Abstract
In order to make save and restore features practical, saved virtual machines (VMs) must be able to quickly restore to normal operation. Unfortunately, fetching a saved memory image from persistent storage can be slow, especially as VMs grow in memory size. One possible solution for reducing this time is to lazily restore memory after the VM starts. However, accesses to unrestored memory after the VM starts can degrade performance, sometimes rendering the VM unusable for even longer. Existing performance metrics do not account for performance degradation after the VM starts, making it difficult to compare lazily restoring memory against other approaches. In this paper, we propose both a better metric for evaluating the performance of different restore techniques and a better scheme for restoring saved VMs.
Existing performance metrics do not reflect what is really important to the user -- the time until the VM returns to normal operation. We introduce the time-to-responsiveness metric, which better characterizes user experience while restoring a saved VM by measuring the time until there is no longer a noticeable performance impact on the restoring VM. We propose a new lazy restore technique, called working set restore, that minimizes performance degradation after the VM starts by prefetching the working set. We also introduce a novel working set estimator based on memory tracing that we use to test working set restore, along with an estimator that uses access-bit scanning. We show that working set restore can improve the performance of restoring a saved VM by more than 89% for some workloads.
- μswsusp. http://suspend.sourceforge.net/.Google Scholar
- AMD64 virtualization codenamed "pacifica" technology: Secure virtual machine architecture reference manual, May 2005. http://enterprise.amd.com/downloadables/Pacifica_Spec.pdf.Google Scholar
- D. F. Bacon, P. Cheng, and V. Rajan. A real-time garbage collector with low overhead and consistent utilization. In Proc. POPL '03, New Orleans, LA, USA, Jan. 2003. Google Scholar
Digital Library
- D. F. Bacon, P. Cheng, and V. Rajan. The Metronome, a simpler approach to garbage collection in real-time systems. In Proc. OTM 2003 Workshops, 2003.Google Scholar
Cross Ref
- S. Bansal and D. S. Modha. CAR: Clock with adaptive replacement. In Proc. FAST '04, 2004. Google Scholar
Digital Library
- G. E. Blelloch and P. Cheng. On bounding time and space for multiprocessor garbage collection. In Proc. PLDI '99, Atlanta, GA, USA, May 1999. Google Scholar
Digital Library
- C. Clark, K. Fraser, S. Hand, J. G. Hansen, E. Jul, C. Limpach, I. Pratt, and A. Warfield. Live migration of virtual machines. In Proc. NSDI '05, 2005. Google Scholar
Digital Library
- F. J. Corbato. A paging experiment with the Multics system. Technical report, MIT Project MAC, May 1969.Google Scholar
- P. J. Denning. The working set model for program behaviour. Commun. ACM, 11 (5), 1968. Google Scholar
Digital Library
- M. R. Hines and K. Gopalan. Post-copy based live virtual machine migration using adaptive pre-paging and dynamic self-ballooning. In Proc. VEE 2009, Washington, DC, USA, 2009. Google Scholar
Digital Library
- T. Holwerda. SuperFetch: How it works & myths, May 2009. http://www.osnews.com/story/21471/SuperFetch_How_it_Works_Myths.Google Scholar
- S. Jiang and X. Zhang. LIRS: an efficient low inter-reference recency set replacement policy to improve buffer cache performance. In Proc. SIGMETRICS '02, Marina del Rey, California, USA, 2002. Google Scholar
Digital Library
- S. Jiang, F. Chen, and X. Zhang. CLOCK-Pro: An effective improvement of the CLOCK replacement. In Proc. USENIX '05, 2005. Google Scholar
Digital Library
- R. E. Jones. Garbage Collection: Algorithms for Automatic Dynamic Memory Management. Wiley, Chichester, July 1996. URL http://www.cs.ukc.ac.uk/people/staff/rej/gcbook/gcbook.html. Google Scholar
Digital Library
- kernel-enhancements-xp. Kernel enhancements for Windows XP, jan 2003. http://www.microsoft.com/whdc/archive/XP_kernel.mspx.Google Scholar
- H. A. Lagar-Cavilla, J. A. Whitney, A. M. Scannell, P. Patchin, S. M. Rumble, E. de Lara, M. Brudno, and M. Satyanarayanan. SnowFlock: rapid virtual machine cloning for cloud computing. In Proc. Eurosys '09, Nuremberg, Germany, 2009. Google Scholar
Digital Library
- R. B. Miller. Response time in man-computer conversational transactions. In Proceedings of the fall joint computer conference, part I, AFIPS '68 (Fall, part I), pages 267--277, New York, NY, USA, 1968. ACM. Google Scholar
Digital Library
- E. J. O'Neil, P. E. O'Neil, and G. Weikum. The LRU-K page replacement algorithm for database disk buffering. In P. Buneman and S. Jajodia, editors, Proc. SIGMOD '93, 1993. Google Scholar
Digital Library
- C. P. Sapuntzakis, R. Chandra, B. Pfaff, J. Chow, M. S. Lam, and M. Rosenblum. Optimizing the migration of virtual computers. SIGOPS Operating Systems Review, 36, 2002. Google Scholar
Digital Library
- Seagate. Product Manual. Barracuda 7200.11 Serial ATA. http://www.seagate.com/staticfiles/support/disc/manuals/desktop/Barracuda 7200.11/100452348g.pdf, Jan. 2009.Google Scholar
- Standard Performance Evaluation Corporation. SPECjbb2005 User's Guide. http://www.spec.org/jbb2005/docs/UserGuide.html, April 2006.Google Scholar
- J. Sugerman, G. Venkitachalam, and B.-H. Lim. Virtualizing I/O devices on VMware Workstation's hosted virtual machine monitor. In Proceedings of the 2001 USENIX Annual Technical Conference, Boston, MA, USA, June 2001. Google Scholar
Digital Library
- N. Tolia, D. G. Andersen, and M. Satyanarayanan. Quantifying interactive user experience on thin clients. IEEE Computer, 39 (3), Mar. 2006. Google Scholar
Digital Library
- VMware. Timekeeping in VMware virtual machines. http://www.vmware.com/vmtn/resources/238, Aug. 2008.Google Scholar
Index Terms
Fast restore of checkpointed memory using working set estimation
Recommendations
Fast restore of checkpointed memory using working set estimation
VEE '11: Proceedings of the 7th ACM SIGPLAN/SIGOPS international conference on Virtual execution environmentsIn order to make save and restore features practical, saved virtual machines (VMs) must be able to quickly restore to normal operation. Unfortunately, fetching a saved memory image from persistent storage can be slow, especially as VMs grow in memory ...
Working Set Size Estimation Techniques in Virtualized Environments: One Size Does not Fit All
SIGMETRICS '18Energy consumption is a primary concern for datacenters' management. Numerous datacenters are relying on virtualization, as it provides flexible resource management means such as virtual machine (VM) checkpoint/restart, migration and consolidation. ...
Memory Reclamation and Compression Using Accurate Working Set Size Estimation
CLOUD '15: Proceedings of the 2015 IEEE 8th International Conference on Cloud ComputingMinimizing the total amount of physical memory consumption of the virtual machines (VM) concurrently running on a physical server is a key to improving a hyper visor's consolidation ratio, defined as the maximum number of VMs that can run on a physical ...







Comments