Safe and Efficient Cluster Communication in Java using Explicit MemoryManagement

Other Titles


This thesis presents a framework for using explicit memory management to improve the communication performance of JavaTM cluster applications. The framework allows programmers to explicitly manage Java communication buffers, called jbufs, which are directly accessed by the DMA engines of high-performance network interfaces and by Java programs as primitive-typed ar-rays. The central idea is to remove the hard separation between Java's gar-bage-collected heap and the non-collected memory region in which DMA buffers must normally be allocated. The programmer controls when a jbuf is part of the garbage-collected heap so that the garbage collector can ensure it is safely re-used or de-allocated, and when it is not so it can be used for DMA transfers. Unlike other techniques, jbufs preserve Java's storage- and type-safety and do not depend on a particular garbage collection scheme. The safety, efficiency, and programmability of jbufs are demonstrated throughout this thesis with implementations of an interface to the Virtual In-terface Architecture, of an Active Messages communication layer, and of Java Remote Method Invocation (RMI). The impact on applications is also evalu-ated using an implementation of cluster matrix multiplication as well as a publicly available RMI benchmark suite. The thesis proposes in-place object de-serialization (de-serialization with-out allocation and copying of objects) to further enhance the performance of RMI on homogeneous clusters. This optimization takes advantage of the zero-copy capabilities of network devices to reduce the per-object de-serialization costs to a constant irrespective of object size, which is particularly beneficial for large objects such as arrays. In-place de-serialization is realized using jstreams, an extension of jbufs with object I/O streams. Jstreams use the ex-plicit memory management offered by jbufs to incorporate de-serialized ob-jects into the receiving Java virtual machine without compromising its integrity, without restricting the usage of those objects, and without making assumptions about the underlying garbage collection scheme. The perform-ance impact of jstreams on Java RMI and the benchmark suite is evaluated.

Journal / Series

Volume & Issue



Date Issued



Cornell University


computer science; technical report


Effective Date

Expiration Date




Union Local


Number of Workers

Committee Chair

Committee Co-Chair

Committee Member

Degree Discipline

Degree Name

Degree Level

Related Version

Related DOI

Related To

Related Part

Based on Related Item

Has Other Format(s)

Part of Related Item

Related To

Related Publication(s)

Link(s) to Related Publication(s)


Link(s) to Reference(s)

Previously Published As

Government Document




Other Identifiers


Rights URI


technical report

Accessibility Feature

Accessibility Hazard

Accessibility Summary

Link(s) to Catalog Record