1,455 Hits in 4.8 sec

Scaling File Service Up and Out

Garth A. Gibson
2004 USENIX Conference on File and Storage Technologies  
: Capacity balancing: file/dir uses space on all storage: evenly file/dir uses space on all storage: evenly Load balancing: Load balancing: dynamic access to file/dir over all storage: evenly dynamic access  ...  of files file, directory or closely linked set of files Scalable capacity: Scalable capacity: file/dir uses space on all storage: can get big file/dir uses space on all storage: can get big Capacity balancing  ...  Cluster storage for scalable Linux clusters Cluster storage for scalable Linux clusters Garth Gibson Garth Gibson  ... 
dblp:conf/fast/Gibson04 fatcat:zvgok7g7dzagjpswo2z3pwjgve

ReconFS: a reconstructable file system on flash storage

Youyou Lu, Jiwu Shu, Wei Wang
2014 USENIX Conference on File and Storage Technologies  
The inverted indices and logs are used respectively to reconstruct the structure and the content of the directory tree on reconstruction.  ...  However, the update patterns of namespace metadata, such as intensive writeback and scattered small updates, exaggerate the writes to flash storage dramatically, which hurts both performance and endurance  ...  Acknowledgments We would like to thank our shepherd Remzi Arpaci-Dusseau and the anonymous reviewers for their comments and suggestions.  ... 
dblp:conf/fast/LuSW14 fatcat:eajaylbwcbgezlcbcfa7tyqjpy

Towards Semantics for Provenance Security

Stephen Chong
2009 USENIX Conference on File and Storage Technologies  
Moreover, the provenance itself may be sensitive information, necessitating restrictions on the use of both data and provenance to enforce security requirements.  ...  The semantic definitions require that provenance information released to the user does not reveal confidential data, and that neither the provenance information given to the user, nor the program's output  ...  Acknowledgments Thanks to Kim Bruce, Melissa O'Neill, and Chris Stone for giving useful feedback on an earlier version of this work.  ... 
dblp:conf/fast/Chong09 fatcat:4eabjh4s2ffcxgwyvh37low7u4

pFSCK: Accelerating File System Checking and Repair for Modern Storage

David Domingo, Sudarsun Kannan
2021 USENIX Conference on File and Storage Technologies  
We propose and design pFSCK, a parallel file system checking and recovery (C/R) tool designed to exploit compute and storage parallelism in modern storage devices. pFSCK enables fine-grained parallelism  ...  file system.  ...  We also thank Rutgers Panic Lab for helping with the storage infrastructure.  ... 
dblp:conf/fast/DomingoK21 fatcat:3e6o55ebdnf6zjeqtiwub77koa

Spiffy: Enabling File-System Aware Storage Applications

Kuei Sun, Daniel Fryer, Joseph Chu, Matthew Lakier, Angela Demke Brown, Ashvin Goel
2018 USENIX Conference on File and Storage Technologies  
, and an online storage layer cache that preferentially caches files for certain users.  ...  We propose Spiffy, an annotation language for specifying the on-disk format of a file system.  ...  Acknowledgements We thank the anonymous reviewers and our shepherd, André Brinkmann, for their valuable feedback.  ... 
dblp:conf/fast/SunFCLBG18 fatcat:in7oo26llbcnhcz4wiqnfm2g5a

Large-Scale Graph Processing on Emerging Storage Devices

Nima Elyasi, Changho Choi, Anand Sivasubramaniam
2019 USENIX Conference on File and Storage Technologies  
However, emerging storage devices, including currently popular SSDs, provide fairly comparable sequential and random accesses, making these prior solutions inefficient.  ...  We show experimentally on an actual platform that our proposal can give 2X better performance than a state-of-the-art solution.  ...  Acknowledgements We thank Tim Harris, our shepherd, and the anonymous reviewers for their constructive feedback.  ... 
dblp:conf/fast/ElyasiCS19 fatcat:q4wilm6i4bbgvdvjymu373a72y

Isotope: Transactional Isolation for Block Storage

Ji-Yong Shin, Mahesh Balakrishnan, Tudor Marian, Hakim Weatherspoon
2016 USENIX Conference on File and Storage Technologies  
Existing storage stacks are top-heavy and expect little from block storage.  ...  ., providing ACID guarantees), and fast (e.g., 415 MB/s for random file writes).  ...  Acknowledgments This work is partially funded and supported by a SLOAN Research Fellowship received by Hakim Weatherspoon, DARPA MRC and CSSG (D11AP00266) and NSF (1422544, 1053757, 0424422, 1151268, 1047540  ... 
dblp:conf/fast/ShinBMW16 fatcat:zkvafci2tbgmjpu6kx3x6vfiua

MixApart: decoupled analytics for shared storage systems

Madalin Mihailescu, Gokul Soundararajan, Cristiana Amza
2013 USENIX Conference on File and Storage Technologies  
Distributed file systems built for data analytics and enterprise storage systems have very different functionality requirements.  ...  MixApart uses an integrated data caching and scheduling solution to allow MapReduce computations to analyze data stored on enterprise storage systems.  ...  Finally, we thank Sethuraman Subbiah, Deepak Kenchammana, and the rest of the Advanced Technology Group (ATG) for their comments on the early ideas and the initial paper drafts.  ... 
dblp:conf/fast/MihailescuSA13 fatcat:k6hwrsz2vnc3fhwvckucwpc5oy

Quiver: An Informed Storage Cache for Deep Learning

Abhishek Vijaya Kumar, Muthian Sivathanu
2020 USENIX Conference on File and Storage Technologies  
First, Quiver uses a secure hash-based addressing to transparently reuse cached data across multiple jobs and even multiple users operating on the same dataset.  ...  We introduce Quiver, an informed storage cache for deep learning training (DLT) jobs in a cluster of GPUs.  ...  Acknowledgments We thank our shepherd Robert Ross and the anonymous reviewers for their valuable comments and suggestions.  ... 
dblp:conf/fast/KumarS20 fatcat:zdmcbedfobadzo2egj7zail6pe

Host-side Filesystem Journaling for Durable Shared Storage

Andromachi Hatzieleftheriou, Stergios V. Anastasiadis
2015 USENIX Conference on File and Storage Technologies  
Storage virtualization at the file interface achieves clear consistency semantics across data and metadata blocks, supports native file sharing between clients over the same or different hosts, and provides  ...  Hostside caching on durable media is already applied at the block level in order to reduce the load of the storage backend.  ...  This research has been supported by the European Union (European Social Fund -ESF) and Greek national funds through the Operational Program "Education and Lifelong Learning" of the National Strategic Reference  ... 
dblp:conf/fast/Hatzieleftheriou15 fatcat:l3p2boxp2jbprh4etk32vhrmoa

Mirador: An Active Control Plane for Datacenter Storage

Jake Wires, Andrew Warfield
2017 USENIX Conference on File and Storage Technologies  
We show that for performance-dense storage clusters, placement decisions informed by the relative capabilities of net- USENIX Association 15th USENIX Conference on File and Storage Technologies 213  ...  Using approaches from dynamic constraint satisfaction, Mirador migrates both data and client network connections in order to continuously adapt and improve the configuration of the storage system.  ...  We would also like to thank John Wilkes for some very frank and direct comments on an early version of the paper, and Mihir Nanavati for all of his help and feedback along the way.  ... 
dblp:conf/fast/WiresW17 fatcat:agfnvys37zaytf2k3bxavhqn6e

The Case for Browser Provenance

Daniel W. Margo, Margo I. Seltzer
2009 USENIX Conference on File and Storage Technologies  
For example, provenance can improve both history and web search by indicating contextual and personal relationships between data items.  ...  A modern browser now also reads email, plays media, edits documents, and runs applications.  ...  There has been a considerable amount of provenance research on efficient storage and query.  ... 
dblp:conf/fast/MargoS09 fatcat:ijiytwcfwbb65po3kug35dtnki

Non-blocking Writes to Files

Daniel Campello, Hector López, Ricardo Koller, Raju Rangaswami, Luis Useche
2015 USENIX Conference on File and Storage Technologies  
Across the Filebench write workloads, non-blocking writes improve benchmark throughput by 7X on average (up to 45.4X) when using disk drives and by 2.1X on average (up to 4.2X) when using SSDs.  ...  When replaying the MobiBench file system traces, non-blocking writes decrease average operation latency by 20-60%.  ...  Many thanks to Eric Johnson who aided in the collection of file system traces from FIU's production servers.  ... 
dblp:conf/fast/CampelloLKRU15 fatcat:v73khblxsfgdhpshycucxks72m

File recipe compression in data deduplication systems

Dirk Meister, André Brinkmann, Tim Süß
2013 USENIX Conference on File and Storage Technologies  
We propose a combination of efficient and scalable compression schemes to shrink the file recipes' size. A trace-based simulation shows that these methods can compress file recipes by up to 93%.  ...  Data deduplication systems discover and exploit redundancies between different data blocks. The most common approach divides data into chunks and identifies redundancies via fingerprints.  ...  This work was partially supported by the German Federal Ministry of Economics and Technology (BMWi) under project grant SIMBA and by the EU commission within the SCALUS early researchers graduate school  ... 
dblp:conf/fast/MeisterBS13 fatcat:23blfyy3cjepleqrvaj2axprai

WAFL Iron: Repairing Live Enterprise File Systems

Ram Kesavan, Harendra Kumar, Sushrut Bhowmik
2018 USENIX Conference on File and Storage Technologies  
Checking and repairing a file system while it is open for general access present unique challenges.  ...  Consistent and timely access to an arbitrarily damaged file system is an important requirement of enterpriseclass systems.  ...  This extra requirement in memory and storage is negligible on all configurations of ONTAP.  ... 
dblp:conf/fast/KesavanKB18 fatcat:xiqdfozefjfepbmcstaccmhl7e
« Previous Showing results 1 — 15 out of 1,455 results