DOCSLIB.ORG

Richer File System Metadata Using Links and Attributes

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Richer File System Metadata Using Links and Attributes Richer File System Metadata Using Links and Attributes Alexander Ames Nikhil Bobb Scott A. Brandt Adam Hiatt [email protected] [email protected] [email protected] [email protected] Carlos Maltzahn Ethan L. Miller Alisa Neeman Deepa Tuteja [email protected] [email protected] [email protected] [email protected] Storage Systems Research Center Jack Baskin School of Engineering University of California, Santa Cruz Abstract clude not only arbitrary, user-specifiable key-value pairs on files but relationships between files in form of links with at- Traditional file systems provide a weak and inadequate tributes. structure for meaningful representations of file interrela- It is now often easier to find a document on the Web tionships and other context-providing metadata. Existing amongst billions of documents than on a local file system. designs, which store additional file-oriented metadata ei- Documents on the Web are embedded in a rich hyperlink ther in a database, on disk, or both are limited by the structure while files typically are not. Companies such as technologies upon which they depend. Moreover, they do Google are able to take advantage of links between Web not provide for user-defined relationships among files. To documents in order to deliver meaningful ranking of search address these issues, we created the Linking File System results using algorithms like PageRank [32]. In contrast, (LiFS), a file system design in which files may have both search tools for traditional file systems do not have infor- arbitrary user- or application-specified attributes, and at- mation about inter-file relationships other than the hierar- tributed links between files. In order to assure performance chical directory structure and ownership of files. when accessing links and attributes, the system is designed The reason for this dearth of relationships between files to store metadata in non-volatile memory. This paper dis- is that the management of file system metadata is expen- cusses several use cases that take advantage of this ap- sive in traditional system architectures where volatile main proach and describes the user-space prototype we devel- memory must be backed by much slower disk-based sec- oped to test the concepts presented. ondary storage. The advent of new non-volatile main mem- ory technologies such as MRAM [8] promises to reduce the cost of accessing file systems in an arbitrary or fine- 1. Introduction grain fashion with the development of novel file system de- signs [16, 27, 49]. Traditional file systems provide a weak and inadequate The promise of non-volatile main memory has prompted structure for meaningful representations of file interrela- designers to use such memory for the persistent storage of tionships and other context-providing metadata. Solving file system metadata. Although memory-resident metadata this problem has become increasingly urgent as users are trivially speeds up certain common file system operations faced with a growing amount of personal data such as (such as stat), a more remarkable benefit is the ability to email, chat communications, digital photography, and on- employ far richer metadata structures. File systems design- line music. Moreover, computational scientists continue ers will no longer be constrained by disk access speed, but to bemoan the lack of mechanisms for cross-archival ac- instead can focus on the needs of the user. Frequent ac- cess, retrofitting of metadata, and identifying groupings of cess, context-aware searches, and other operations that are related results needed for data mining [38, 39]. A content- prohibitively expensive under traditional, disk-oriented file neutral, file system-based mechanism for storage of arbi- system architectures will become feasible. trary metadata provides one solution to this weakness. As The Linking File System’s ability to assign attributes to an application of this principle, we introduce the Linking and establish attributed links between files in a standard- File System (LiFS). It extends file system metadata to in- ized fashion forms a powerful infrastructure capable of sup- 101 porting a variety of different and extremely useful user, ap- linkingfs. plication, and system operations. Attributed files directly pdf dependency: support enhanced file system searches. Attributed links source will support a number of recent efforts to extend hierar- chical directory structures with more user-friendly and per- hierarchy:main linkingfs. sonalized file organizations [18, 28, 31, 35, 43]. Weighted tex mmst05- hierarchy:main links between files can also be used to record access pat- linkingfs terns that are useful for pre-fetching, hoarding, indexing, dependency:included and search result ranking. Indeed, these links provide an abstract model for file interrelationships previously un- hierarchy:main available at a file system level. ieee.cls 2. The Design of LiFS cvs hierarchy:main hierarchy:main The key features of LiFS are links between files and at- hierarchy:main tributes on both files and links. To ensure the performance templates and reliability of LiFS, the design relies on both the non- volatility and low latency of MRAM. At the most basic fast02- level, a search within the file system traverses a series of miller links across a graph of metadata. An in-memory structure is crucial for this operation; random seeks on disk to in- odes, even with some caching, would be prohibitively slow. Figure 1. Example use of links with attributes: The low latency of MRAM will allow metadata operations organizing files into a directory-like hierarchy and searches to be performed almost instantaneously. called “main” and keeping track of depen- dencies between files. 2.1. Links of each key and its corresponding value is also currently Each link in LiFS has a source file, a target file, and a unconstrained. We anticipate putting an efficiency ceiling non-empty set of attributes consisting of key-value pairs. on this size into the final implementation; excessively long LiFS links differ from POSIX links in that LiFS links rep- key and value strings might affect the ability to deliver high resent a relationship between files instead of simply a refer- performance. ence to a file. The attributes of the links express the nature Both the key and value members of an attribute can con- of the relationship. tain arbitrary data, including binary data. This allows appli- Any file can potentially contain a link to any other file. cations to have rich metadata such as thumbnails, preview As a consequence, every file is also a directory and the dis- video clips, and cached printer spool files without the over- tinction between files and directories is eliminated. The tra- head of special encoding. Sharing of metadata via file and ditional notion of containment of a file within a directory is link attributes provides a powerful infrastructure for appli- simply one relationship among many that can be expressed cation integration. with links. The main benefit of attributes is that they enable users, The key benefit of links is that they provide native sup- applications, and the system itself to annotate files and port for a variety of relationships between files that are cur- links. This allows for fast and effective file searching, cat- rently supported in an ad hoc manner by individual appli- egorizing, partitioning, and manipulation, and provides in- cations and the operating system. In addition to contain- frastructure for other features that may not have been con- ment, links can express a variety of other useful relation- sider by the file system designer. It also provides a far ships such as included-in, referenced-by, dependent-upon, richer context in which files can include information about created-by, opened-by, and others. Links also allow for dy- provenance, intended use, type, contents, creator, modifi- namically customizable views of the file system based on cation history, version, and other information that a user, the type of link followed. application, or system may want to keep. We also allow attributes to be executable. A special case 2.2. Attributes of executable file attributes are file triggers. A file trigger on a file specifies a pattern/action pair. A pattern specifies In our initial design both files and links can carry a num- the file system operation (such as a read or write) on the ber of attributes limited only by available memory. The size file with which the executable attribute is associated. An 102 action specifies code that is executed whenever the asso- cites the target. This allows finding documents by specify- ciated operation is invoked on the file. File triggers are a ing “citation paths.” powerful mechanism that simplifies the implementation of Nonetheless, in order for users to properly identify files a wide variety of file system services such as versioning, that they may wish to open, there must be a traversal of mirroring, and others (as discussed in Section 3.5). File identifiable links beginning with a file system entry point to triggers raise a rich set of interesting security and language the target file, akin to root, or a path relative to the current design issues that we are actively investigating. working directory. Unlike conventional file systems, there need not be the same root directory for all users; in fact, a 2.3. Interface user may be able to choose between one of many roots. LiFS supports an enhanced version of the POSIX inter- 2.3.3. Link Accesses, Updates and Deletions In LiFS, face. Here, we present the key examples of the linking API. we introduce API calls to retrieve a single link, perform up- dates to the link, and delete a link. Just as users must open 2.3.1. Link Creation We introduce a new system call, a single file in conventional systems, there are cases where rellink, to create new (relational) links.
Load more

Recommended publications
  • A Searchable-By-Content File System
    A Searchable-by-Content File System Srinath Sridhar, Jeffrey Stylos and Noam Zeilberger May 2, 2005 Abstract Indexed searching of desktop documents has recently become popular- ized by applications from Google, AOL, Yahoo!, MSN and others. How- ever, each of these is an application separate from the file system. In our project, we explore the performance tradeoffs and other issues encountered when implementing file indexing inside the file system. We developed a novel virtual-directory interface to allow application and user access to the index. In addition, we found that by deferring indexing slightly, we could coalesce file indexing tasks to reduce total work while still getting good performance. 1 Introduction Searching has always been one of the fundamental problems of computer science, and from their beginnings computer systems were designed to support and when possible automate these searches. This support ranges from the simple-minded (e.g., text editors that allow searching for keywords) to the extremely powerful (e.g., relational database query languages). But for the mundane task of per- forming searches on users’ files, available tools still leave much to be desired. For the most part, searches are limited to queries on a directory namespace—Unix shell wildcard patterns are useful for matching against small numbers of files, GNU locate for finding files in the directory hierarchy. Searches on file data are much more difficult. While from a logical point of view, grep combined with the Unix pipe mechanisms provides a nearly universal solution to content-based searches, realistically it can be used only for searching within small numbers of (small-sized) files, because of its inherent linear complexity.
    [Show full text]
  • A Semantic File System for Integrated Product Data Management', Advanced Engineering Informatics, Vol
    Citation for published version: Eck, O & Schaefer, D 2011, 'A Semantic File System for Integrated Product Data Management', Advanced Engineering Informatics, vol. 25, no. 2, pp. 177-184. https://doi.org/10.1016/j.aei.2010.08.005 DOI: 10.1016/j.aei.2010.08.005 Publication date: 2011 Document Version Publisher's PDF, also known as Version of record Link to publication University of Bath Alternative formats If you require this document in an alternative format, please contact: [email protected] General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. Take down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Download date: 02. Oct. 2021 Advanced Engineering Informatics 25 (2011) 177–184 Contents lists available at ScienceDirect Advanced Engineering Informatics journal homepage: www.elsevier.com/locate/aei A semantic file system for integrated product data management ⇑ Oliver Eck a, , Dirk Schaefer b a Department of Computer Science, HTWG Konstanz, Germany b Systems Realization Laboratory, Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA, USA article info abstract Article history: A mechatronic system is a synergistic integration of mechanical, electrical, electronic and software tech- Received 31 October 2009 nologies into electromechanical systems. Unfortunately, mechanical, electrical, and software data are Received in revised form 10 June 2010 often handled in separate Product Data Management (PDM) systems with no automated sharing of data Accepted 17 August 2010 between them or links between their data.
    [Show full text]
  • Using Virtual Directories Prashanth Mohan, Raghuraman, Venkateswaran S and Dr
    1 Semantic File Retrieval in File Systems using Virtual Directories Prashanth Mohan, Raghuraman, Venkateswaran S and Dr. Arul Siromoney {prashmohan, raaghum2222, wenkat.s}@gmail.com and [email protected] Abstract— Hard Disk capacity is no longer a problem. How- ‘/home/user/docs/univ/project’ directory lists the ever, increasing disk capacity has brought with it a new problem, file. The ‘Database File System’ [2], encounters this prob- the problem of locating files. Retrieving a document from a lem by listing recursively all the files within each directory. myriad of files and directories is no easy task. Industry solutions are being created to address this short coming. Thereby a funneling of the files is done by moving through We propose to create an extendable UNIX based File System sub-directories. which will integrate searching as a basic function of the file The objective of our project (henceforth reffered to as system. The File System will provide Virtual Directories which SemFS) is to give the user the choice between a traditional list the results of a query. The contents of the Virtual Directory heirarchial mode of access and a query based mechanism is formed at runtime. Although, the Virtual Directory is used mainly to facilitate the searching of file, It can also be used by which will be presented using virtual directories [1], [5]. These plugins to interpret other queries. virtual directories do not exist on the disk as separate files but will be created in memory at runtime, as per the directory Index Terms— Semantic File System, Virtual Directory, Meta Data name.
    [Show full text]
  • Privacy Engineering for Social Networks
    UCAM-CL-TR-825 Technical Report ISSN 1476-2986 Number 825 Computer Laboratory Privacy engineering for social networks Jonathan Anderson December 2012 15 JJ Thomson Avenue Cambridge CB3 0FD United Kingdom phone +44 1223 763500 http://www.cl.cam.ac.uk/ c 2012 Jonathan Anderson This technical report is based on a dissertation submitted July 2012 by the author for the degree of Doctor of Philosophy to the University of Cambridge, Trinity College. Technical reports published by the University of Cambridge Computer Laboratory are freely available via the Internet: http://www.cl.cam.ac.uk/techreports/ ISSN 1476-2986 Privacy engineering for social networks Jonathan Anderson In this dissertation, I enumerate several privacy problems in online social net- works (OSNs) and describe a system called Footlights that addresses them. Foot- lights is a platform for distributed social applications that allows users to control the sharing of private information. It is designed to compete with the performance of today’s centralised OSNs, but it does not trust centralised infrastructure to en- force security properties. Based on several socio-technical scenarios, I extract concrete technical problems to be solved and show how the existing research literature does not solve them. Addressing these problems fully would fundamentally change users’ interactions with OSNs, providing real control over online sharing. I also demonstrate that today’s OSNs do not provide this control: both user data and the social graph are vulnerable to practical privacy attacks. Footlights’ storage substrate provides private, scalable, sharable storage using untrusted servers. Under realistic assumptions, the direct cost of operating this storage system is less than one US dollar per user-year.
    [Show full text]
  • Modern Infrastructure for Dummies®, Dell EMC #Getmodern Special Edition
    Modern Infrastructure Dell EMC #GetModern Special Edition by Lawrence C. Miller These materials are © 2017 John Wiley & Sons, Ltd. Any dissemination, distribution, or unauthorized use is strictly prohibited. Modern Infrastructure For Dummies®, Dell EMC #GetModern Special Edition Published by: John Wiley & Sons Singapore Pte Ltd., 1 Fusionopolis Walk, #07-01 Solaris South Tower, Singapore 138628, www.wiley.com © 2017 by John Wiley & Sons Singapore Pte Ltd. Registered Office John Wiley & Sons Singapore Pte Ltd., 1 Fusionopolis Walk, #07-01 Solaris South Tower, Singapore 138628 All rights reserved No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, except as permitted by the Copyright Act of Singapore 1987, without the prior written permission of the Publisher. For information about how to apply for permission to reuse the copyright material in this book, please see our website http://www.wiley.com/go/ permissions. Trademarks: Wiley, For Dummies, the Dummies Man logo, The Dummies Way, Dummies.com, Making Everything Easier, and related trade dress are trademarks or registered trademarks of John Wiley & Sons, Inc. and/or its affiliates in the United States and other countries, and may not be used without written permission. Dell EMC and the Dell EMC logo are registered trademarks of Dell EMC. All other trademarks are the property of their respective owners. John Wiley & Sons Singapore Pte Ltd., is not associated with any product or vendor mentioned in this book. LIMIT OF LIABILITY/DISCLAIMER OF WARRANTY: WHILE THE PUBLISHER AND AUTHOR HAVE USED THEIR BEST EFFORTS IN PREPARING THIS BOOK, THEY MAKE NO REPRESENTATIONS OR WARRANTIES WITH RESPECT TO THE ACCURACY OR COMPLETENESS OF THE CONTENTS OF THIS BOOK AND SPECIFICALLY DISCLAIM ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
    [Show full text]
  • Orion File System : File-Level Host-Based Virtualization
    Orion File System : File-level Host-based Virtualization Amruta Joshi Faraz Shaikh Sapna Todwal Pune Institute of Computer Pune Institute of Computer Pune Institute of Computer Technology, Technology, Technology, Dhankavadi, Pune 411043, India Dhankavadi, Pune 411043, India Dhankavadi, Pune 411043, India 020-2437-1101 020-2437-1101 020-2437-1101 [email protected] [email protected] [email protected] Abstract— The aim of Orion is to implement a solution that The automatic indexing of files and directories is called provides file-level host-based virtualization that provides for "semantic" because user programmable transducers use better aggregation of content/information based on information about these semantics of files to extract the semantics and properties. File-system organization today properties for indexing. The extracted properties are then very closely mirrors storage paradigms rather than user- stored in a relational database so that queries can be run access paradigms and semantic grouping. All file-system against them. Experimental results from our semantic file hierarchies are containers that are expressed based on their system implementation ORION show that semantic file physical presence (a separate drive letter on Windows, or a systems present a more effective storage abstraction than the particular mount point based on the volume in Unix). traditional tree structured file systems for information We have implemented a solution that will allow sharing, storage and retrieval. users to organize their files
    [Show full text]
  • The Sile Model — a Semantic File System Infrastructure for the Desktop
    The Sile Model — A Semantic File System Infrastructure for the Desktop Bernhard Schandl and Bernhard Haslhofer University of Vienna, Department of Distributed and Multimedia Systems {bernhard.schandl,bernhard.haslhofer}@univie.ac.at Abstract. With the increasing storage capacity of personal computing devices, the problems of information overload and information fragmen- tation become apparent on users’ desktops. For the Web, semantic tech- nologies aim at solving this problem by adding a machine-interpretable information layer on top of existing resources, and it has been shown that the application of these technologies to desktop environments is helpful for end users. Certain characteristics of the Semantic Web archi- tecture that are commonly accepted in the Web context, however, are not desirable for desktops; e.g., incomplete information, broken links, or disruption of content and annotations. To overcome these limitations, we propose the sile model, an intermediate data model that combines attributes of the Semantic Web and file systems. This model is intended to be the conceptual foundation of the Semantic Desktop, and to serve as underlying infrastructure on which applications and further services, e.g., virtual file systems, can be built. In this paper, we present the sile model, discuss Semantic Web vocabularies that can be used in the context of this model to annotate desktop data, and analyze the performance of typical operations on a virtual file system implementation that is based on this model. 1 Introduction A large amount of information is stored on personal desktops. We use our per- sonal computing devices—both mobile and stationary—to communicate, to write documents, to organize multimedia content, to search for and retrieve informa- tion, and much more.
    [Show full text]
  • Propeller: a Scalable Metadata Organization for a Versatile Searchable File System
    University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln Computer Science and Engineering, Department CSE Technical reports of Spring 3-29-2011 Propeller: A Scalable Metadata Organization for A Versatile Searchable File System Lei Xu University of Nebraska-Lincoln, [email protected] Hong Jiang University of Nebraska-Lincoln, [email protected] Xue Liu University of Nebraska-Lincoln, [email protected] Lei Tian University of Nebraska-Lincoln, [email protected] Yu Hua University of Nebraska-Lincoln, [email protected] See next page for additional authors Follow this and additional works at: https://digitalcommons.unl.edu/csetechreports Part of the Computer and Systems Architecture Commons, Databases and Information Systems Commons, and the OS and Networks Commons Xu, Lei; Jiang, Hong; Liu, Xue; Tian, Lei; Hua, Yu; and Hu, Jian, "Propeller: A Scalable Metadata Organization for A Versatile Searchable File System" (2011). CSE Technical reports. 119. https://digitalcommons.unl.edu/csetechreports/119 This Article is brought to you for free and open access by the Computer Science and Engineering, Department of at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in CSE Technical reports by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. Authors Lei Xu, Hong Jiang, Xue Liu, Lei Tian, Yu Hua, and Jian Hu This article is available at DigitalCommons@University of Nebraska - Lincoln: https://digitalcommons.unl.edu/ csetechreports/119 Propeller: A Scalable Metadata Organization for A Versatile Searchable File System Lei Xu Hong Jiang Xue Liu University of Nebraska-Lincoln University of Nebraska-Lincoln University of Nebraska-Lincoln [email protected] [email protected] [email protected] Lei Tian Yu Hua Jian Hu University of Nebraska-Lincoln University of Nebraska-Lincoln University of Nebraska-Lincoln [email protected] [email protected] [email protected] ABSTRACT prone.
    [Show full text]
  • Dynamic Metadata Management in Semantic File Systems
    M. Suguna Int. Journal of Engineering Research and Applications www.ijera.com ISSN : 2248-9622, Vol. 5, Issue 3, ( Part -5) March 2015, pp.44-47 RESEARCH ARTICLE OPEN ACCESS Dynamic Metadata Management in Semantic File Systems M. Suguna1, T. Anand2 1P.G Student& A.V.C College of Engineering, Mayiladuthurai, Tamil Nadu. 2Professor & A.V.C College of Engineering, Mayiladuthurai, Tamil Nadu. ABSTRACT The progression in data capacity and difficulty inflicts great challenges for file systems. To address these contests, an inventive namespace management scheme is in distracted need to deliver both the ease and competence of data access. For scalability, each server makes only local, autonomous decisions about relocation for load balancing. Associative access is provided by a traditional extension to present tree-structured file system conventions, and by protocols that are intended specifically for content based access.Rapid attribute- based accesstofile system contents is fulfilled by instinctive extraction and indexing of key properties of file system objects. The programmed indexing of files and calendars is called “semantic” because user programmable transducers use data about the semantics of efficient file system objects to extract the properties for indexing. Tentative results from a semantic file system execution support the thesis that semantic file systems present a more active storage abstraction than do traditional tree planned file systems for data sharing and command level programming. Semantic file system is executed as a middleware in predictable file systems and works orthogonally with categorized directory trees. The semantic relationships and file groups recognized in file systems can also be used to facilitate file prefetching among other system-level optimizations.
    [Show full text]
  • Rlinks: a Mechanism for Navigating to Related Files
    University of Kentucky UKnowledge University of Kentucky Master's Theses Graduate School 2007 RLINKS: A MECHANISM FOR NAVIGATING TO RELATED FILES Naveen Akarapu University of Kentucky, [email protected] Right click to open a feedback form in a new tab to let us know how this document benefits ou.y Recommended Citation Akarapu, Naveen, "RLINKS: A MECHANISM FOR NAVIGATING TO RELATED FILES" (2007). University of Kentucky Master's Theses. 467. https://uknowledge.uky.edu/gradschool_theses/467 This Thesis is brought to you for free and open access by the Graduate School at UKnowledge. It has been accepted for inclusion in University of Kentucky Master's Theses by an authorized administrator of UKnowledge. For more information, please contact [email protected]. ABSTRACT OF THESIS RLINKS: A MECHANISM FOR NAVIGATING TO RELATED FILES This thesis introduces Relative links or rlinks, which are directed labeled links from one file to another in a file system. Rlinks provide a clean way to build and share related-file information without creating additional files and directories. Rlinks form overlay graphs between files of a file system, thus providing useful alternate views of the file system. This thesis implements rlinks for the Linux kernel and modifies the storage structure of the Ext2 file system to store the rlinks. KEYWORDS: Rlinks, Linux kernel, Ext2, related files, file system Naveen Akarapu 08/03/2007 RLINKS: A MECHANISM FOR NAVIGATING TO RELATED FILES By Naveen Akarapu Dr. Raphael A. Finkel (Director of Thesis) Dr. Raphael A. Finkel (Director of Graduate Studies) August 3, 2007 RULES FOR THE USE OF THESES Unpublished theses submitted for the Master s degree and deposited in the University of Kentucky Library are as a rule open for inspection, but are to be used only with due regard to the rights of the authors.
    [Show full text]
  • A Semantic File System for Integrated Product Data Management', Advanced Engineering Informatics, Vol
    View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by University of Bath Research Portal Citation for published version: Eck, O & Schaefer, D 2011, 'A Semantic File System for Integrated Product Data Management', Advanced Engineering Informatics, vol. 25, no. 2, pp. 177-184. https://doi.org/10.1016/j.aei.2010.08.005 DOI: 10.1016/j.aei.2010.08.005 Publication date: 2011 Document Version Publisher's PDF, also known as Version of record Link to publication University of Bath General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. Take down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Download date: 13. May. 2019 Advanced Engineering Informatics 25 (2011) 177–184 Contents lists available at ScienceDirect Advanced Engineering Informatics journal homepage: www.elsevier.com/locate/aei A semantic file system for integrated product data management ⇑ Oliver Eck a, , Dirk Schaefer b a Department of Computer Science, HTWG Konstanz, Germany b Systems Realization Laboratory, Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA, USA article info abstract Article history: A mechatronic system is a synergistic integration of mechanical, electrical, electronic and software tech- Received 31 October 2009 nologies into electromechanical systems. Unfortunately, mechanical, electrical, and software data are Received in revised form 10 June 2010 often handled in separate Product Data Management (PDM) systems with no automated sharing of data Accepted 17 August 2010 between them or links between their data.
    [Show full text]
  • A Semantic File System for Embedded Systems
    JOINFS: a Semantic File System for Embedded Systems Matthew Harlan Gabriel Parmer Computer Science Department The George Washington University Washington, DC fmharlan,[email protected] Abstract (global) coordination and classification, and is naturally or- ganized by search. Fundamentally, complex data often has Hierarchical file systems are the de-facto abstraction for no single natural category, and instead has many attributes, storing information on modern computing systems. Though any one of which might be useful as a discriminator for a they are useful for providing structure for categorical data, user or program. their failings are most pronounced on consumer embedded Others have claimed that “hierarchical file systems are devices. The limited interface for organizing a directory hi- dead”[6], and that search-based data-access should be the erarchy, and abundant structured, network-accessible data dominant form of organization. Indeed, commodity operat- complicate the management of the hierarchy. ing systems have integrated search functionality in the form JOINFS attempts to enhance the tradition hierarchical of Windows search [9] and Apple Spotlight [7]. Such con- file system abstraction by marrying it with an efficient query cepts are not new: twenty years ago, the semantic filesys- system. Special dynamic directories are populated not by a tem [3] was proposed in which directories are organized by rigid hierarchy, but by an active matching of query terms attributes, and in ’93 the Inversion FS implemented a file to meta-data associated with the files in the system. Hier- system on top of a data-base [4] while exposing querying archical dynamic directories provide the novel concept of functionality.
    [Show full text]