• Journal of KIISE:Computing Practices and Letters
• /
• v.15 no.12
• /
• pp.963-967
• /
• 2009

The Log-structured File System (LFS) collects all modified data into a memory buffer and writes them sequentially to a segment on disk. Therefore, it has the potential to utilize the maximum bandwidth of storage devices where sequential writes are much faster than random writes. However, as disk space is finite, LFS has to conduct cleaning to produce free segments. This cleaning operation is the main reason LFS performance deteriorates when file system utilization is high. To overcome painful cleaning and reduced performance of LFS, we propose the segment space recycling (SSR) scheme that directly writes modified data to invalid areas of the segments and describe the classification method of data and segment to consider locality of reference for optimizing SSR scheme. We implement U-LFS, which employs our segment space recycling scheme in LFS, and experimental results show that SSR scheme increases performance of WOLF by up to 1.9 times in HDD and 1.6 times in SSD when file system utilization is high.

• The KIPS Transactions:PartA
• /
• v.16A no.2
• /
• pp.101-112
• /
• 2009

Existing techniques for defragmentation of the file system need intensive disk operation for some periods at specific time such as disk defragmentation program. In this paper, for solving this problem, we design and implement the automatic and continuous defragmentation free system by distributing the disk operation. We propose the Automatic Layout Scoring(ALS) mechanism for measuring defragmentation degree and suggest the Lazy Copy mechanism that copies the defragmented data at idle time for scattering the disk operation. We search the defragmented file by Automatic Layout Scoring mechanism and then find for empty spaces for that searched file. After lazy copy of searched fils to empty space for preventing that file from being lost, the algorithm solves the defragmentation problem by updating the I-node of that file. We implement these algorithms in Linux and evaluate them for small and defragmented file to get the layout scoring. We outperform the Linux EXT2 file system by $2.4%{\sim}10.4%$ in layout scoring evaluation. And the performance of read and write for various file size is better than the EXT2 by $1%{\sim}8.5%$ for write performance and by $1.2%{\sim}7.5%$ for read performance. We suggest this system for solving the problem of defragmentation automatically without disturbing the I/O task and manual management.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.22 no.6
• /
• pp.1266-1272
• /
• 1997

Due to the price and performanceof uniprocessor workstations and off-the shelf networking, network of workstations(NOW) ae now a cost-effective parallel processing platform tht is competitive with supercomputers. Meanwhile, current network fiile system protocols rely heavily on a central server to coordinate file activity among client workstations. This central server can become a bottleneck that limits scalibility for environments with large numbers of clients. In this paper, we propsoe a highly reliable and effective software RAID file system on the network of workstation environment. We present results form a trace-driven simulation study that shows that the designed software RAID file system is more effective in the aspect of elapsed time when compared with client/server file systems.

• KIPS Transactions on Software and Data Engineering
• /
• v.2 no.2
• /
• pp.91-96
• /
• 2013

The application field of supercomputing systems are changing to support into the field for both a large-volume data processing and high-performance computing at the same time such as bio-applications. These applications require high-performance distributed file system for storage management and efficient high-speed processing of large amounts of data that occurs. In this paper, we introduce MAHA-FS for supercomputing systems for processing large amounts of data and high-performance computing, providing excellent metadata operation performance and IO performance. It is shown through performance analysis that MAHA-FS provides excellent performance in terms of the metadata processing and random IO processing.

• The KIPS Transactions:PartA
• /
• v.13A no.3 s.100
• /
• pp.241-252
• /
• 2006

In cluster file systems, the availability of files has been supported by replicating entire files or generating parities on parity servers. However, those methods require very large temporal and spatial cost, and cannot handle massive failures situation on the file system. So we propose HA-PVFS, a cluster file system supporting high data availability adaptive to temporal locality. HA-PVFS restricts replication or parity generation to some important files, for that it employs an efficient algorithm to estimate file access patterns from limited information. Moreover, in order to minimize the performance degradation of the file system, it uses delayed update method and relay replication.

• Journal of KIISE
• /
• v.43 no.7
• /
• pp.744-750
• /
• 2016

Recent cloud infrastructures provide competitive performances and operation costs for many internet services through pay-per-use model. Particularly, object storages are highlighted, as they have unlimited file holding capacity and allow users to access the stored files anytime and anywhere. Several lines of research are based on cloud-backed file systems, which support traditional POSIX interface rather than RESTful APIs via HTTP. However, these existing file systems handle all files with uniform size backing objects. Consequently, the accesses to cloud object storages are likely to be inefficient. In our research, files are profiled according to characteristics, and appropriate backing unit sizes are determined. We experimentally verify that different backing unit sizes for the object storage improve the performance of cloud-backed file systems. In our comparative experiments with S3QL, our prototype cloud-backed file system shows faster performance by 18.6% on average.

• The Journal of the Korea Contents Association
• /
• v.21 no.7
• /
• pp.157-163
• /
• 2021

A single write operation in a file system can modify multiple data, but these changes in the file system are not atomically written to disk. Thus, for the consistency of the file system, conventional journaling guarantees crash consistency instead of sacrificing the system performance. It is known that using non-volatile memory as a journal space can alleviate performance degradation due to low latency and byte-level accessibility of non-volatile memory. However, none of the journaling techniques considering non-volatile memory provide scalability. In this paper, journal space on non-volatile memory is divided into multiple regions for scalable journaling, thus dispersing concentrated operations in one region. Second, the journal area-specific operator structure is used to accelerate data write operations to storage devices. We apply the proposed technique to JFS to evaluate it on multi-core servers equipped with high-performance storage devices. The evaluation results show that the proposed technique performs better than the existing technique of the NVM-based journaling file system.

• The Transactions of the Korea Information Processing Society
• /
• v.4 no.11
• /
• pp.2849-2860
• /
• 1997

In this paper, we present efficient cache replacement policies for the software RAID file system. The performance of this policies is compared to two other policies previously proposed for conventional file systems and adapted for the software RAID file system. As in hardware RAID systems, we found small-writes to be the performance bottleneck in software RAID file systems. To tackle this small-write problem, we propose cache replacement policies. Using trace driven simulations we show that the proposed policies improve performance in the aspect of the average response time and the average system busy time.

• Journal of the Society of Disaster Information
• /
• v.17 no.1
• /
• pp.176-183
• /
• 2021

Purpose: The introduction of smart factories that reflect the 4th industrial revolution technologies such as AI, IoT, and VR, has been actively promoted in Korea. However, in order to solve various problems arising from existing file-based operating systems, this research will focus on identifying and verifying non-file system-based data protection technology. Method: The research will measure security storage that cannot be identified or controlled by the operating system. How to activate secure storage based on the input of digital key values. Establish a control unit that provides input and output information based on BIOS activation. Observe non-file-type structure so that mapping behavior using second meta-data can be performed according to the activation of the secure storage. Result: First, the creation of non-file system-based secure storage's data input/output were found to match the hash function value of the sample data with the hash function value of the normal storage and data. Second, the data protection performance experiments in secure storage were compared to the hash function value of the original file with the hash function value of the secure storage after ransomware activity to verify data protection performance against malicious ransomware. Conclusion: Smart factory technology is a nationally promoted technology that is being introduced to the public and this research implemented and experimented on a new concept of data protection technology to protect crucial data within the information system. In order to protect sensitive data, implementation of non-file-type secure storage technology that is non-dependent on file system is highly recommended. This research has proven the security and safety of such technology and verified its purpose.

• Journal of KIISE:Computer Systems and Theory
• /
• v.29 no.8
• /
• pp.472-483
• /
• 2002

This paper presents the design and implementation of a cache file system, umcFS, which is specifically designed to provide an efficient caching and transmission of streaming media. The proposed file system is based on the concept of file disk and implemented as an application level on top of a general-purpose file system. The file disk favors the continuity of cached media and provides an efficient I/O mechanism for cache server. umcFS statically allocates control blocks as well as media cache blocks. These blocks are referenced by the single-level indirect management structure. As the file system is designed as an application level, it is easy to develop and port to other systems. The performance of the implemented system shows that umcFS performs about 13% better than the native file system in randomly accessing the cache blocks of 1024KB.