• Proceedings of the Korean Information Science Society Conference
• /
• 2000.04a
• /
• pp.139-141
• /
• 2000

검사점 및 복구 도구는 응용 프로그램의 상태를 주기적으로 안정된 저장소에 저장을 하고, 결함이 발생하였을 경우 가장 최근의 검사점으로부터 효율적으로 복구하게 하는 도구이다. 특히 검사점 및 복구 도구는 장시간 수행되는 프로세서에게는 아주 중요한 의미를 지니며, 결함으로 인해 장시간 수행되는 프로세스에 의해 생성된 중간 결과를 잃어버리지 않게 한다. 본 논문에서는 UnixWare 커널 수준의 검사점 및 복구 도구인 Kckpt의 설계 및 구현 내용을 제시하고, Kckpt의 성능을 사용자 수준에서 구현한 검사점 라이브러리와 비교한 결과를 제시한다. Kckpt를 사용함으로 해서 UnixWare는 소스 코드 수준에서 사용자가 초기화한 장소에서 검사점을 만들 수 있을 뿐만 아니라 실행 가능한 파일만을 가진 사용자에게도 완전한 투명성을 제공할 수 있다. 또한 Kckpt를 사용자 라이브러리 수준에서 구현한 검사점 도구와 성능을 비교한 결과 오버헤드가 훨씬 줄어들었음을 관찰할 수 있었다.

• Development and Reproduction
• /
• v.18 no.1
• /
• pp.65-71
• /
• 2014

Cell cycle process is regulated by a number of protein kinases and among them, serine/threonine kinases carry phosphate group from ATP to substrates. The most important three kinase families are Cyclin-dependent kinase (Cdk), Polo-like kinase (Plk), and Aurora kinase. Polo-like kinase family consists of 5 members (Plk1-Plk5) and they are involved in multiple functions in eukaryotic cell division. It regulates a variety of aspects such as, centrosome maturation, checkpoint recovery, spindle assembly, cytokinesis, apoptosis and many other features. Recently, it has been reported that Plks are related to tumor development and over-expressed in many kinds of tumor cells. When injected the anti-Plk antibody into human cells, the cells show aneuploidy, and if inhibit Plks, most of the mitotic cell division does not proceed properly. For that reasons, many inhibitors of Plk have been recently emerged as new target for remedy of the cancer therapeutic research. In this paper, we reviewed briefly the characteristics of Plk families and how Plks work in regulating cell cycles and cancer formation, and the possibilities of Plks as target for cancer therapy.

• Journal of KIISE:Computer Systems and Theory
• /
• v.29 no.2
• /
• pp.75-86
• /
• 2002

The selection of the optimal checkpointing interval has been a very critical issue to implement a checkpointing recovery scheme for the fault tolerant distributed system. This paper presents a new scheme that allows a process to select the proper checkpointing interval dynamically. A process in the system evaluates the cost of checkpointing and possible rollback for each checkpointing interval and selects the proper time interval for the next checkpointing Unlike the other scheme, the overhead incurred by both of the checkpointing and rollback activities are considered for the cost evaluation and current communication pattern is reflected in the selection of the checkpointing interval. Moreover, the proposed scheme requires no extra message communication for the checkpointing interval selection and can easily be incorporated into the existing checkpointing coordination schemes.

• The Transactions of the Korea Information Processing Society
• /
• v.7 no.6
• /
• pp.1753-1765
• /
• 2000

As the areas of computer application are expanded, real-time application environments that must process as many transactions as possible within their deadlines, such as a stock transaction systems, ATM switching systems etc, have been increased recently. The reason why the conventional database systems can't process soft real-time applications is the lack of prediction and poor performance on processing transaction's deadline. If transactions want to access data stored at the secondary storage, they can not satisfy requirements of real-time applications because of the disk delay time. This paper designs a main-memory database prototype systems to be suitable to real-time applications and then this system can produce rapid results without disk i/o as all of the information are loaded in main memory database. In thesis proposed the improved techniques with respect to logging, checkpointing, and recovering in our environment. In order to improve the performance of the system, a) the frequency of log analysis and redo processing is reduced by the proposed redo technique at system failure, b) database consistency is maintained by improved fuzzy checkpointing. The performance model is proposed which consists of two parts. The first part evaluates log processing time for recovery and compares with other research activities. The second part examines checkpointing behavior.

• Journal of KIISE:Software and Applications
• /
• v.32 no.11
• /
• pp.1084-1098
• /
• 2005

In this paper, we propose a migration technique for autonomous mobile agents suitable to dynamic environments. The proposed migration technique dynamically creates the itinerary of agents that considers states of networks and systems. In the migration of the agent, it first sends prefetching message to the next system. The system receives necessary data for migration in advance. Through this, we reduce the amount of the sending data and save the time for creating the instance of the agent. And it improves the execution efficiency by using the checkpoint-based recovery method that does not execute the agent again and recovers the process states even though the errors take place. To show superiority of the proposed technique, we compare the proposed method with the existing methods through various simulations.

• Proceedings of the Korean Information Science Society Conference
• /
• 2011.06c
• /
• pp.88-91
• /
• 2011

NAND 플래시 메모리는 저전력과 빠른 접근 속도의 특징 때문에 차세대 저장장치로 주목 받고 있다. 특히 플래시 메모리로 만들어진 SSD(solid state disk)는 인터페이스가 기존의 하드디스크와 동일하고 대용량화 되고 있기 때문에 가까운 미래에 다양한 저장시스템의 저장장치로 사용될 것으로 예상된다. 그러나 NAND 플래시메모리 기반 저장장치는 쓰기 전 소거 구조와 같은 독특한 하드웨어 특징을 가지고 있기 때문에 특정 지역에 반복적인 쓰기 요청을 발생하는 B트리를 구축하는 것은 심각한 성능저하를 야기 할 것이다. 이러한 문제를 해결하기 위해 버퍼를 이용하여 B트리 구축 성능을 개선한 방법들이 제안되었다. 그러나 이러한 기법들은 갑작스러운 전원 차단 시 버퍼에 유지하고 있던 데이터를 모두 유실하기 때문에 고장회복을 위한 추가적인 방법이 필요하다. 따라서 본 논문에서는 버퍼를 이용한 방법 중 IBSF기법을 기반으로 NAND 플래시 메모리 기반 저장장치에서 고성능의B트리 구축 방법뿐만 아니라 전원 차단시 효율적인 고장회복을 할 수 있는 기법을 제안한다. 본 논문에서 제안하는 기법은 B트리 변경시 변경 된 정보를 로그에 저장하여 관리한다. 또한 루트노드가 변경될 때 검사점(checkpoint)을 수행한다. 마지막으로 다양한 실험을 통하여 본 논문의 고장회복 성능을 보여준다.

• Journal of KIISE:Software and Applications
• /
• v.26 no.5
• /
• pp.621-638
• /
• 1999

재시작 회복시 수행해야 할 작업략을 줄이기 위하여 주기적으로 수행되는 검사점 중 퍼지 검사점은 정상적인 트랜잭션들의 수행과 병행하여 수행된다. 트랜잭션들과 퍼지 검사점의 수행시, 로그 레코드 기록의 순서화를 위한 로그 래취를 획득한 상태에서 로그 레코드를 기록할수있으며, 트랜잭션 엔트리의 무결성을 보장하기 위한 트랜잭션 래취를 획득한 상태에서 트랜잭션은 로그 레코드의 기록과 관련된 트랜잭션 상태를 변경할수 있으며 검사점 프로세스는 트랜잭션의 상태를 검사할 수 있다. 퍼지 검사점과 트랜잭션들이 수행이 비동기로 수행되는 상황에서 로그래취와 트랜잭션 래취의 획득 및 해제 방법이 적절하지 못한 경우, 퍼지 검사점이 어떤 트랜잭션의 트랜잭션 상태를 검사하기 전에 그 트랜잭션에 의하여 기록된 로그 레코드이지만 그 퍼지 검사점이 그 트랜잭션 상태를 검사할때까지 그 로그 레코드의 기록상황이 그 트랜잭션의 트랜잭션 상태에 반영되지 않은 경우가 발생할수 있다. 이 경우, 기록 상황이 트랜잭션 상태에 반영되지 않은 그 로그 레코드를 그 퍼지 검사점에 대한 그 트랜잭션의 상태미방영 로그 레코드라고 한다. 본 논문은 상태미반영 로그 레코드를 발생시키는 상황들을 분석하고, 상태미반영 로그 레코드의 발생에 따른 문제점들을 기술하며, 상태미반영 로그 레코드의 발생에 대하여 전면 방지 기법, 일부 방지 기법 그리고 전면 허용하지만 회복시 이를 해결하는 기법등 세가지 해결책을 제시한다.

• Journal of KIISE:Computer Systems and Theory
• /
• v.31 no.3_4
• /
• pp.204-212
• /
• 2004

Existing distributed systems have higher probability of failures occurrence than stand-alone system, so many fault tolerant techniques have been developed. Because of insufficient storage resulting from the increased fault tolerance information stored, the performance of system has been degraded. To avoid performance degradation, it needs delete useless fault tolerance information. In this paper, we propose a garbage collection algorithm for fault tolerance information. And we define and design the garbage collection agent for garbage collection of fault tolerance information, the information agent for management of fault tolerant data, and the facilitator agent for communication between agents. Also, we propose the garbage collection algorithm using the garbage collection agent. For rollback recovery, we use independent checkpointing protocol and sender based pessimistic message logging protocol. In our proposed garbage collection algorithm, the garbage collection, information, and facilitator agent is created with process, and the information agent constructs domain knowledge with its checkpoints and non-determistic events. And the garbage collection agent decides garbage collection time, and it deletes useless fault tolerance information in cooperation with the information and facilitator agent. For propriety of proposed garbage collection technique using agents, we compare domain knowledge of system that performs garbage collection after rollback recovery and domain knowledge of system that doesn't perform garbage collection.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.8 no.1
• /
• pp.137-149
• /
• 2004

A new software paradigm is required on the development of network and various service requirements. With this, many studies on a mobile agent have been made. For the execution of the mobile agent, migration is the most important factor that influences the performance of the mobile agent. In this paper we propose the method that leads to high migration efficiency in order to improve the performance. The features of our migration technique are as follows. First, the migration technique creates the dynamic itinerary that appropriately copes with the network conditions and the platform changes to improve the agent execution efficiency. Second, it perfecters an executed code to reduce the amount of the mobile data and reduces the execution time by instantiating the agent in advance. Third, it improves the execution efficiency by using the checkpoint-based recovery method that does not execute the agent again and recovers the process states even though the errors take place. Though the simulation we compare the proposed method with the existing methods. The simulation result shows that the proposed method outperform the existing methods in terms of migration.

• The KIPS Transactions:PartA
• /
• v.10A no.2
• /
• pp.123-130
• /
• 2003

This paper presents a new striped checkpointing scheme for serverless cluster computers, where the local disks are attached to the cluster nodes collectively form a distributed RAID with a single I/O space. Striping enables parallel I/O on the distributed disks and staggering avoids network bottleneck in the distributed RAID. We demonstrate how to reduce the checkpointing overhead and increase the availability by striping and staggering dynamically for communication intensive applications. Linpack HPC Benchamark and MPI programs are applied to these checkpointing schemes for performance evaluation on the 16-nodes cluster system. Benchmark results prove the benefits of the striped checkpointing scheme compare to the existing schemes, and these results are useful to design the efficient checkpointing scheme for fast rollback recovery from any single node failure in a cluster system.