PAPI-V: Performance Monitoring for Virtual Machines

Johnson, Matthew; McCraw, Heike; Moore, Shirley; Mucci, Philip J.; Nelson, John; Terpstra, Dan; Weaver, Vince; Mohan, Tushar

doi:10.1109/icppw.2012.29

Cited by 16 publications

(16 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…During the execution of co-located programs, we can identify if a program is data-intensive or compute-intensive based on the number of instructions and the number of memory accesses that can be obtained easily through performance counter or performance application programming interface (PAPI) [14]. During the execution of co-located programs, we can identify if a program is data-intensive or compute-intensive based on the number of instructions and the number of memory accesses that can be obtained easily through performance counter or performance application programming interface (PAPI) [14].…”

Section: Discussionmentioning

confidence: 99%

Adaptive demand‐aware work‐stealing in multi‐programmed multi‐core architectures

Chen

Zheng

Guo

2015

Concurrency and Computation

View full text Add to dashboard Cite

SUMMARYModern multi-core computers often execute multiple programs concurrently, but traditional work-stealing schedulers assume that only a single program executes at a time. If multiple work-stealing programs run on a multi-core computer concurrently, the performance of these co-located programs is poor and unbalanced due to the interference among their workers. To relieve this problem, this paper proposes demand-aware work-stealing (DWS) for a single work-stealing program. If multiple programs that adopt DWS run simultaneously on a multi-core system, the computation resources are evenly allocated to the co-located programs at the beginning. At runtime, if a program does not need all its allocated cores, it donates its free cores. Otherwise, if a program needs more cores, it steals the free cores donated by its co-located programs. In addition, each program that adopts DWS dynamically puts its workers to sleep and wakes them up according to the number of its queued tasks. Experimental results show that DWS can improve the co-located programs' performance up to 42.6% and can also balance the performance of the co-located programs compared with traditional work-stealing.

show abstract

Section: Discussionmentioning

confidence: 99%

Adaptive demand‐aware work‐stealing in multi‐programmed multi‐core architectures

Chen

Zheng

Guo

2015

Concurrency and Computation

View full text Add to dashboard Cite

show abstract

“…PAPI (Browne et al, 2000;Johnson et al, 2012) is a well-known and widely applied programming interface for accessing the hardware counters within an application's source code. PAPI provides both a simple, high level interface for the acquisition of simple measurements and a fully programmable, low level interface to the underlying counter hardware.…”

Section: Methodsmentioning

confidence: 99%

A framework for comparative performance study on virtualised machines

Zhao

Tao

Fürlinger

2014

IJAHUC

View full text Add to dashboard Cite

The virtualisation technology has been widely used today in various research fields, including high performance computing, grid computing, cloud computing as well as server-client systems. The virtualisation introduces advantages such as on-demand customised resource provision, easy management and support for multiple Operating Systems, etc. However, this technology also results in a performance loss due to the additional virtualisation layer. To help application developers study the performance issues on virtualised multicore machines, we developed a tool framework with a set of performance tools that collect the runtime performance data and demonstrate the different runtime application behaviour on both the physical machine and the virtual machine within a single view. We appliedthetoolframeworktoanalyseanumberofbenchmarkapplicationsandconductedoptimisations based on the analysis discovery. The initial experimental results show a significant performance gain. She is currently a senior research associate at the Steinbuch Centre for Computing, Karlsruhe Institute of Technology. Her research work targets mainly on parallel and distributed computing, data-intensive computing and grid and cLOUD computing. She has published more than 100 articles in peer-reviewed international journals and leading conferences. She serves as co-chair or PC member in a set of international conferences and workshops.

show abstract

“…Lampe et al measured the impact of inaccurate timing in publicly available cloud computing services [8]. Johnson et al describe extensions to the Performance API (PAPI) project which provide more accurate timers and shared resource usage statistics in VMs [5]. However, these extensions only provide system-wide steal time information, and the paper emphasizes the need for per-process steal time accounting.…”

Section: Related Workmentioning

confidence: 98%

Sampling-based Steal Time Accounting under Hardware Virtualization

Höfer

Hörschläger

Mössenböck

2015

Proceedings of the 6th ACM/SPEC International Conference on Performance Engineering

View full text Add to dashboard Cite

Virtualization enables the efficient sharing of hardware resources among multiple virtual machines (VMs). Because the physical resources are limited, the scheduler must often suspend one VM to allow some other VM to run. The operating system in a VM is typically unaware of the suspension and accounts periods of suspension as CPU time to the executing application thread. This misrepresentation of resource usage makes it difficult to tell whether a performance problem is caused by an actual bottleneck in the application or by the virtualization infrastructure.We present a novel approach to compute to what degree the threads of an application in a virtual machine are affected by suspension. Our approach does not require any modifications to the operating system or to the virtualization software. It periodically samples the system-wide amount of "steal time" that is reported by the virtualization infrastructure, and divides it among the monitored threads according to their CPU usage. With a prototype implementation, we demonstrate that our approach accounts accurate amounts of steal time to application threads, that it can be used to compute the true resource usage of an application, and that it incurs only negligible performance overhead.

show abstract

PAPI-V: Performance Monitoring for Virtual Machines

Cited by 16 publications

References 12 publications

Adaptive demand‐aware work‐stealing in multi‐programmed multi‐core architectures

Adaptive demand‐aware work‐stealing in multi‐programmed multi‐core architectures

A framework for comparative performance study on virtualised machines

Sampling-based Steal Time Accounting under Hardware Virtualization

Contact Info

Product

Resources

About