Thread migration and communication minimization in DSM systems

Thitikamol, K.; Keleher, Peter J.

doi:10.1109/5.747869

Cited by 38 publications

(2 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Barnes-Hut and Water-Spatial achieved less perfect results but all classes are consistently over 92% accurate. V. RELATED WORK Active correlation tracking is first proposed in D-CVM [12,25] and extended in later work [26]. The system deliberately disables preemptive thread scheduling and sets each page to be invalid for invoking access log.…”

Section: ) Accuracymentioning

confidence: 99%

Adaptive sampling-based profiling techniques for optimizing the distributed JVM runtime

Lam

Luo

Wang

2010

2010 IEEE International Symposium on Parallel &Amp; Distributed Processing (IPDPS)

View full text Add to dashboard Cite

Abstract-Extending the standard Java virtual machine (JVM) for cluster-awareness is a transparent approach to scaling out multithreaded Java applications. While this clustering solution is gaining momentum in recent years, efficient runtime support for fine-grained object sharing over the distributed JVM remains a challenge. The system efficiency is strongly connected to the global object sharing profile that determines the overall communication cost. Once the sharing or correlation between threads is known, access locality can be optimized by collocating highly correlated threads via dynamic thread migrations. Although correlation tracking techniques have been studied in some page-based software DSM systems, they would entail prohibitively high overheads and low accuracy when ported to fine-grained object-based sharing systems. In this paper, we propose a lightweight sampling-based profiling technique for tracking inter-thread sharing. To preserve locality across migrations, we also propose a thread stack sampling mechanism for profiling the set of objects which are tightly coupled with a migrant thread. Sampling rates in both techniques can vary adaptively to strike a balance between preciseness and overhead. Such adaptive techniques are particularly useful for applications whose sharing patterns could change dynamically. The profiling results can be exploited for effective thread-tocore placement and dynamic load balancing in systems ranging from a multicore machine up to a cloud computing environment. We present the design and preliminary performance result of the JESSICA2 distributed JVM with the profiling implemented. Experimental results show that the profiling is able to obtain over 95% accurate global sharing profiles at a cost of only a few percents of execution time increase for fineto medium-grained applications.

show abstract

Section: ) Accuracymentioning

confidence: 99%

Adaptive sampling-based profiling techniques for optimizing the distributed JVM runtime

Lam

Luo

Wang

2010

2010 IEEE International Symposium on Parallel &Amp; Distributed Processing (IPDPS)

View full text Add to dashboard Cite

show abstract

“…As a matter of fact, many automatic adaptive schemes [5,6,7,8,9,10,11] have been proposed. These include adaptation between the single-writer and multiple-writer protocols, adaptation between the write-invalidate and writeupdate protocols, and even process/thread migration.…”

Section: One Of the Important Variant Of Rc Is The Home-based Releasementioning

confidence: 99%

ORION: an adaptive home-based software distributed shared memory system

Wong

Proceedings Seventh International Conference on Parallel and Distributed Systems (Cat. No.PR00568)

View full text Add to dashboard Cite

In this paper, we describe a multithreaded software distributed shared memory (DSM) system named Orion. It has been developed to provide POSIX-thread (pthread) The two fully automatic schemes aim to involve minimal user intervention and yet deliver good performances with some speedups ranging from 2% to 79% observed in some 8 benchmarks tested. KeywordsDistributed shared memory, virtual shared memory, release consistency, home-based release consistency and multithreading.

show abstract