Earticle

최근 인터넷 뱅킹, 전자결제, 전자상거래, 홈쇼핑, 증권 거래, 민원 업무 등 미션 크리티컬한 비즈니스가 증가함에 따라 무정지 서비스에 대한 요구가 높아지면서, 기존의 단일 데이터베이스에서 클러스터링, 이중화 기법 등에 관한 고가용성 기법에 관한 연구가 증가하고 있다. ALTIBASETM Log Analyzer(이하 ALA)는 이중화 기법과는 별개로 Active Log를 기반으로 API를 제공하여, 이기종 또는 동일 기종간의 통신 및 확장성을 제공한다. 본 논문에서는 ALA를 사용하여 고가용성과 고확장성, 실시간 동기화 기능을 모두 만족시킬 수 있는 데이터베이스 시스템의 설계를 제시하고 ALA에 대한 성능평가를 하였다.

Recently, the need for non-stop service is increasing by the business mission-critical Internet banking, e-payment, e-commerce, home shopping, securities trading, and petition business increases, clustered in a single database of existing, redundant research on high-availability technologies related to technique, etc. is increasing. It provides an API based on the Active Log in addition to the technique of redundancy, ALTIBASETM Log Analyzer (below, ALA), provides scalability and communication of the same model or between heterogeneous. In this paper, we evaluated the performance of ALA by presenting the design of the database system that you can use the ALA, to satisfy all the synchronization features high scalability and high availability, real-time.

최근 SNS의 발전과 더불어 급격히 증가하고 있는 대용량 데이터들을 효과적으로 활용하기 위해서는, 대용량 데이터에 대한 효율적이고 유연한 분산 처리 기술이 필수적이다. 그러나 기존의 관계형 데이터베이스는 데이터의 폭증 상황에 대한 QoS를 보장하기 어려운 단점이 존재한다. 따라서 본 논문에서는 SNS에서와 같은 대용량 데이터 서비스 지원 시 발생하는 기존 RDBMS의 문제점을 제시한다. 또한, 이러한 문제점을 해결하기 위해 Cassandra 기반 분산 처리 데이터베이스 시스템을 설계 및 구현한다. 아울러, 구현한 시스템을 SNS 프로토타입에 적용하여 HBase와의 성능비교를 수행한다. 마지막으로, 성능결과를 분석함으로써 구현한 시스템이 데이터 폭증 상황에 유연한 플랫폼임을 검증한다.

Recently, in order to effectively utilize the big data which is rapidly increasing with the growth of SNS, efficient and flexible distributed processing of big data is essential. However, the conventional DBMS cannot guarantee Quality of Service(QoS) for data explosion. Therefore, in this paper, we present the problem of the existing RDMS for supporting the big data management service. Also, in order to solve the problem, we design and implement the Cassandra based distributed database system for managing the big data. Furthermore, we do the performance comparison of the Cassandra with HBase by applying the developed system to the SNS prototype. Finally, we verify that our system is more flexible and efficient platform by analyzing its performance result.

A lot of learning objects have been distributed to schools, but they are not fully utilized because many teachers do not know where the data are and how to use them. Furthermore, the internet search engines produce large amount of data most of which are of no use. Thus, it is needed to develop an efficient method of finding data necessary for instruction and learning. In this study, a learning objects management system is constructed that utilizes these unused servers by connecting distributed databases using JDBC, which is a technology for distributed database. In addition, we developed an learning objects sharing function based on a metadata standard that we revised using the data collected from Sharable Content Object Reference Model(SCORM), a metadata standard from Korea Education & Research Information Service(KERIS).

데이터베이스는 정보 시스템의 핵심 기술로서 많은 양의 데이터를 효율적으로 구성하고 저장하는 역할을 한다. 본고에서는 먼저 데이터베이스 기술을 위한 분산 시스템의 데이터 일관성 및 보안 요구 사항을 분 석하고자 한다. 데이터 공유 및 리소스 공유에 있어 NewSQL 데이터베이스의 장점을 소개하고 분산 시 스템 설계를 기반으로 하는 NewSQL 데이터베이스의 핵심 기술을 분석한다. 분산 스토리지 엔진, 실시간 분석 및 조회 엔진, HTAP 하이브리드 워크로드 아키텍처, 분산 트랜잭션 등은 예로 들 수가 있다.

Synchronization of clocks in distributed system has been an important area of research over the last decade. There are various methods of achieving clock synchronization depending on the requirements of the situation. A clock synchronization service ensures that spatially dispersed and heterogeneous processors in a distributed system share a common notion of time. In order to behave as a single, unified computing resource, distributed systems have need for a synchronization of drifting clocks and several algorithms have been proposed on this topic. Our paper highlights total network connection between different processors (alias nodes) of the system, We provide a simple, efficient, and unified algorithm for clock synchronization that can minimize the total network connection of the system.

An efficient functionality of any distributed database system “DDBS” is highly dependent on its proper design in terms of the adopted fragmentation and allocation methods. However, an optimal fragmentation and allocation design techniques could be very complex and requires good experiences and knowledge to reach. Fragmentation of a large, global databases are performed by dividing the database relations horizontally, vertically or as a combination of both. In order to enable distributed database systems to work efficiently, these fragments have to be allocated across the available sites in such a way that reduces communication cost i.e. to minimize the total volume of data transmitted during queries execution over sites. This paper presents a new data re-allocation model for replicated and non-replicated constrained DDBSs by bringing a change to data access pattern. This approach assumes that the distribution of fragments over network sites was initially performed according to a properly forecasted set of query frequency values that could be employed over sites. Our model takes sites constraints into account in the re-allocation phase. It proposes an efficient plan to re-allocate data fragments across sites based on communication and update cost values for each fragment individually. The re-allocation process will be performed by selecting the maximal update cost value for each fragment and making the re-allocation accordingly. Experimental results confirmed that the proposed technique will effectively contribute in solving fragments re-allocation problem in a dynamic distributed relational databases environment.

In this paper, it is to focus on two areas, primarily data inconsistency problems faced by the mobile database users when they are trying to accessing or updating the database data while mobility in work, in such environment demand for appropriate strategies arises, to overcome this problem. When users come in contact with other network node, updated data is not reflected or propagated at the other sites or nodes, which is previously updated or changed at previous visited site or at activity center. Secondly, require advance replication architecture and effective network protocol based algorithms follows effective replication strategy that securely supports databases changes activities anytime and anywhere in mobile computing network environment.

The performance of a distributed database system depends particularly on the site-allocation of the fragments. Queries access different fragments among the sites, and an originating site exists for each query. A data allocation algorithm should distribute the fragments to minimize the transfer and settlement costs of executing the query plans. The primary cost for a data allocation algorithm is the cost of the data transmission across the network. The data allocation problem in a distributed database is NP-complete, and scalable evolutionary algorithms were developed to minimize the execution costs of the query plans. In this paper, quadratic assignment problem heuristics were designed and implemented for the data allocation problem. The proposed algorithms find near-optimal solutions for the data allocation problem. In addition to the fast ant colony, robust tabu search, and genetic algorithm solutions to this problem, we propose a fast and scalable hybrid genetic multi-start tabu search algorithm that outperforms the other well-known heuristics in terms of execution time and solution quality.

A replication server is considered to be one of the most effective tools to cope with the problems that may be caused by the complex data replications in distributed database systems. In the distributed environment, locking a table is inevitable and it is the main reason to coerce the system practically. This paper presents an Asynchronous Replicator Scheme (ARS) that basically utilizes the system log as files named differential files to refresh the distributed data files with complicated queries, and that it prevents (normally, huge) base tables from being locked. We take join operations as the complicated queries, not only because the join operation covers almost all the operations, but also because it is one of the most time-consuming and data intensive operations in query processings.

Although numerous distributed database design models and solution algorithms have been developed, very few have been validated. Validation is critical to the successful application of such models to distributed database design. In this paper, we develop a simulation system for distributed database design. We then analyze and validate an average response time model using simulation. The simulation results demonstrate that the average response time model is reasonably accurate.

데이터베이스 기술의 적용분야가 점차 확대되어감에 따라 작업처리율을 증대시키기 위한 다양한 형태의 거래 처리 모형들을 필요로 하는 추세이다. 그러나 기존의 syntax위주의 직렬성 이론만 가지고서는 거래의 실행시간상 차별화 특성을 수용하면서 다수의 거래에 대한 단위시간당 처리 생산성을 높이기는 힘든 형편이다. 이러한 상황을 해결하기 위하여 이타적 잠금기법(altruistic locking: AL)은 거래가 객체를 사용한 다음 더이상 그 객체를 요구하지 않을 때 다른 거래들이 그 객체를 로크할 수 있도록 미리 객체에 대한 로크를 해제함으로써 거래들의 대기시간을 줄이기 위한 취지에서 제안된 것이다. 확장형 이타적 잠금(extended altruistic locking: XAL)기법은 AL을 자취의 확장 측면에서 개선한 잠금기법으로서 AL이 근본적으로 안고 있는 반드시 기부된 객체만을 처리해야 한다는 부담을 보다 완화한 기법이다. 본 논문에서는 우선 장기거래로 인한 단기거래의 장기적 대기현상 완화 측면에서의 AL과 XAL의 공통적 한계점을 분석하였다. 분산 환경하에서 장기거래로 인한 단기거래의 장기적 대기현상을 최소화하도록 줄임으로써 동시성 제어의 정도를 높이는 반면, 거래간의 평균 대기시간을 줄일 수 있는 새로운 확장형 이타적 잠금기법인 전후진방식의 신형 확장 기법인 2DL(two-way donation locking)을 제안하였다. 기법의 적용 광범위성을 위해 분산 계산 환경에서도 작동될 수 있게끔 설계하였다. 모의실험에 의한 성능평가 결과 장기거래의 길이가 5이상, 9이하인 상황에서 2DL은 2PL보다 작업 처리율과 거래의 평균 대기시간 면에서 우수한 결과를 나타내었다.

Database correctness is guaranteed by standard transaction scheduling schemes like two-phase locking for the context of concurrent execution environment in which short-lived ones are normally mixed with long-lived ones. Traditional syntax-oriented serializability notions are considered to be not enough to handle in particular various types of transaction in terms of duration of execution. To deal with this situation, altruistic locking has attempted to reduce delay effect associated with lock release moment by use of the idea of donation. An improved form of altruism has also been deployed in extended altruistic locking in a way that scope of data to be early released is enlarged to include even data initially not intended to be donated. In this paper, we first of all investigated limitations inherent in both altruistic schemes from the perspective of alleviating starvation occasions for transactions in particular of short-lived nature. The idea of two-way donation locking(2DL) has then been experimented to see the effect of more than single donation in distributed database systems. Simulation experiments shows that 2DL outperforms the conventional two-phase locking in terms of the degree of concurrency and average transaction waiting time under the circumstances that the size of long-transaction is in between 5 and 9.

질의실행시간최소화는 분산 데이타베이스 설계에 있어 가장 중요한 목적중의 하나이다. 총시간최소화는 온라인거래처리시스템의 목적인 반면, 반응시간최소화는 의사결정지원 질의시스템의 목적이다. 본 논문에서는 질의실행시간최소화를 달성하기 위해 질의를 세분화하여 최적의 데이터베이스 사이트에 할당하는 분석모델을 개발하였으며, 문제해결방법으로 유전자알고리즘을 채택하였다. 총시간최소화 관점에서의 질의실행 계획은 반응시간최소화 관점의 질의실행계획에는 적합하지 않다는 것을 증명하였으며, 그 반대의 경우도 증명하였다. 최대 20개의 조인이 포함되는 질의를 설계하여 시뮬레이션 실험을 통해 테스트를 수행하였고, 유전자알고리즘과 완전한 전수조사와의 결과를 비교함으로써 모든 경우에 유전자알고리즘을 채택한 해결책이 최적의 결과를 도출하였음을 증명하였다.

Query execution time minimization is an important objective in distributed database design. While total time minimization is an objective for On Line Transaction Processing (OLTP), response time minimization is for Decision Support queries. We formulate the sub-query allocation problem using analytical models and solve with genetic algorithm (GA). We show that query execution plans with total time minimization objective are inefficient from response time perspective and vice versa. The procedure is tested with simulation experiments for queries of up to 20 joins. Comparison with exhaustive enumeration indicates that GA produced optimal solutions in all cases in much less time.

This research presents distributed database and knowledge-base modeling approach for intelligent product design system. The product design information in this study is described by a collection of rules and design knowledge that are utilized according to the product development procedures. In this work, a network-based architecture has been developed to enable dispersed designers to simultaneously accomplish remote design tasks. A client/server communication diagram has also been proposed to facilitate consistent primary information modeling for multi-user access and reuse of designed results. An intelligent product design system has been studied with the concepts of distributed database and network-based architecture in order to support concurrent engineering design and automatic design part assembly. The system provides the capability of composing new designs from proper design elements stored in the database and knowledge-base. The distributed intelligent product design is applied to the design of an automobile part as an example.