{"title":"A fault secure dictionary machine","authors":"A. Reddy, P. Banerjee","doi":"10.1109/ICDE.1987.7272363","DOIUrl":null,"url":null,"abstract":"A fault-secure dictionary machine is presented in this paper. The symmetry of the binary tree architecture for a dictionary machine is exploited to obtain fault-secureness with little overhead. The proposed design utilizes only one extra processor and with some other modifications to the structure of the processors, can detect a single failure of a processor or a link. The proposed design keeps two copies of each record and whenever a record is extracted from the machine the two copies are compared to detect if any fault has occurred. The low overhead is a result of observing the fact that at any given time all the processors of the machine need not be active and these potentially-idle processors are used to do redundant processing to enable detecting single faults.","PeriodicalId":145433,"journal":{"name":"1987 IEEE Third International Conference on Data Engineering","volume":"8 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1987-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"3","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"1987 IEEE Third International Conference on Data Engineering","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICDE.1987.7272363","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 3
Abstract
A fault-secure dictionary machine is presented in this paper. The symmetry of the binary tree architecture for a dictionary machine is exploited to obtain fault-secureness with little overhead. The proposed design utilizes only one extra processor and with some other modifications to the structure of the processors, can detect a single failure of a processor or a link. The proposed design keeps two copies of each record and whenever a record is extracted from the machine the two copies are compared to detect if any fault has occurred. The low overhead is a result of observing the fact that at any given time all the processors of the machine need not be active and these potentially-idle processors are used to do redundant processing to enable detecting single faults.