sample
aki Powered By Docstoc
					CoreLab: An Emerging Network Testbed
    towards Network Virtualization

     Network Virtualization Research Lab


            Akihiro NAKAO
           Associate Professor
           University of Tokyo
                 NICT
PlanetLab

     The largest and most popular overlay network test-bed

     Currently consists of 900+ nodes at 450+ sites

     800+ Projects/ 1000+ Researchers

     180+ high-level academic publications enabled




                                                        2
      09-3-2             Akihiro NAKAO, APAN 2009
Brief History of PlanetLab

   PlanetLab 1.0 (2002-3)
   
    UNIX account slivers

   PlanetLab 2.0 (2003-4)
   
    Vserver slivers

   PlanetLab 3.0 (2004-2006)
   
    PLCAPI 1.0

   PlanetLab 4.0 (2007-)
   
    MyPLC1.0
   
    Federation Idea

   PlanetLab 4.2 (2008-9)          We are here

   
   Federation (PLE, PLJ, (PLK, PLC,=>PLA?))
   
   New Development using Private PlanetLab
    (RSpec/GENI-wrapper, New Node Type, Monitoring, QA,VINI)
                                                         

What have we learned from PlanetLab?

  Concept
  
   Slice-based   experimental facility

  “Slice” Mechanism
  
   Resource   Isolation through Virtualization

  Management Framework
  
   Node   : Physical Machine
  
   Sliver : Virtual Machine
  
   Slice : A collection of Virtual Machines
  
   RSpec: Resource Description
  
   PLC: Central Management Servers
   09-3-2                Akihiro NAKAO, APAN 2009   

Slivers
                Sliver
 Sliver
 Sliver
              Sliver

                                                                Auditing service
     Node
      Owner
                                          Monitoring services
                          VM1
      VM2
     …
      VMn

     Mgr
        VM
                                            Brokerage services
                                                                Provisioning services



       Virtual Machine Monitor (VMM)



                                 Linux kernel (Fedora Core)
                                 + Vservers (namespace isolation)
                                 + Schedulers (performance isolation)
                                 + VNET (network virtualization)

             - From http://www.planet-lab.org/files/presentation-2007-05-01-planetlab.ppt


  09-3-2                         Akihiro NAKAO, APAN 2009                         

PlanetLab Architecture
                                         Slice1
 Slice2





                                         Sliver
   Sliver





           PLC

                             Node





                                         Sliver
   Sliver





                             Node


  09-3-2          Akihiro NAKAO, APAN 2009                   

Limitations in PlanetLab

  Flexibility is sacrificed over Scalability/Performance
  
   Sliver: Container-based Resource Isolation
  
   Shares crucial resources
      •  Single kernel (Linux with VServer patch)
      •  Network stack and modules (NAT, Firewall, etc)
      •  Network resources (IP address and ports)
      •  Vnet (virtual network device ; need work for multi-homing)
  
   Cannot   run arbitrary kernel/network stack & modules

  Software Engineering Issue
  
   Cannot   keep up with the latest kernel (VServer)
      •  The latest VServer patch 2.6.26 (PlanetLab 2.6.22)
      •  Misses out support for new devices/chip features
    09-3-2                Akihiro NAKAO, APAN 2009               

Design Principle for a New Test-bed

   Performance
   
    Network I/O

   Scalability
   
    The number of slices

   Isolation
   
    Resource must be isolated per slice

   Flexibility
   
    Arbitrary kernel/network stacks/resources

   Code-Reusability                                   Largely Disregarded


   
    Catch up with the latest technology
   
    Minimize software engineering effort

    09-3-2                  Akihiro NAKAO, APAN 2009                 

VEE Candidate for Sliver Implementation
                                       Manager
     Apps
      apps
                     Apps       Apps
                  Apps        Apps     MngOS
       GuestOS
   GuestOS
                  GuestOS    GuestOS
  Apps            VE
         VE
      VM
          VM
        VM
            Apps       VM
        VM

  HostOS                               VMM
                                   HostOS

  Hardware                             Hardware
                              Hardware

          Resource Container                 Hypervisor-Based                            Host-Based

Design     Performance Scalability                 Security Performance Flexibility Code
Principles                                         Isolation Isolation              Reusability

Resource                 +             +                                  +
Container         advantageous

                                                                                         -      disadvantageous

 Hyper-                  +                             +                  +              +
  visor
                                                                                                          -
 Hosted-                  ?            ?               +                  ?              +                +
 VMM
                         room for improvement

         09-3-2                            Akihiro NAKAO, APAN 2009                                        

CoreLab

  Design
           VEE : Hosted Virtual Machine Monitor
  
   Sliver
  
   VM Disk Image Transfer : Multipath & Caching
  
   Network Configuration : Multihoming
  
   Management Tools: Reuse PlanetLab’s design
               (GENI-wrapper/Federation possible)
   Comparison Chart          PlanetLab                   CoreLab
   Sliver                    Resource Container          Hosted VMM
   Kernel                    Shared                      Separated
   Guest OS                  Various Linux Distros       Any OSes
   IP & Port                 Shared                      Shared (managed)
   Privileged System Calls   Shared                      Isolated

   09-3-2                     Akihiro NAKAO, APAN 2009                      

 CoreLab Prototype

  Implementation
           VEE : KVM + QEMU on Kernel 2.6.28.7
  
   Sliver
  
   VM Disk Image Transfer : BitTorrent
  
   Network Configuration : TunTap / Managed IP/Port
  
   Management Tools: MyPLC
  
   Optimizations
      •  Compression of VM Image
      •  Increase Sharable Buffers
      •  Decrease Influence of VM Boots
  
   Misc.
      •  SSH Login (permitopen, authorized_key) and VNC
      •  Compatibility with PlanetLab (PLCAPI + CoreLab APIs)
    09-3-2                Akihiro NAKAO, APAN 2009              

KVM Hosted Virtualization
Full/Para Virtualization (Virtio)




   09-3-2           Akihiro NAKAO, APAN 2009   

Network Configuration




           PlanetLab Model
                              CoreLab Model

  09-3-2                      Akihiro NAKAO, APAN 2009                    

VM Image Dissemination


                                        Introduce helper nodes
                                        •  BT Node Master
                                        •  BT Node Slave

                                        Introduce components
                                        •  BTManager
                                        •  BTSeed




  09-3-2     Akihiro NAKAO, APAN 2009                       

Scalability Check (1/2) pkt-size=64B
Total packet receive rate (Kpps)




                                                Probably our system is CPU bound
                                                QEMUs nicely schedule by Linux Kernel





                      Max at 48kpps (slightly worse than Xen by Trellis)
                      Little overhead in term of # of VMs (16 – 224 VMs)
                      Great resource fairness among VMs (~1.8% deviation)

                            Offered Load via pktgen (Kpps)
     09-3-2                   Akihiro NAKAO, APAN 2009                          

CoreLab Deployment (on-going)

  Phase1 : 12 Sites over JGN2Plus + SINET + others
            10 in operation

  Phase2 : Asian Region




                                 HP DL 580 (4U)
                                 Intel Xeon 16-core (quad 4-core)
                                 128GB (Max 256GB) memory
                                 300GB RAID5
                                 (accommodate up to 254 slices)
  09-3-2          Akihiro NAKAO, APAN 2009                    

CoreLab : In Action…




             http://www.corelab.jp


  09-3-2      Akihiro NAKAO, APAN 2009   

Conclusions

  Contributions
  
   Identify   design principles for alternative test-beds
     •  Performance, Scalability, Isolation
     •  Flexibility, Code-Reusability
  
   Implement     prototype test-bed CoreLab
     •  Hosted VMM as sliver to achieve
  
   Report   various optimization and preliminary evaluations
     •  Throughput 30-70% of Vanilla Linux and PlanetLab OS

  Future Work
  
   Build Full-Fledged CoreLab
  
   Optimize performance and scalability
                         Hosted VMM
      Virtual Router via Akihiro NAKAO, APAN 2009
  
 09-3-2                                                     

More Future Directions

  CoreLab Enhancement
  
   Geographically   (Japan to Asia)

  Public PlanetLab
  
   J-Lab   as a regional Public PlanetLab

  Federation
             CoreLab, J-Lab with PlanetLab, OneLab2
  
   Federate
  
   GENI Spiral 1 Solicitation 2


  Virtual Router Project
  
   ExtremeEnhancement to CoreLab
  
   OpenFlow / NetFPGA
  
   UTokyo, NICT, NEC, Hitachi, NTT, Fujitsu        

Overlay to Network Virtualization
                                                                     A Slice of
                                                                Computation Resources

   Virtualize not only edges…
                                                  PC Server

   …but also network core




                                                              Mgmt S/W
                                                                         VM(Slice)
                                                                                     VM(Slice)



                                                                                                 VM(Slice)
                                                   switch           VMM/Hypervisor




   PlanetLab to GENI…Similar efforts in several countries…
Planetary-Scale Network Virtualization
                                                Internet-Wide / ISP-Wide
                                                         Virtual Links




                       Virtual Core
                         Routers



                                                                         Virtual Edge
                                                                           Routers

  Virtual End System


                       Substrate Link (could be any exisiting technology)
Roadmap
       2007
          2008
             2009
             2010
     2011

                              Nov.


                         Deploy(1st Phase)
 Federation
                     Fiscal Year

Research PlanetLab

                   CoreLab: PlanetLab Enhancement + VR
                                         Federate
Private PlanetLab
            Deploy


                                 J-lab : Private PlanetLab
                                 (JGN2+SINET+…)
Public PlanetLab

                     Public PlanetLab (Federation; Migration)


                   Network Virtualization Lab
                   (CoE on Network Virtualization)
                            AKARI Architecture Project                  2nd Phase
                        Service Platform Advanced Research Center


   09-3-2   JGN2                          JGN2plus                  New Testbed
                                                                             

Contact Information

   Network Virtualization / Overlay is one of the key
    technologies for proceeding further in defining NwGN
    architecture(s)

   Test-bed designs are crucial part of this research…
            Net-Virtualization Research Lab
             Contact: nv-lab@ml.nict.go.jp
                 http://www.nvlab.org



   09-3-2             Akihiro NAKAO, APAN 2009        23

				
DOCUMENT INFO
Shared By:
Categories:
Stats:
views:58
posted:3/5/2010
language:English
pages:23