PERFORMANCE EVALUATION OF THE ORCA SHARED-OBJECT SYSTEM

Orca is a portable, object-based distributed shared memory (DSM) syste m. This article studies and evaluates the design choices made in the O rca system and compares Orca with other DSMs. The article gives a quan titative analysis of Orca's coherence protocol (based on write-updates with function shipping), the totally ordered group communication prot ocol, the strategy for object placement, and the all-software, user-sp ace architecture. Performance measurements for 10 parallel application s illustrate the trade-offs made in the design of Orca and show that e ssentially the right design decisions have been made. A write-update p rotocol with function shipping is effective for Orca, especially since it is used in combination with techniques that avoid replicating obje cts that have a low read/write ratio. The overhead of totally ordered group communication on application performance is low. The Orca system is able to make near-optimal decisions for object placement and repli cation. In addition, the article compares the performance of Orca with that of a page-based DSM (TreadMarks) and another object-based DSM (C RL). It also analyzes the communication overhead of the DSMs for sever al applications. All performance measurements are done on a 32-node Pe ntium Pro cluster with Myrinet and Fast Ethernet networks. The results show that the Orca programs send fewer messages and less data than th e TreadMarks and CRL programs and obtain better speedups.