{"title":"Network file for ISDN","authors":"R. Arai, K. Itao","doi":"10.1109/MASS.1990.113586","DOIUrl":"https://doi.org/10.1109/MASS.1990.113586","url":null,"abstract":"The necessity of file-storage systems for network systems such as ISDN (integrated services digital network) and B-ISDN (broadband ISDN) is discussed. The stored information has a wide range of characteristics depending on the storage medium and it is accessed and transmitted through heterogeneous networks. Since visual services will increase with the progress of network systems, the file-storage system must provide large capacity, low storage cost, and high transfer rate. In ISDN and B-ISDN, these requirements will be met by optical-disk storage systems. In the future, the file-storage system will need to provide a huge storage capacity and an ultra-high data transfer rate sufficient for high-quality visual services at a low cost. This will require innovative technologies in memory devices and materials to surpass the real recording-density limits of present technologies. >","PeriodicalId":282025,"journal":{"name":"[1990] Digest of papers. Tenth IEEE Symposium on Mass Storage Systems@m_Crisis in Mass Storage","volume":"70 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116010770","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}
{"title":"Impact of HPPI and fibre channel standards on data delivery","authors":"P. Savage","doi":"10.1109/MASS.1990.113593","DOIUrl":"https://doi.org/10.1109/MASS.1990.113593","url":null,"abstract":"Two ANSI X3T9.3 standards activities are discussed. HPPI (high performance parallel interface) is a simple electrical point-to-point device that is implementable at either 100 or 200 MB. A HPPI channel is unidirectional; only a few return wires are used for flow control by the destination. There are no means for retransmitting after error, although the data is protected by adequate parity and longitudinal redundancy bits. Error bursts are delivered with status=error, or optionally dropped in the bit bucket, and any corrective action must be initiated by an upper-layer protocol. Thus, HPPI provides a datagram service. The fiber channel standardization project is a direct response to the need to extend the range of SCSI, IPI, and HPPI, all of which are electrical systems. Accordingly, the fiber channel standards working group established a goal of defining implementations that will accommodate SCSI, IPI, and HPPI within a common compatible lower-level set of protocols. It appears that other important transmission protocols can be modulated/carried on the fiber channel; examples are BlockMux channel and FDDI.<<ETX>>","PeriodicalId":282025,"journal":{"name":"[1990] Digest of papers. Tenth IEEE Symposium on Mass Storage Systems@m_Crisis in Mass Storage","volume":"93 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123819512","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}
{"title":"A 1990s solution to the crisis in mass storage","authors":"Ernie Ohrenstein, C. Masters","doi":"10.1109/MASS.1990.113592","DOIUrl":"https://doi.org/10.1109/MASS.1990.113592","url":null,"abstract":"A cost-effective implementation of disk-array technology for a scalable data-storage system which satisfies HPPI (high performance parallel interface) performance at the host interface level and also meets the objectives of large storage capacities, high sustained throughput, reliability, and efficient software interfacing is described. The design parameters for a data-storage system to satisfy a broad range of applications within the supercomputing environment are presented. It is shown that high-performance storage systems based on the HPPI industry standard will break the I/O (input/output) bottleneck and make it possible for the supercomputer users of the 1990s to efficiently support the throughput and capacity requirements of emerging applications which, until now, would have been too I/O-intensive to undertake. >","PeriodicalId":282025,"journal":{"name":"[1990] Digest of papers. Tenth IEEE Symposium on Mass Storage Systems@m_Crisis in Mass Storage","volume":"85 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131055360","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}
{"title":"Hiding mass storage under Unix: NASA's MSS-II architecture","authors":"D. Tweten","doi":"10.1109/MASS.1990.113585","DOIUrl":"https://doi.org/10.1109/MASS.1990.113585","url":null,"abstract":"A second-generation mass storage system, MSS-II, is described. It achieves a 10-MB transfer rate into and out of the central-network mass-storage node, while preserving the illusion to users that it is simply an ordinary Unix system with an infinite supply of very fast disk space. It does so by implementing a Berkeley RAID-style file system in kernel software, by instituting striping network access, by implementing a transparent hierarchical storage manager, and by extending the concepts of file ownership and permission to removable volumes. The major software components (i.e. the high-performance file system, rapid-access storage hierarchy, volume manager, and striping network access) and file integrity are discussed. >","PeriodicalId":282025,"journal":{"name":"[1990] Digest of papers. Tenth IEEE Symposium on Mass Storage Systems@m_Crisis in Mass Storage","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127494531","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}
{"title":"Magnetic tape technology in the 1990s","authors":"Harry C. Hinz","doi":"10.1109/MASS.1990.113566","DOIUrl":"https://doi.org/10.1109/MASS.1990.113566","url":null,"abstract":"The extensibility of the 8-mm recording technology as it concerns generation-to-generations compatibility, system transfer rates, physical size, and volumetric capacity is explored. A practical progression of tape generations from early inception to the 6-MB/s transfer rate of the future and an accompanying capacity increase to upwards of 67 is described. Possible technology paths toward greater volumetric efficiency in both media- and recording-device physical sizes are discussed, moving the latter from a 5.25-in full height to a 3.5-in form factor. >","PeriodicalId":282025,"journal":{"name":"[1990] Digest of papers. Tenth IEEE Symposium on Mass Storage Systems@m_Crisis in Mass Storage","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114496756","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}
{"title":"Automated cartridge system library server","authors":"Dennis Reed, G. Mueller","doi":"10.1109/MASS.1990.113578","DOIUrl":"https://doi.org/10.1109/MASS.1990.113578","url":null,"abstract":"A history and description of the automated cartridge system library server (ACSLS) are given. The ACSLS is a software product that introduces the StorageTek 4400 automated cartridge system (ACS) into a networked, heterogeneous computer environment. ACSLS software allows the StorageTek 4400 ACS robotic tape subsystem to be used by network-attached computer systems. There are two primary requirements for the ACSLS software: to provide portable high-level software allowing the StorageTek 4400 ACS to be used with a wide variety of computer platforms; and to design the software in layers of functionality that could serve as building blocks for future products. It is shown how these goals were achieved.<<ETX>>","PeriodicalId":282025,"journal":{"name":"[1990] Digest of papers. Tenth IEEE Symposium on Mass Storage Systems@m_Crisis in Mass Storage","volume":"73 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126264848","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}
{"title":"Review of IEEE-CS MSS reference model","authors":"Stephen W. Miller","doi":"10.1109/MASS.1990.113561","DOIUrl":"https://doi.org/10.1109/MASS.1990.113561","url":null,"abstract":"Since publication of the IEEE computer society MSS (mass storage system) reference model, two implementation strategies have been competing in the development of the model. Both strategies stress location transparency of files. However, in one strategy the location transparency is confined within a node on the network, requiring cooperation between differing types of nodes. The other strategy stresses global location transparency across the ecosystem of computation. These strategies are described, and the steps taken to resolve the current issues and pursue the requisite standards are discussed.<<ETX>>","PeriodicalId":282025,"journal":{"name":"[1990] Digest of papers. Tenth IEEE Symposium on Mass Storage Systems@m_Crisis in Mass Storage","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134180066","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}
{"title":"Mass storage system advances at Los Alamos","authors":"B. Collins, C. Mercier, Tom Stup","doi":"10.1109/MASS.1990.113572","DOIUrl":"https://doi.org/10.1109/MASS.1990.113572","url":null,"abstract":"A new packet-switching network being developed as a testbed to evaluate new technologies, including the ANSI standard high-performance parallel interface (HPPI), crossbar switches, and packet buffering and protocol-processing distribution is described. The methods by which mass-storage systems would use and benefit from the new network are being investigated. The testbed includes an IBM 3090-600E running a test version of the Los Alamos common file system (CFS) with a HPPI implementation and a disk-stripping system. >","PeriodicalId":282025,"journal":{"name":"[1990] Digest of papers. Tenth IEEE Symposium on Mass Storage Systems@m_Crisis in Mass Storage","volume":"78 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126059384","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}
J. Foglesong, George Richmond, L. Cassell, C. B. Hogan, J. Kordas, M. Nemanic
{"title":"The Livermore distributed storage system: implementation and experiences","authors":"J. Foglesong, George Richmond, L. Cassell, C. B. Hogan, J. Kordas, M. Nemanic","doi":"10.1109/MASS.1990.113563","DOIUrl":"https://doi.org/10.1109/MASS.1990.113563","url":null,"abstract":"Several key design goals and implementation areas which support the extensibility, modularity, and flexibility of the LINCS storage system are discussed. The separation of data and control messages has had a positive impact on performance and modularity by allowing third-party copies without the actual passing of data through the bitfile servers. Efficient management of storage media and bitfile headers has increased storage utilization and provided integrity of the header information. A network-wide locking mechanism that preserves an object's consistency when accessed concurrently by multiple applications has been designed. The separation of the human-oriented naming mechanism from the other object servers, has given the system and its clients flexibility, extensibility, and modularity not found in an integral naming mechanism. >","PeriodicalId":282025,"journal":{"name":"[1990] Digest of papers. Tenth IEEE Symposium on Mass Storage Systems@m_Crisis in Mass Storage","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130590957","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}
{"title":"Storage management issues for Cray Research","authors":"Thomas W. Lanzatella, P. Rutherford","doi":"10.1109/MASS.1990.113591","DOIUrl":"https://doi.org/10.1109/MASS.1990.113591","url":null,"abstract":"The issues facing Cray Research as a supercomputer vendor resulting from the availability of high-speed networks (100 MB/s) connected to high-capacity storage devices (>1 TB) are discussed. Current capabilities in the area of storage management in the UNICOS operating system are reviewed. Operational requirements stemming from the changing environment along with functional extensions to the UNICOS operating system are discussed. >","PeriodicalId":282025,"journal":{"name":"[1990] Digest of papers. Tenth IEEE Symposium on Mass Storage Systems@m_Crisis in Mass Storage","volume":"59 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125782161","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}