DOI: 10.3897/jucs.2020.006

Authors: Kohei Hiraga, Osamu Tatebe, Hideyuki Kawashima

Abstract: Metadata performance scalability is critically important in high-performance computing when accessing many small files from millions of clients. This paper proposes a design of a scalable distributed metadata server, PPMDS, for parallel file systems using multiple key-value servers. In PPMDS, hierarchical namespace of a file system is efficiently managed by multiple servers. Multiple entries can be atomically updated using a nonblocking distributed transaction based on an algorithm of dynamic software transactional memory. This paper also proposes optimizations to further improve the metadata performance by introducing a server-side transaction processing, multiple readers, and a shared lock mode, which reduce the number of remote procedure calls and prevent unnecessary blocking. Performance evaluation shows the scalable performance up to 3 servers, and achieves 62,000 operations per second, which is 2.58x performance improvement compared to a single metadata performance.

HTML

XML

PDF

DOI: 10.3217/jucs-025-09-1089

Authors: Mohammed Al-Saleh, Hanan Hamdan

Abstract: The Antivirus (AV) is of an important concern to the end-users community. Mainly, the AV achieves security by scanning data against its database of virus signatures. In addition, the AV tries to reach a pleasant balance between security and United States of Americability. When to scan data is an important design decision an AV has to make. Because AVs are equipped with on-access scanners that scan files when necessary, we want to have a fine-grained approach that provides us with high precision explanation of the performance impact of the AVs on different file system operations. Microsofts minifilter driver technology helps us achieve exactly what we want. By deploying a minifilter driver, we show that most overhead of the tested AVs are greatly imposed on the OPEN operation. Interestingly, we also show that the AV greatly enhances the timing for the READ operation. Finally, the WRITE and CLEANUP operations show almost no differences in terms of performance.

HTML

XML

PDF

DOI: 10.3217/jucs-025-02-0074

Authors: Jędrzej Bieniasz, Krzysztof Szczypiorski

Abstract: This paper summarizes research on methods for information hiding in Open Social Networks. The first contribution is the idea of StegHash, which is based on the use of hashtags in various open social networks to connect multimedia files (such as images, movies, songs) with embedded hidden data. The proof of concept was implemented and tested using a few social media services. The experiments confirmed the initial idea. Next, SocialStegDisc was designed as an application of the StegHash method by combining it with the theory of filesystems. SocialStegDisc provides the basic set of operations for files, such as creation, reading or deletion, by implementing the mechanism of a linked list. It establishes a new kind of mass-storage characterized by unlimited data space, but limited address space where the limitation is the number of the hashtags' unique permutations. The operations of the original StegHash method were optimized by trade-offs between the memory requirements and computation time. Features and limitations were identified and discussed. The proposed system broadens research on a completely new area of threats in social networks.

HTML

XML

PDF

DOI: 10.3217/jucs-011-01-0020

Authors: Sriranjani Sitaraman, S. Venkatesan

Abstract: An important problem in digital forensics is to record a checkpoint of a disk drive mounted as a file system on a host machine without disrupting the disk s normal operations. We present a checkpointing methodology for a disk that has a Unix-like file system. While our algorithm is built around the Unix file system, it can be used to checkpoint disks formatted for other file systems such as NTFS, etc. Our algorithm satisfies several correctness conditions.

HTML

XML

PDF

DOI: 10.3217/jucs-009-08-0813

Authors: Rafael Lins

Abstract: Reference counting is a widely employed memory management technique, in which garbage collection operations are interleaved with computation. Standard reference counting has the major drawback of being unable to handle cyclic structures. This paper presents an important optimisation to a recently published algorithm for cyclic reference counting. Proofs of the correctness of the original and lazy algorithms are provided, together with performance figures.

HTML

XML

PDF

DOI: 10.3217/jucs-001-06-0399

Authors: Artur Klauser, Reinhard Posch

Abstract: Changing relative performance of processors, networks, and disks makes it necessary to reconsider algorithms using these three resources. As networks get faster and less congested topologies emerge, it becomes important to use network resources more aggressively to obtain good performance. Substitution of local disk accesses by accesses to remote memory can lead to better balanced resource usage and thus to faster systems. In this work we address the issue of file caching in a networked file system configuration. Distributed block-level in-memory caches are considered. We show that carefully constructed distributed concepts can lead to lower server load and better overall system performance than centralized concepts. Oversimplification, although aimed at gaining performance for single components, may deteriorate overall performance as a result of unbalanced resource usage.

HTML

XML

PDF