No abstract

No abstract

Abstract. Adaptation of legacy software to grid services environment is gradually gaining in significance both in academic and commercial settings but presently no comprehensive framework addressing this area is available and the scope of research work covering this field is still unsatisfactory. The main contribution of this paper is the proposal of a versatile architecture designed to facilitate the process of transition to grid services platform. We provide thorough analysis of the presented solution and confront it with fundamental grid requirements. In addition, the results of performance evaluation of a prototype implementation are demonstrated.

As software becomes increasingly complex and difficult to analyze, it is more and more common for developers to use high-level, type-safe, object-oriented (OO) programming languages and to architect systems that comprise multiple components. Different components are often implemented in different programming languages. In state-of-the-art multicomponent, multi-language systems, cross-component communication relies on remote procedure calls (RPC) and message passing. As components are increasingly co-located on the same physical machine to ensure high utilization of multi-core systems, there is a growing potential for using shared memory for cross-language cross-runtime communication.We present the design and implementation of Co-Located Runtime Sharing (CoLoRS), a system that enables crosslanguage, cross-runtime type-safe, transparent shared memory. CoLoRS provides object sharing for co-located OO runtimes for both static and dynamic languages. CoLoRS defines a language-neutral object/class model, which is a staticdynamic hybrid and enables class evolution while maintaining the space/time efficiency of a static model. CoLoRS uses type mapping and class versioning to transparently map shared types to private types. CoLoRS also contributes a synchronization mechanism and a parallel, concurrent, on-thefly GC algorithm, both designed to facilitate cross-language cross-runtime object sharing.We implement CoLoRS in open-source, productionquality runtimes for Python and Java. Our empirical evaluation shows that CoLoRS extensions impose low overhead. We also investigate RPC over CoLoRS and find that using shared memory to implement co-located RPC significantly Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. OOPSLA/SPLASH'10, October 17-21, 2010, Reno/Tahoe, Nevada, USA. Copyright c 2010 ACM 978-1-4503-0203-6/10/10. . . $10.00 improves both communication throughput and latency by avoiding data structure serialization.

The growth in complexity of modern systems makes it increasingly difficult to extract high-performance. The software stacks for such systems typically consist of multiple layers and include managed runtime environments (MREs). In this paper, we investigate techniques to improve cooperation between these layers and the hardware to increase the efficacy of automatic memory management in MREs.General-purpose MREs commonly implement parallel and/or concurrent garbage collection and employ compaction to eliminate heap fragmentation. Moreover, most systems trigger collection based on the amount of heap a program uses. Our analysis shows that in many cases this strategy leads to ineffective collections that are unable to reclaim sufficient space to justify the incurred cost. To avoid such collections, we exploit the observation that dead objects tend to cluster together and form large, never-referenced, regions in the address space that correlate well with virtual pages that have not recently been referenced by the application. We leverage this correlation to design a new, simple and light-weight, yield predictor that estimates the amount of reclaimable space in the heap using hardware page reference bits. Our predictor allows MREs to avoid low-yield collections and thereby improve resource management.We integrate this predictor into three state-of-the-art parallel compactors, implemented in the HotSpot JVM, that represent distinct canonical heap layouts. Our empirical evaluation, based on standard Java benchmarks and opensource applications, indicates that inexpensive and accurate yield prediction can improve performance significantly.