300 likes | 506 Views
Christopher Clark, Keir Fraser, Steven Hand, Jacob Gorm Hansen, Eric Jul, Christian Limpach, Ian Pratt, Andrew Warfield University of Cambridge Computer Laboratory, Department of Computer Science University of Copenhagen Presentation by Marty Krogel. Live Migration of Virtual Machines.
E N D
Christopher Clark, Keir Fraser, Steven Hand, Jacob Gorm Hansen, Eric Jul, Christian Limpach, Ian Pratt, Andrew Warfield University of Cambridge Computer Laboratory, Department of Computer Science University of Copenhagen Presentation by Marty Krogel Live Migration of Virtual Machines
Outline • Goal • Why VM migration? • Related works • Design Choices • Implementation • Handling Local Resources • Design Overview • Testing the Implementation • Benchmarks • Future Work • Conclusion
To find a quick and efficient way to transfer services between physical servers. Goal
Challenges • Minimizing downtime. • Keeping total migration time down. • Avoid disrupting active services.
Benefits of Migrating Virtual Machines Instead of Processes • Avoids 'residual dependencies'. • Can transfer in-memory state information. • Allows separation of concern between users and operator of a data center or cluster.
Related Work • The Collective project • Zap • NomadBIOS • VMotion • Process migration • Sprite, MOSIX • Java and .NET even suffer
Memory Migration Options • Push phase • Stop-and-copy phase • Pull phase
Implementation • Pre-copy migration • Bounded iterative push phase • Rounds • Writable Working Set • Short stop-and-copy phase • Careful to avoid service degradation
Handling Local Resouces • Open network connections • Migrating VM can keep IP and MAC address. • Broadcasts ARP new routing information • Some routers might ignore to prevent spoofing • A guest OS aware of migration can avoid this problem • Local storage • Network Attached Storage
Tracking Writable Working Set • Xen inserts shadow pages under the guest OS, populated using guest OS's page tables. • The shadow pages are marked read-only. • If OS tries to write to a page, the resulting page fault is trapped by Xen. • Xen checks the OS's original page table and forwards the appropriate write permission. • If the page is not read-only in the OS's PTE, Xen marks the page as dirty.
Implementation Issues • Managed Migration vs. Self Migration • Dynamic Rate-Limiting • Rapid Page Dirtying • Paravirtualized Optimizations • Stunning Rogue Processes • Freeing Page Cache Pages
Test Setup • 2 Dell PE-2650 server-class machines • Dual Xeon 2GHz CPUs • Only used 1 CPU • HyperThreading enabled • 2GB Memory • Broadcom TG3 network interface • Gigabit Ethernet switch • NetApp F840 Network attached storage server using iSCSI protocol • XenLinux 2.4.27 OS
By integrating live OS migration into the Xen virtual machine monitor, rapid movement of interactive workload across a cluster or data center is possible. Using pre-copying and dynamically adapting network-bandwith total downtime can be reduced to imperceptible levels, even for complex interactive services. Conclusion
Future Work • Cluster Management • Wide Area Network Redirection • Migrating Block Devices