A five-year study of file-system metadata

Agrawal, Nitin; Bolosky, William J.; Douceur, John R.; Lorch, Jacob R.

doi:10.1145/1288783.1288788

Cited by 236 publications

(176 citation statements)

References 22 publications

Supporting

Mentioning

169

Contrasting

Order By: Relevance

“…The distributions from 1993 and 2005 shown in Figure 9.13 are remarkably similar, with just a slight shift to higher values. Other comparisons of the file-size distributions at the same location over a 20-year span (1984 and 2005) and a 4-year span (2000)(2001)(2002)(2003)(2004) reached essentially the same conclusion [673,18].…”

Section: The Distribution Of File Sizesmentioning

confidence: 61%

See 1 more Smart Citation

Workload Modeling for Computer Systems Performance Evaluation

Feitelson

2015

197

131

View full text Add to dashboard Cite

Reliable performance evaluations require the use of representative workloads. This is no easy task since modern computer systems and their workloads are complex, with many interrelated attributes and complicated structures. Experts often use sophisticated mathematics to analyze and describe workload models, making these models difficult for practitioners to grasp. This book aims to close this gap by emphasizing the intuition and the reasoning behind the definitions and derivations related to the workload models. It provides numerous examples from real production systems, with hundreds of graphs. Using this book, readers will be able to analyze collected workload data and clean it if necessary, derive statistical models that include skewed marginal distributions and correlations, and consider the need for generative models and feedback from the system. The descriptive statistics techniques covered are also useful for other domains.

show abstract

Section: The Distribution Of File Sizesmentioning

confidence: 61%

“…The model can also be extended to allow for link deletions [255]. Essentially the same model has also been suggested in other domains (e.g., the creation of subdirectories in a file system [18]). …”

Section: Preferential Attachmentmentioning

confidence: 98%

Workload Modeling for Computer Systems Performance Evaluation

Feitelson

2015

197

131

View full text Add to dashboard Cite

show abstract

“…Therefore, we may relax the consistency model for media files as they are often non-editable, while the replicas of other documents have to be consistent as they are frequently edited. In addition to that, fetching the document files on demand is not expensive compared to that for media files as the media files size [16] is often bigger than that of documents files [21] on average.…”

Section: Consistency and Availabilitymentioning

confidence: 99%

“…Thus, the access to update rate is often high for these files. Moreover, the size of these media files is often larger than the editable files such as doc or xml [16] [21]. Caching and replicating these files on every possible device improves the access time to those files and results in less network traffic, but it also increases the overhead of maintaining consistency between copies.…”

Section: High Availability With Weak Consistencymentioning

confidence: 99%

vUPS: Virtually Unifying Personal Storage for Fast and Pervasive Data Accesses

Hassan

Bhattarai

Chen

2013

Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering

View full text Add to dashboard Cite

Abstract. More and more overlapping functions on all kinds of mobile devices with these on traditional computers have significantly expanded the usage of mobile devices in our daily life. This also causes the demand surge of pervasively and quickly accessing files across different personal devices owned by a user. Most existing solutions, such as DropBox and SkyDrive, rely on some centralized infrastructure (e.g., cloud storage) to synchronize files across different devices. Therefore, these solutions come with potential risks of user privacy and data secrecy. In addition, continuously maintaining strong consistency among multiple replicas of a file is very costly. On the other hand, today a common user often owns sufficiently large storage space across her personal home desktop, office computer, and mobile devices. Therefore, in this paper, we aim to design and implement a system to virtually Unify Personal Storage (vUPS) for fast and pervasive accesses of personal data across different devices. vUPS provides similar services as offered by existing cloud-based storage services, but (1) vUPS consists of only personal computers without involving any third party, thus it minimizes the risks of user privacy and data secrecy; (2) vUPS organizes all storage in a distributed fashion so that it is not prone to the single point of failure; (3) vUPS differentiates files and maintains different consistency policies in order to reduce the consistency maintenance cost. Having implemented vUPS with HTML5, we conduct extensive experiments to evaluate its performance. The results show that vUPS offers similar user performance when compared to DropBox.

show abstract

“…Moreover, the recent price drop for magnetic disk drives [1] has accelerated the explosive increase in the number of files within typical file systems [2].…”

Section: Introductionmentioning

confidence: 99%

A Method for Searching Keyword-Lacking Files Based on Interfile Relationships

Watanabe

Kobayashi

Yokota

2008

On the Move to Meaningful Internet Systems: OTM 2008 Workshops

View full text Add to dashboard Cite

Abstract. Current information technologies require file systems to contain so many files that searching for desired files is a major problem. To address this problem, desktop search tools using full-text search techniques have been developed. However, those files lacking any given keywords, such as picture files and the source data of experiments, cannot be found by tools based on full-text searches, even if they are related to the keywords. It is even harder to find files located in different directories from the files that include the keywords. In this paper, we propose a method for searching for files that lack keywords but do have an association with them. The proposed method derives relationship information from file access logs in the file server, based on the concept that those files opened by a user in a particular time period are related. We have implemented the proposed method, and evaluated its effectiveness by experiment. The evaluation results indicate that the proposed method is capable of searching keyword-lacking files and has superior precision and recall compared with full-text and directorysearch methods.

show abstract

A five-year study of file-system metadata

Cited by 236 publications

References 22 publications

Workload Modeling for Computer Systems Performance Evaluation

Workload Modeling for Computer Systems Performance Evaluation

vUPS: Virtually Unifying Personal Storage for Fast and Pervasive Data Accesses

A Method for Searching Keyword-Lacking Files Based on Interfile Relationships

Contact Info

Product

Resources

About