Pub Date : 2000-11-28DOI: 10.1109/CLUSTER.2000.10013
Chei-Yol Kim, J. Cho, Dae-Wha Seo
An adaptive Dual-Cache Scheme(DCS) and dynamic prefetching scheme were designed to improve the performance of the Parallel File System for Linux(PFSL). PFSL is a parallel file system for a clustering environment and is implemented using a multi-threaded programming technique with POSIX thread libraries supported by Linux. The proposed adaptive DCS and dynamic prefetching scheme both reflect the current file request. The former reduces the first file request response time, whereas the latter improves the file system performance by adjusting the amount of data for prefetching.
{"title":"Adaptive Dual-Cache Scheme with Dynamic Prefetching Scheme in Parallel File System","authors":"Chei-Yol Kim, J. Cho, Dae-Wha Seo","doi":"10.1109/CLUSTER.2000.10013","DOIUrl":"https://doi.org/10.1109/CLUSTER.2000.10013","url":null,"abstract":"An adaptive Dual-Cache Scheme(DCS) and dynamic prefetching scheme were designed to improve the performance of the Parallel File System for Linux(PFSL). PFSL is a parallel file system for a clustering environment and is implemented using a multi-threaded programming technique with POSIX thread libraries supported by Linux. The proposed adaptive DCS and dynamic prefetching scheme both reflect the current file request. The former reduces the first file request response time, whereas the latter improves the file system performance by adjusting the amount of data for prefetching.","PeriodicalId":92128,"journal":{"name":"Proceedings. IEEE International Conference on Cluster Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2000-11-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77210892","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2000-01-01DOI: 10.1109/CLUSTER.2000.10000
G. Ciaccio, G. Chiola
In this paper we report about the recently completed porting of the Genoa Active Message MAchine (GAMMA) to the Netgear GA620 Gigabit Ethernet adapter. Such device is a low-cost (less than 300 US dollars at the time of writing) Gigabit Ethernet adapter with a state-of-art internal architecture based on two programmable on-board processors and substantial amount of on-board RAM, which makes this product a very appealing, cheap alternative to Myrinet. A combination of low end-to-end latency (32 s) and high transmission throughput (103 MByte/s end-to-end) demonstrates the potential for Gigabit Ethernet lightweight protocols to yield messaging performance comparable to the best Myrinet protocols. This result is of interest, given the envisaged drop in cost of Gigabit Ethernet due to the forthcoming transition from fiber optic to UTP cabling and ever increasing mass market production of such standard interconnect.
{"title":"Low-cost Gigabit Ethernet at Work","authors":"G. Ciaccio, G. Chiola","doi":"10.1109/CLUSTER.2000.10000","DOIUrl":"https://doi.org/10.1109/CLUSTER.2000.10000","url":null,"abstract":"In this paper we report about the recently completed porting of the Genoa Active Message MAchine (GAMMA) to the Netgear GA620 Gigabit Ethernet adapter. Such device is a low-cost (less than 300 US dollars at the time of writing) Gigabit Ethernet adapter with a state-of-art internal architecture based on two programmable on-board processors and substantial amount of on-board RAM, which makes this product a very appealing, cheap alternative to Myrinet. A combination of low end-to-end latency (32 s) and high transmission throughput (103 MByte/s end-to-end) demonstrates the potential for Gigabit Ethernet lightweight protocols to yield messaging performance comparable to the best Myrinet protocols. This result is of interest, given the envisaged drop in cost of Gigabit Ethernet due to the forthcoming transition from fiber optic to UTP cabling and ever increasing mass market production of such standard interconnect.","PeriodicalId":92128,"journal":{"name":"Proceedings. IEEE International Conference on Cluster Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2000-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73795375","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2000-01-01DOI: 10.1109/CLUSTER.2000.10014
G. Pfister
{"title":"The Future of Cluster Interconnects: InfiniBand or Ethernet?","authors":"G. Pfister","doi":"10.1109/CLUSTER.2000.10014","DOIUrl":"https://doi.org/10.1109/CLUSTER.2000.10014","url":null,"abstract":"","PeriodicalId":92128,"journal":{"name":"Proceedings. IEEE International Conference on Cluster Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2000-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73348997","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2000-01-01DOI: 10.1109/CLUSTER.2000.10012
Joachim Worringen
An ear cushioning device is disclosed for use with plug type earphones mounted on a springable carrier, as are commonly dispensed on commercial airplanes. The cushioning device is of a soft, resilient foam material which provides comfort through a spreading of the pressure from the earphones over the entire ear of the user. A preferred embodiment of the device is formed integrally from a single foam pad, with the ear cushions at the ends and a stretchable band between the cushions, for extension over the top of the user's head. The stretchable band is preferably a convoluted formation which acts as a light spring, tending to pull the ear cushions toward one another. Openings through the ear cushions receive the plug type earphones, so that the latter do not extend into the user's ears. In another embodiment, a semi-rigid head band fits over the head with a pair of soft foam ear cushions slidably mounted on the ends of the head band. An elongated opening at each end of the head band receives an earphone plug for transmission of sound to the ear, and the head band and cushions accept and spread the pressure of the earphones. In either embodiment, soft pliable inserts may be positioned in the ear cushion openings to receive the earphone plugs and carry the sound into the ears.
{"title":"Efficient Parallel I/O on SCI Connected Clusters","authors":"Joachim Worringen","doi":"10.1109/CLUSTER.2000.10012","DOIUrl":"https://doi.org/10.1109/CLUSTER.2000.10012","url":null,"abstract":"An ear cushioning device is disclosed for use with plug type earphones mounted on a springable carrier, as are commonly dispensed on commercial airplanes. The cushioning device is of a soft, resilient foam material which provides comfort through a spreading of the pressure from the earphones over the entire ear of the user. A preferred embodiment of the device is formed integrally from a single foam pad, with the ear cushions at the ends and a stretchable band between the cushions, for extension over the top of the user's head. The stretchable band is preferably a convoluted formation which acts as a light spring, tending to pull the ear cushions toward one another. Openings through the ear cushions receive the plug type earphones, so that the latter do not extend into the user's ears. In another embodiment, a semi-rigid head band fits over the head with a pair of soft foam ear cushions slidably mounted on the ends of the head band. An elongated opening at each end of the head band receives an earphone plug for transmission of sound to the ear, and the head band and cushions accept and spread the pressure of the earphones. In either embodiment, soft pliable inserts may be positioned in the ear cushion openings to receive the earphone plugs and carry the sound into the ears.","PeriodicalId":92128,"journal":{"name":"Proceedings. IEEE International Conference on Cluster Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2000-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78391932","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2000-01-01DOI: 10.1109/CLUSTER.2000.10033
V. Tran, L. Hluchý, Giang T. Nguyen
In this paper, we present a new model for parallel program development for cluster computing called Data Driven Graph (DDG). DDG automatically analyzes data dependence among tasks, synchronizes data, generates task graphs and schedules. Programming in DDG is easy and reliable; most of work is done automatically by DDG, what does not only minimize the amount of work done by programmers but also removes most frequent errors likes race conditions and deadlocks. The integrated scheduler makes parallel programs in DDG run efficiently. Our experiments demonstrate the simplicity and efficiency of programs written in DDG.
{"title":"Parallel Programming Environment for Cluster Computing","authors":"V. Tran, L. Hluchý, Giang T. Nguyen","doi":"10.1109/CLUSTER.2000.10033","DOIUrl":"https://doi.org/10.1109/CLUSTER.2000.10033","url":null,"abstract":"In this paper, we present a new model for parallel program development for cluster computing called Data Driven Graph (DDG). DDG automatically analyzes data dependence among tasks, synchronizes data, generates task graphs and schedules. Programming in DDG is easy and reliable; most of work is done automatically by DDG, what does not only minimize the amount of work done by programmers but also removes most frequent errors likes race conditions and deadlocks. The integrated scheduler makes parallel programs in DDG run efficiently. Our experiments demonstrate the simplicity and efficiency of programs written in DDG.","PeriodicalId":92128,"journal":{"name":"Proceedings. IEEE International Conference on Cluster Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2000-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77032047","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2000-01-01DOI: 10.1109/CLUSTER.2000.10024
E. Cecchet, Cyril Perrin
The SIRAC laboratory has developed SciFS, a Distrib- uted Shared Memory (DSM) tightly integrated with the Linux operating system kernel, that exploits the high performance and remote addressing capabilities of SCI (Scalable Coherent Interface) networks. We experiment with using SciFS to build large scale clustered web caches. An efficient DSM does not only provide a fast and large data storage space, but it also allows the design of new algorithms to handle user requests. * We propose a clustered web cache prototype with a new request distribution algorithm, called PPBL (Parallel Pull-Based LRU), especially designed for use with a DSM system. Unlike other distribution algorithms the decision is distributed over all nodes thus providing better scal- ability. In this short paper, we only present PPBL princi- ples but additional information can be found on our web site at http://sci-serv.inrialpes.fr.
{"title":"Parallel Pull-Based LRU : a New Request Distribution Algorithm for Large Scale SCI Clustered Web Caches","authors":"E. Cecchet, Cyril Perrin","doi":"10.1109/CLUSTER.2000.10024","DOIUrl":"https://doi.org/10.1109/CLUSTER.2000.10024","url":null,"abstract":"The SIRAC laboratory has developed SciFS, a Distrib- uted Shared Memory (DSM) tightly integrated with the Linux operating system kernel, that exploits the high performance and remote addressing capabilities of SCI (Scalable Coherent Interface) networks. We experiment with using SciFS to build large scale clustered web caches. An efficient DSM does not only provide a fast and large data storage space, but it also allows the design of new algorithms to handle user requests. * We propose a clustered web cache prototype with a new request distribution algorithm, called PPBL (Parallel Pull-Based LRU), especially designed for use with a DSM system. Unlike other distribution algorithms the decision is distributed over all nodes thus providing better scal- ability. In this short paper, we only present PPBL princi- ples but additional information can be found on our web site at http://sci-serv.inrialpes.fr.","PeriodicalId":92128,"journal":{"name":"Proceedings. IEEE International Conference on Cluster Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2000-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82083329","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2000-01-01DOI: 10.1109/CLUSTER.2000.10005
Kazuki Hyoudou, Yasuichi Nakayama
An optical alignment system for optically determining the translational and rotational orientation of portions of an object with respect to a system reference utilizing a reflective target (8) positioned on the object portion and adapted to receive an optical reference beam (6) transmitted to the target (8) from the reference position. The reflective target (8) includes a photoelectric sensor (22) for producing an electrical signal indicative of the lateral displacement of the reflective target (8) from the optical reference beam (6) and a reflective surface (mirror 24) for returning a portion (return beam 26) of the optical reference beam (6) to a photosensor (28) positioned adjacent the reference beam source (12) to provide information relative to the angular position of the reflective target (8). A method for operating the system is also disclosed.
{"title":"FireCluster: PC Cluster System Employing IEEE 1394","authors":"Kazuki Hyoudou, Yasuichi Nakayama","doi":"10.1109/CLUSTER.2000.10005","DOIUrl":"https://doi.org/10.1109/CLUSTER.2000.10005","url":null,"abstract":"An optical alignment system for optically determining the translational and rotational orientation of portions of an object with respect to a system reference utilizing a reflective target (8) positioned on the object portion and adapted to receive an optical reference beam (6) transmitted to the target (8) from the reference position. The reflective target (8) includes a photoelectric sensor (22) for producing an electrical signal indicative of the lateral displacement of the reflective target (8) from the optical reference beam (6) and a reflective surface (mirror 24) for returning a portion (return beam 26) of the optical reference beam (6) to a photosensor (28) positioned adjacent the reference beam source (12) to provide information relative to the angular position of the reflective target (8). A method for operating the system is also disclosed.","PeriodicalId":92128,"journal":{"name":"Proceedings. IEEE International Conference on Cluster Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2000-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83413455","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2000-01-01DOI: 10.1109/CLUSTER.2000.10017
G. Goldszmidt
The Océano project is developing a prototype of a scaleable infrastructure that enables multienterprise hosting on a virtualized collection of hardware resources. Hosted customers increasingly require support for peak loads that are orders of magnitude larger than what they experience in their normal steady state, particularly for commercial Web workloads. Océano provides a hosting environment with fast response time in adjusting the resources (bandwidth, servers, and storage), assigned to each hosted customer, to the dynamically fluctuating workload. It creates dynamically carved domain clusters for each hosted enterprise, and introduces high levels of automation to control resources. Our objective is to research hosting architectures that enable high availability and dynamic scalability, for delivering networked applications and services, and are capable of handling large surges in traffic. We have implemented a pilot over a set of 50 Linux and 6 AIX servers.
{"title":"The Océano Project - A Multi-Domain Cluster for a Computing Utility","authors":"G. Goldszmidt","doi":"10.1109/CLUSTER.2000.10017","DOIUrl":"https://doi.org/10.1109/CLUSTER.2000.10017","url":null,"abstract":"The Océano project is developing a prototype of a scaleable infrastructure that enables multienterprise hosting on a virtualized collection of hardware resources. Hosted customers increasingly require support for peak loads that are orders of magnitude larger than what they experience in their normal steady state, particularly for commercial Web workloads. Océano provides a hosting environment with fast response time in adjusting the resources (bandwidth, servers, and storage), assigned to each hosted customer, to the dynamically fluctuating workload. It creates dynamically carved domain clusters for each hosted enterprise, and introduces high levels of automation to control resources. Our objective is to research hosting architectures that enable high availability and dynamic scalability, for delivering networked applications and services, and are capable of handling large surges in traffic. We have implemented a pilot over a set of 50 Linux and 6 AIX servers.","PeriodicalId":92128,"journal":{"name":"Proceedings. IEEE International Conference on Cluster Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2000-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75120047","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2000-01-01DOI: 10.1109/CLUSTER.2000.10020
M. Herz
{"title":"Remote System Management Principles","authors":"M. Herz","doi":"10.1109/CLUSTER.2000.10020","DOIUrl":"https://doi.org/10.1109/CLUSTER.2000.10020","url":null,"abstract":"","PeriodicalId":92128,"journal":{"name":"Proceedings. IEEE International Conference on Cluster Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2000-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81176934","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2000-01-01DOI: 10.1109/CLUSTER.2000.10011
A. F. Díaz, Jesus Ferreira, J. Ortega, A. Cañas, A. Prieto
A layer for efficient communication on clusters of Personal Computers (PCs) using the Linux operating system is proposed. Besides optimizing communication performance, by reducing the latencies and increasing the bandwidth figures even for short messages, the proposed communication system also meets other requirements such as multiprogramming, portability, protection against corrupted programs, reliable message delivery, and direct access to the network for all applications.
{"title":"CLIC: Fast Communication on Linux Clusters","authors":"A. F. Díaz, Jesus Ferreira, J. Ortega, A. Cañas, A. Prieto","doi":"10.1109/CLUSTER.2000.10011","DOIUrl":"https://doi.org/10.1109/CLUSTER.2000.10011","url":null,"abstract":"A layer for efficient communication on clusters of Personal Computers (PCs) using the Linux operating system is proposed. Besides optimizing communication performance, by reducing the latencies and increasing the bandwidth figures even for short messages, the proposed communication system also meets other requirements such as multiprogramming, portability, protection against corrupted programs, reliable message delivery, and direct access to the network for all applications.","PeriodicalId":92128,"journal":{"name":"Proceedings. IEEE International Conference on Cluster Computing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2000-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76174599","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}