• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.17 no.1
• /
• pp.90-100
• /
• 2013

Active-Standby Duplication Techniques are conventionally used for fault-tolerant systems. But in this paper we researched on the Active-Active Duplication Techniques for Fault-tolerant DNS System. Our Active-Active Duplication made the 1st DNS periodically copied to the 2nd DNS and maintained the same status by using Rsync and Crontab. Even though the 1st or the 2nd DNS stops due to some critical errors, the remaining DNS can take over and provide continuous services.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.18 no.1
• /
• pp.63-71
• /
• 2014

In this paper we researched on the Duplication Techniques for Active-Active Web Servers. Rsync and crontab utilities make copy periodically between web servers and maintain the same status. Load Balancing Server makes web servers load balanced and fast servicing by executing web servers alternatively. Even though one web server stops due to some critical errors, the remaining web server can take over and provide services continuously.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2014.04a
• /
• pp.105-108
• /
• 2014

기업에서 사용하는 시스템에서는 대량의 데이터를 실시간으로 처리해야 하며 또한 시스템 장애로 인한 생산 중단이 발생하지 않도록 하기 위해 서버 이중화 구성에 많은 투자를 하고 있다. 장애를 최소화하기 위해 서버 구성을 Active-Standby 구조로 이중화를 구성하여 사용하고 있으며, 이로 인해 시스템에 대한 많은 비용을 투자하고 있다. 서버 부하 감소 및 서버 가용성 향상을 통한 투자 비용을 절감하기 위해 서버 분산 처리에 대한 방법이 나오고 있다. 그러나 Network 스위치나 라우터 등의 Hardware 적인 분산 시스템을 이용한 방법으로는 특정한 데이터만을 분산하여 처리할 수 없고 Hardware 의 투자 비용 또한 증가하게 된다. 또한 서버의 Active-Standby 구성에서 Standby 서버는 Active 서버의 문제 발생시에만 이용하게 됨으로 인해 서버의 가용성이 떨어지는 불합리가 발생함으로 인해 이를 개선하기 위한 연구를 통해 Active-Active 구성을 통해 서버의 가용성을 확보하고자 한다. 본 논문에서는 서버 가용성 향상을 위한 방법으로 DQ(Distributed Queue)를 이용하여 응용프로그램에서의 데이터 분산을 통한 응용프로그램에서의 부하를 분산하여 서버 내 리소스를 최대한 활용하면서 응용프로그램에서의 부하를 감소화할 수 있는 방법을 검증하고자 한다.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.25 no.8A
• /
• pp.1104-1114
• /
• 2000

If the control system that works important job fails, economical loss occurred. Hence, to guarantee high reliability, it must be duplicated. In the case of traditional duplication mechanism, dedicated operating system with duplication functions were built. This required much development and maintenance cost. They can be saved, if we use commercial operating system and its development environment.This paper proposes a duplication mechanism for the system based on commercial 0S. The system that explained in this paper is BSC(Base Station Controller). The duplicated BSC system uses concurrent write memory for synchronization and VxWorks as an operating system. We propose how the task supporting duplication functions is executed without delay and preemption, how to synchronize standby's memory with active' s, and how to use concurrent-write memory easily with VxWorks' s partition. We also describe the takeover procedure when the active detects its hardware fault and when the standby recognizes the failure of the active.

• Proceedings of the IEEK Conference
• /
• 2000.11a
• /
• pp.267-270
• /
• 2000

In this paper, we consider the simple redundant structures with the function of hardware based active/standby control. The system includes two switch modules. The switch module is connected to a data bus, but only the active switch module has control of the data bus. The standby unit takes over the function of the active unit when the active unit failure or mode command are asserted. And this paper illustrate the high-speed router system and the overall redundant system architecture. The proposed redundant architecture for 80G Router system is verified and implemented with experiment.

• The KIPS Transactions:PartA
• /
• v.16A no.4
• /
• pp.235-244
• /
• 2009

The mission-critical computer for air defense has to maintain its operation without any fault for a long mission time and is required to implement at low cost. Now the reliability of the mission critical-computer using Active Sparing Redundancy fault-tolerant technique is inferior to that of the computer using TMR technique. So in this paper are proposed Extended ASR(EASR) technique that provides higher reliability than that of the computer using TMR technique. The fault-tolerant performance of the implemented mission-critical computer is proven through reliability analysis and numbers of fault recovery test. Also, the reliability of the mission-critical computer using EASR technique is compared with those of computer using ASR and TMR techniques. EASR technique is very suitable to the mission-critical computer.

• Journal of the Korean Institute of Telematics and Electronics D
• /
• v.35D no.6
• /
• pp.46-53
• /
• 1998

A new method to form the double structured active layers of a-Si/a-SiN$_{x}$ of polycrystalline thin film transistor is proposed and poly-Si TFTs employed double structure active film are fabricated. Nitrogen ions were added to bottom amorphous silicon active film(a-SiN$_{x}$ ) and pure a-Si film deposition on a-SiN$_{x}$ was followed. The XeCl excimer laser was irradiated to crystallize double structure active film. The grain growth of upper a-Si film was also promoted in the double structured active layers of a-Si/a-SiN$_{x}$ due to the mitigation of solidification process of lower a-SiN$_{x}$ layer. Our experimental results show that the ratio of NH$_3$/SiH$_4$ is required to maintain below 0.11 for the reduction of contact resistance of n$^{+}$ poly-SiN$_{x}$ layer.r.

• Convergence Security Journal
• /
• v.8 no.4
• /
• pp.51-56
• /
• 2008

A combination of VPNs and dial-up access, such as DSL and Cable, usually provides the cost-effective solution as the substitution of private networks on high-cost leased line. The business demand for high availability has increased with VPN spreading. This paper presents the schemes for a high availability of network access and a load balancing of network traffic in VPN gateways by using multiple links or multihoming capability based on active-active approach. The high availability and load balancing of network links can be achieved by duplicating external network access into multiple independent links. This can provide a continuous network connection to internal users even if one of the links is failed. Moreover, it can provide twice network bandwidth by distributing the traffic into the links. Static and dynamic algorithms are proposed as the load balancing algorithms.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.27 no.1C
• /
• pp.103-111
• /
• 2002

In RNC of IMT-2000, main control processors such as ASP, ACP and OMP are responsible for call control function, and the high reliability and real-time property should be provided for it. So, the study of real-time fault-tolerant for it is needed. In this paper, we proposes an Task based duplication method, in which Tasks in active side operated on message unit and send the updated data to standby side after operation, log in the message to standby side for recovery during take-over. This scheme decreases the dual down and the complexity of synchronization procedure, and performs the synchronization more exactly because Tasks control the synchronization of system. This paper also proposes the fault detection and the fault handing method for effective implementation of Task based duplication. This scheme focus on increasing the fault detection rate and intercepting originally that fault data is send to standby side.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.18 no.11
• /
• pp.2634-2639
• /
• 2014

The idea of redundancy is used in order to improve the availability of networks and systems and there are various methods for implementing redundancy. To perform the availability analysis various stochastic models have been used. In this paper, 2N redundancy with one active service unit and one standby service unit is considered. To evaluate the expected availability, we model 2N redundancy using Stochastic Reward Nets. This model can be solved using the SPNP package.