• The KIPS Transactions:PartA
• /
• v.8A no.4
• /
• pp.359-366
• /
• 2001

This paper overviews the design of SANique$^{TM}$ -a shred file system for Linux cluster based on SAN environment. SANique$^{TM}$ has the capability of transferring user data from network-attached SAN disks to client applcations directly without the control of centralized file server system. The paper also presents the characteristics of each SANique$^{TM}$ subsystem: CFM(Cluster File Manager), CVM(Cluster Volume Manager), CLM(Cluster Lock Manager), CBM(Cluster Buffer Manager) and CRM(Cluster Recovery Manager). Under the SANique$^{TM}$ design layout, then, the syndrome of '||'&'||'quot;split-brain'||'&'||'quot; in shared file system environments is described and defined. The work first generalizes and illustrates possible situations in each of which a shared file system environment may split into two or more pieces of separate brain. Finally, the work describes the SANique$^{TM}$ approach to the given "split-brain"problem using SAN disk named "split-brain" and develops the overall recovery procedure of shared file systems.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.11
• /
• pp.1-6
• /
• 2019

Multicast is an efficient method for real-time transmission in many multimedia applications. It is important to recover lost packets and to manage multicast groups according to the network status in order to improve the reliability of multicast transmissions. In this paper, we propose a method that can efficiently recover lost packets in a large multicast group. In the proposed method, we create a recovery cluster (RC) using a multicast domain (MD) for recovery of lost packets. In the conventional methods, clusters send a request message for lost packets to the senders in order to recover the packets lost from many multicast applications. This increases packet delay time and overhead because of the feedback messages and retransmitted packets. In the proposed method, we improve these problems using the RC, which consists of many MDs (which have overlay multicast senders), and many cluster heads (CHs). We divide the message into blocks, and divide each block into many segments for packet recovery using the CHs. When packet loss occurs, all CHs share the segment information and recover the lost segments at the same time. Simulation results show that the proposed method could improve the packet recovery ratio by about 50% compared to the conventional methods.

• Journal of Korea Multimedia Society
• /
• v.7 no.4
• /
• pp.505-517
• /
• 2004

This paper overviews the design details of the cluster file system $\textrm{SANique}^{TM}$ on the SAN environment. $\textrm{SANique}^{TM}$ has the capability of transferring user data from shared SAN disk to client application without control of centralized file server. We, especially, focus on the characteristics and functions of recovery manager CRM of $\textrm{SANique}^{TM}$. The process component for failure detection and its overall procedure are described. We define the split-brain problem that cannot be easily detected in cluster file systems and also propose the recovery management method based on SAN disk in order to detect and solve the split-brain situation.

• The KIPS Transactions:PartC
• /
• v.16C no.3
• /
• pp.363-372
• /
• 2009

For large scale VOD service, cluster servers are spotlighted to their high performance and low cost. A cluster server usually consists of a front-end node and multiple back-end nodes. Though increasing the number of back-end nodes can result in the more QoS streams for clients, the possibility of failures in back-end nodes is proportionally increased. The failure causes not only the stop of all streaming service but also the loss of the current playing positions. In this paper, when a back-end node becomes a failed state, the recovery mechanisms are studied to support the unceasing streaming service. For the actual VOD service environment, we implement a cluster-based VOD servers composed of general PCs and adopt the parallel processing for MPEG movies. From the implemented VOD server, a video block recovery mechanism is designed on parity algorithms. However, without considering the architecture of cluster-based VOD server, the application of the basic technique causes the performance bottleneck of the internal network for recovery and also results in the inefficiency CPU usage of back-end nodes. To address these problems, we propose a new failure recovery mechanism based on the pipeline computing concept.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.16 no.4
• /
• pp.9-17
• /
• 2020

When IoT sensor nodes are deployed in areas where data collection is challenging, sensors must be relocated if sensing holes occur due to improper placement of sensors or energy depletion, and data collection is impossible. The sensing hole's cluster header transmits a request message for sensor relocation to an adjacent cluster header through a specific relay node. However, since a specific relay node is frequently used, a member sensor located in a specific cluster area adjacent to the sensing hole can continuously receive the movement message. In this paper, we propose a method that avoids the situation in which the sensing hole cluster header monopolizes a specific relay node and allows the cluster header to use multiple relay nodes fairly. Unlike the existing method in which the relay node immediately responds to the request of the header, the method proposed in this paper solves a ping-pong problem and a problem that the request message is concentrated on a specific relay node by applying a method of responding to the request of the header using a timer. OMNeT++ simulator was used to analyze the performance of the proposed method.

• The Journal of Fisheries Business Administration
• /
• v.40 no.3
• /
• pp.127-146
• /
• 2009

This study is to estimate the factor weights of the reasons for decreases in marine fishery resources using the Analytical Hierarchy Process. Furthermore, it classifies 20 fishes under a fishery resource recovery plan into various groups of fishes according to these factor weights using the non-hierarchial cluster analysis. The factors of decreases in marine fishery resources are identified as bio-ecological, technology-system, economic-business, and fishing village-society factors. Two of the most important factors of decreases in resource are turned out to be the economic-business and bio-ecological factors, estimated as 31% and 30% respectively. The technology-system and fishing village-society factors are estimated as 21% and 18% respectively. The study utilizes non-hierarchical cluster analysis in order to classify 20 fishes into 2, 3, and 4 groups. K-means cluster analysis is applied for grouping in conjunction with ANOVA to identify statistical differences in factors. Once again, the economic-business and bio-economic factors play main role in grouping 2-groups of fishes case. The third group of fishes in addition to the previous 2 groups of fishes appears as those 4 factors of decrease evenly play about the same role at a 3-groups of fishes case. Finally, the economic-business and bio-economic factors are turned out to be evenly important in the 4th group once there are 4-groups of fishes.

• International Journal of Industrial Entomology and Biomaterials
• /
• v.14 no.1
• /
• pp.23-32
• /
• 2007

The nature of genetic variation and diversity among the 65 multivoltine silkworm genotypes was evaluated for 16 post cocoon characters. The components of genetic variation revealed higher PCV (60.487%) and GCV (44.56%) for evenness (variation 1) followed by cohesion (PCV=55.38%, GCV=40.36%) and non-broken filament length (PCV=32.05%, GCV=31.28%). The higher heritability ($h^2$ in broad sense) was observed for boil-off loss (95.6%) followed by non-broken filament length (95.22%). The both genotypic and phenotypic correlation indicated significant positive correlation of filament length with non-broken filament length, silk recovery, raw silk, neatness, and low neatness; and negative correlation with denier, renditta and silk waste. The principal component analysis (PCA) revealed 75.381 % of total variance from the five principal components extracted. On the basis of Mahalonobis' $D^2$ values (Ward's minimum variance), the sixty-five multivoltine silkworm genotypes were classified in to 9 clusters with substantial inter and intra cluster distances. Number of genotypes included in different clusters varied from 3 to 17. The results indicated that the optimum distance obtained in cluster VII (15.059) along with higher cluster mean values especially for filament length, non broken filament length, renditta, silk recovery, silk waste, and raw silk emphasized the utilization of these genotypes in the conventional silkworm breeding programme for improvement of multivoltine silk fibre quality. The possibility of exploiting genetic variation in post cocoon traits for efficient breeding programme is discussed.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.11 no.4
• /
• pp.17-22
• /
• 2011

The underwater environments are quite different from the terrestrial ones in terms of the communication channel and constrains. In underwater wireless sensor network, the probability of node failure is high because sensor nodes are deployed in more harsh environments than the ground based networks and moved by waves and currents. There are researches considering the communication environments of underwater to improve the data transmission throughput. In this paper, we present a checkpointing scheme of the cluster heads that recoveries from a cluster head failure quickly. Experimental results show that the proposed scheme enhances the reliability of the networks and more efficient in terms of the energy consumption and the recovery latency than without checkpointing.

• Proceedings of the IEEK Conference
• /
• 2003.07e
• /
• pp.2371-2374
• /
• 2003

In transmitting compressed video bit-stream over Internet, packet loss causes error propagation in both spatial and temporal domain, which in turn leads to severe degradation in image qualify In this paper, a new approach for the recovery of lost or erroneous Motion Vector(MV)s by clustering the movements of neighboring blocks by their homogeneity is proposed. MVs of neighboring blocks are clustered according to ALA(Average Linkage Algorithm) clustering and a representative value for each cluster is determined to obtain the candidate MV set. By computing the distortion of the candidates, a MV with the minimum distortion is selected. Experimental results show that the proposed algorithm exhibits better performance in many cases than existing methods.

• The KIPS Transactions:PartA
• /
• v.13A no.7 s.104
• /
• pp.607-614
• /
• 2006

The checkpointing schemes should reduce the process delay through managing the checkpoints of each node to fit the network load to enhance the performance of the process running on the cluster system that write the checkpoints into its global stable storage. For this reason, a cluster system with single IO space on a distributed RAID chooses a suitable checkpointng scheme to get the maximum IO performance and the best rollback recovery efficiency. In this paper, we improved the striped checkpointing scheme with dynamic stripe group size by adapting to the network bandwidth variation at the point of checkpointing. To analyze the performance of the multi striped checkpointing scheme, we applied Linpack HPC benchmark with MPI on our own cluster system with maximum 512 virtual nodes. The benchmark results showed that the multistriped checkpointing scheme has better performance than the striped checkpointing scheme on the checkpoint writing efficiency and rollback recovery at heavy system load.