• Journal of Internet Computing and Services
• /
• v.17 no.4
• /
• pp.1-9
• /
• 2016

Algorithms implementing distributed shared memory (DSM) were developed for ensuring consistency. The performance of DSM algorithms is dependent on system and usage parameters. However, ensuring these algorithms to tolerate faults is a problem that needs to be researched. In this study, we proposed fault-tolerant scheme for DSM system and analyzed reliability and fault-tolerant overhead. Using our analysis, we can choose a proper algorithm for DSM on error prone environment.

• The Transactions of the Korea Information Processing Society
• /
• v.3 no.6
• /
• pp.1433-1442
• /
• 1996

Data distribution of SPMD(Single Program Multiple Data) pattern is one of main features of HPF (High Performance Fortran). This paper describes design is sues for such data distribution and its efficient execution model on TICOM IV computer, named SPAX(Scalable Parallel Architecture computer based on X-bar network). SPAX has a hierarchical clustering structure that uses distributed shared memory(DSM). In such memory structure, it cannot make a full system utilization to apply unanimously either SMDD(shared Memory Data Distribution) or DMDD(Distributed Memory Data Distribution). Here we propose another data distribution model, called DSMDD(Distributed Shared Memory Data Distribution), a data distribution model based on hierarchical masters-slaves scheme. In this model, a remote master and slaves are designated in each node, shared address scheme is used within a node and message passing scheme between nodes. In our simulation, assuming a node size in which system performance degradation is minimized,DSMDD is more effective than SMDD and DMDD. Especially,the larger number of logical processors and the less data dependency between distributed data,the better performace is obtained.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.04a
• /
• pp.116-118
• /
• 2000

분산 공유 메모리(Distributed Shared Memory)시스템은 사용자에게 간단한 공유메모리 개념을 제공하기 때문에 사용자는 노드 사이의 데이터 이동에 관여할 필요가 없다. DSM에서 일치서 유지를 위한 프로토콜을 선택하는 것은 통신부하를 줄이는데 중요한 역할을 한다. 본 논문은 DSM 프로토콜을 효과적으로 선택하기 위한 새로운 성능평가 모델을 제시한다. 본 연구에서 제안하는 성능평가 모델을 사용함으로써 무효화방식(invalidate protocol), 갱신 방식(update protocol) 그리고 이주방식(migratory protocol)의 성능예측이 가능하다. 본 성능평가모델은 노드들 사이의 데이터 일치성(consistency)유지를 위한 부담을 최소화하는 최적의 DSM 프로토콜을 결정하는데 사용된다.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.2
• /
• pp.476-484
• /
• 1999

Many message logging and checkpointing schemes are proposed for fault tolerance in distributed systems in which nodes communicate by message passing. Most researches for recoverable distributed shared memory (DSM) also adopt similar schemes used in message passing systems. However, schemes used in message passing systems are not always appropriate to be directly used in DSM systems because the two systems, message passing systems and DSM systems, have different natures (function shipping and data shipping). Many modified schemes have been proposed for DSM systems to resolve these differences. In this paper, an object oriented approach is proposed for recoverable DSM. We present a new dependency tracking scheme between pages instead of processes. Based on this scheme, we propose new checkpointing and recovery schemes that can reduce overhead to make DSM recoverable.

• Journal of KIISE:Computing Practices and Letters
• /
• v.5 no.6
• /
• pp.700-707
• /
• 1999

본 논문에서는 GIS 관련 연산을 실시간에 효율적으로 처리하기 위한 분산공유메모리 기반 병렬처리 시스템을 제안한다. 본 논문의 분산공유메모리 시스템은 메시지전달 방식의 분산메모리 MIMD 컴퓨터 상에 소프트웨어 기반 분산공유메모리 모듈을 탑재함으로써 구현되었다. 또한 GIS 연산의 기본이 되는 공간 객체를 공유의 기본 단위로 설정하고, GIS 데이타의 특성을 반영하여 읽기전용 공유데이타 타입을 추가하였으며, 네트워크 오버헤드를 줄이기 위하여 복수의 객체를 한번에 읽어오는 bulk access가 가능하도록 하였다. 본 시스템에서는 GIS 데이타의 효율적인 분배를 위하여 부하균등화 기법으로 guided self scheduling을 사용하였다. 실험결과 본 시스템은 네트워크 캐쉬의 효율적인 활용을 통하여 소프트웨어 기반 분산메모리 시스템의 오버헤드에도 불구하고 MPI 기반 메시지전달 방식에 비하여 향상된 성능을 얻을 수 있었다.Abstract In this paper, we propose a distributed shared memory(DSM) based parallel processing system to process GIS related computations efficiently in real time. The system is based on a software DSM module implemented on top of a distributed MIMD computer. In the DSM system, spatial object, which is a fundamental structure to represent GIS data, is used as a basic unit for sharing, and a read-only shared data type is added to reflect the characteristics of GIS data. In addition, a bulk access to multiple shared data is made possible to reduce the network overhead. A guided self scheduling method is devised for efficient load balancing in distributing GIS data to parallel processors. The experimental results show that the DSM system performs better than an MPI based message-passing system through the efficient utilization of network cache in spite of the system's software overhead.

• Journal of Digital Contents Society
• /
• v.9 no.1
• /
• pp.77-86
• /
• 2008

False sharing is a result of co-location of unrelated data in the same unit of memory coherency, and is one source of unnecessary overhead being of no help to keep the memory coherency in multiprocessor systems. Moreover, the damage caused by false sharing becomes large in proportion to the granularity of memory coherency. To reduce false sharing in page-based DSM systems, it is necessary to allocate unrelated data objects that have different access patterns into the separate shared pages. In this paper we propose sized and call-site tracing-based shared memory allocator, shortly SCSTallocator. SCSTallocator places each data object requested from the different call-sites into the separate shared pages, and at the same time places each data object that has different size into different shared pages. Consequently data objects that have the different call-site and different object size prohibited from being allocated to the same shared page. Our observations show that our SCSTallocator outperforms the existing dynamic shared memory allocators. By combining the two existing allocation technique, we can reduce a considerable amount of false sharing misses.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2005.05a
• /
• pp.933-936
• /
• 2005

네트워크 기술의 발전과 멀티미디어 데이터의 전송을 가능하게 되어 VOD 서버등의 연구가 활발하게 이루어지고 있다. 본 논문에서는 리눅스 시스템에 다윈 스트리밍 서버와 DSM(Distributed Shared Memory)을 구축하고, 구축된 시스템을 이용하여 DSM 기반에서 다윈 스트리밍 서버가 실행이 될 수 있도록 설계한다.

• 제어로봇시스템학회:학술대회논문집
• /
• 2002.10a
• /
• pp.94.1-94
• /
• 2002

This paper is concerned with parallel FFT and Quick-Merge Sort. They are implemented on computers interconnected by VMIC 5579 reflective memory and a cluster of workstations (PCs) interconnected via Fast Ethernet. Message passing interface (MPI) parallel library is used for communication in a cluster of workstations. An improved parallel FFT is also presented to decrease an execution time in the case of a small number of hosts. Distributed shared memory (DSM), VMIC 5579 reflective memory (RM), a cluster of workstations (COW) and message passing interface (MPI) parallel library are described.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.4
• /
• pp.1098-1105
• /
• 1999

The caches in distributed shared memory systems enhance the performance by reducing memory access latency and communication overhead, but they must solve the cache coherence problem. This paper proposes a new directory protocol to solve the cache coherence problem and to improve the system performance in distributed shared memory systems. To maintain the cache coherence of shared data, processors within a limited distance reduce the communication overhead by using a bit-vector like the full directory scheme. Processors over a limited distance store pointers in a directory pool. Since the bit-vector and the directory pool remove the unnecessary cache invalidations, the proposed scheme reduces the communication traffic and improves the system performance. The dynamic limited directory scheme reduces the communication traffic up to 66 percents compared with the limited directory scheme and the number of directory access up to 27 percents compared with the dynamic pointer allocation scheme.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.8
• /
• pp.2245-2252
• /
• 1999

Since DSM provides a user a simple shared memory abstraction, the user does not have to be concerned with data movement between hosts. Each node in DSM systems has processor, memory, and connection to a network. Memory is divided into pages, and a page can have multiple copies in different nodes. To maintain data consistency between nodes, two conventional protocols are used : write-update protocol and invalidate protocol. The performance of these protocols depends on the system parameters and the memory access patterns. for adapting to memory access patterns, competitive update protocol updates those copies of a page that are expected to be used in the near future, while selectively invalidating other copies. We present weighted competitive update protocols that consider different communication bandwidth for each connection a of two nodes. Test result by simulation show that the weighted competitive update protocol improves performance.