• KIISE Transactions on Computing Practices
• /
• v.22 no.1
• /
• pp.26-31
• /
• 2016

Migrim(Migration enhanced Grid Middleware) is a communication middleware between embedded devices and multiple servers. In traditional client-server communication, users' requests are sent to and processed by a designated server even though the server may suffer from a heavy load. In addition, the designated server may not have proper information to process the user's request correctly. Proposed connection migration mechanism and transaction migration mechanism are designed to improve the performance and accuracy of request processing. The connection migration is a procedure for delegating a connection to another server, which results in a well-distributed balancing of load among the servers. The transaction migration is a procedure for delegating a transaction to another server, and improves the accuracy of response.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.6
• /
• pp.83-90
• /
• 2009

Since traffic is aggregated to a MPP that acts as an Internet gateway, if traffic load is not balanced among the MPPs in a WMN, the overall performance of a WMN becomes poor even though the total traffic load is far below the capacity of the WMN. Therefore, in this paper, we propose a stochastic load balancing scheme where each MP (Mesh Point) probabilistically selects its serving MPP according to the congestion levels of MPPs. Through extensive simulations using ns-2, we have verified that our scheme can stabilize a WMN fast when congestion occurs and reduce packet loss rate by distributing traffic load of a congested MPP to multiple MPPs in the inverse proportional to their congestion levels. Compared to queue-based load balancing scheme, our method can decrease network stabilization time by 34 seconds, and reduce packet loss rate by 7.6%. Since the proposed scheme can reduce network stabilization time by efficiently using network resource, it is expected to contribute to the reliable operation of a WMN.

• The Journal of the Korea Contents Association
• /
• v.16 no.7
• /
• pp.667-680
• /
• 2016

As distributed computing platforms like Hadoop MapReduce have been developed, it is necessary to perform the conventional query processing techniques, which have been executed in a single computing machine, in distributed computing environments efficiently. Especially, studies on similarity join query processing in distributed computing environments have been done where similarity join means retrieving all data pairs with high similarity between given two data sets. But the existing similarity join query processing schemes for distributed computing environments have a problem of skewed computing load balance between clusters because they consider only the data transmission cost. In this paper, we propose Matrix-based Load-balancing Algorithm for efficient similarity join query processing in distributed computing environment. In order to uniform load balancing of clusters, the proposed algorithm estimates expected computing cost by using matrix and generates partitions based on the estimated cost. In addition, it can reduce computing loads by filtering out data which are not used in query processing in clusters. Finally, it is shown from our performance evaluation that the proposed algorithm is better on query processing performance than the existing one.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.10c
• /
• pp.53-55
• /
• 2000

인터넷의 급성장은 웹서버의 요청율을 급격하게 증가시키고 심각한 정체를 일으키고 있다. 이를 해결하기 위하여 클라이언트의 요청을 분산된 여러 서버들에게로 전달시키는 방법이 많이 사용되고 있다. 이 방법은 웹서버의 처리량을 향상시키고 확장성과 가용성을 제공하지만 서버들의 부하를 균등하게 하는 방법이 필요하게 된다. 이러한 부하 균등을 위해 크게 DNS(Domain Name Server)를 활용한 정책과 IP 단계에서 분배기(Dispatcher)를 사용한 정책이 제안되었다. 본 논문에서는 전자와 후자의 단점을 보완하고 성능을 향상시킨 다중 분배기를 제안하고 실험하였다.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.22 no.12
• /
• pp.2635-2643
• /
• 1997

In this paepr, a routing algorithm is proposed in order to reduce average end-to-end delay and congesting in the high speed network. The algorithm proposed here uses the existing one which adaptively modifies routes and the amount of traffic allocated to each link as user traffic partterns flutuate. This algorithm is ELB(Equitable Load Balancing). Also, the new algorithm considers the proportional to the distance between source and destination. It reduces congestion from the ELB and average end-to-end delay from the propagation dealy concepts respectively. Through the simulation, it shows that the algorithm proposed here reduces average end-to-end delay over low load to high load, and it also guarantees the congestion control.

• Proceedings of the Korean Information Science Society Conference
• /
• 2008.06b
• /
• pp.346-350
• /
• 2008

AMD에서 사용한 HyperTransport 기술 기반 다중 처리기가 좋은 성능을 보이면서 최근 NUMA(Non Uniform Memory Access) 환경에 대한 관심이 증가하고 있다. 본 논문에서는 NUMA 시스템을 위한 부하균등 모델을 제안한다. 다중 처리기 시스템에서 운영체제는 특정 처리기에 부하가 많아지는 것을 부하가적은 처리기로 나누어 주기 위해 부하 균등 기법들을 가지고 있다. 이런 부하 균등 기법은 처리기가 가지고 있는 태스크 개수에 의존적인 연구가 많다. 본 연구에서는 NUMA 시스템의 메모리 접근 비용이 위치에 따라 다른 것을 반영한 부하 균등 기법의 모델을 제시한다. 이를 위해 모의 실험 환경을 구축하고 특정 상황들에 대한 실험을 통해 증명한다.

• Journal of Korea Multimedia Society
• /
• v.11 no.3
• /
• pp.331-340
• /
• 2008

In order to make load balancing, workload information of nodes should be informed to the network. In a load balancing algorithm[13] based on the SBIBD(Symmetric Balanced Incomplete Block Design), each node receives global workload information by only two round message exchange with O (${\upsilon}{\sqrt{\upsilon}}$) traffic overhead, where ${\upsilon}$ is the number of nodes. It is very efficient but works well only when ${\upsilon}=p^2+p+1$ for a prime number p. In this paper, we generate a special incidence structure in order for the algorithm works well for an arbitrary number of nodes. In the experiment with w,($5{\leq}w{\leq}5,000$), nodes and more than 80% of receiving workload information, traffic overhead was less than O ($w{\sqrt{w}})$ and the result for standard deviation of traffic overhead showed that each node has largely balanced amount of traffic overhead.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.04a
• /
• pp.151-153
• /
• 2000

본 논문은 분산 서버 시스템을 위한 Load Balancing Architecture를 제안한다. 이 구조는 기존에 구현된 클라이언트/서버 시스템을 그대로 사용하면서 규모를 확장할 수 있는 특징을 가지고 있다. Load Balancing Service가 동작하기 위해서는 서버의 성능과 구성상태에 대한 정보를 주고받는 것이 필요하다. 이를 위하여 본 논문에서는 Load Balacing Information Transfer Protocol을 제시한다. 본 논문이 제시하는 구조는 DCOM 및 EJB 등의 Component 서버에서도 사용되어서 확장성을 가지게 할 수 있다.

• Journal of Korea Multimedia Society
• /
• v.7 no.11
• /
• pp.1580-1587
• /
• 2004

Nowadays, VPN(Virtual Private Network) devices supporting multi-link connections only utilize the second link to backup the fail-off primary link. In practice, however, the occurrence of the link fail-off is so rare that the capacity available on the second link is wasted. In this paper, a scheme of establishing a VPN with multi-links, and load balancing between the links under normal circumstances is proposed in order to take advantage of multi-links, and to eventually increase the effective bandwidth of the VPN. Additionally, the transition functionality is also applied for the link fail-off case like existing VPN devices. Consequently, the proposed scheme enables VPNs with multi-links not only to maintain the higher availability but also to highly increase the effective bandwidth.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.26 no.11C
• /
• pp.140-147
• /
• 2001

In this paper, we propose a scheme of calculating resources\` costs for precomputation in QoS routing schemes. This scheme is effective in respect to the g1oba1 network utilization and the balanced use of network resources. We also propose a QoS routing scheme for transit and intra traffic in a large scale of domain-based network. For a domain in the network, the routing scheme first precomputes multiple paths between all pairs of ingress and egress border routers while considering balancing of the expected load and then selects paths with the probability of path selection. We, therefore, expect that the paths are better than any other paths in respect to reserving the network resources on paths. The routing scheme combines inter and intra domain routings seamlessly and uses the same cost calculation scheme. And for analysis of a load balancing cost calculation proposed a simulation is done.