{"title":"A fault-tolerant bridging scheme for dynamic load balancing","authors":"T. Sheu","doi":"10.1109/LCN.1991.208061","DOIUrl":null,"url":null,"abstract":"Presents a fault-tolerant bridging scheme for interconnecting multiple LANs. Fault tolerance is built by linking a LAN to two bridge ports located on two different processing elements, respectively. Two load-partitioning schemes are developed to prevent multiple copies of an arriving frame. A static partitioning scheme requires little hardware or software augmentation. However, the load may not be balanced due to non-uniform traffic patterns. On the other hand, a dynamic partitioning scheme can offer dynamic load balancing, although it requires mechanisms to monitor the traffic and to synchronize inter-processor communication. The bridge performance in terms of frame response time is evaluated through simulation. In the simulation, the impact of the two partitioning schemes on bridge performance are compared. Finally interprocessor synchronization cost is analyzed.<<ETX>>","PeriodicalId":163373,"journal":{"name":"[1991] Proceedings 16th Conference on Local Computer Networks","volume":"9 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1991-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"[1991] Proceedings 16th Conference on Local Computer Networks","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/LCN.1991.208061","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Presents a fault-tolerant bridging scheme for interconnecting multiple LANs. Fault tolerance is built by linking a LAN to two bridge ports located on two different processing elements, respectively. Two load-partitioning schemes are developed to prevent multiple copies of an arriving frame. A static partitioning scheme requires little hardware or software augmentation. However, the load may not be balanced due to non-uniform traffic patterns. On the other hand, a dynamic partitioning scheme can offer dynamic load balancing, although it requires mechanisms to monitor the traffic and to synchronize inter-processor communication. The bridge performance in terms of frame response time is evaluated through simulation. In the simulation, the impact of the two partitioning schemes on bridge performance are compared. Finally interprocessor synchronization cost is analyzed.<>