本文已被:浏览 1161次 下载 2287次
Received:October 22, 2015 Revised:November 25, 2015
Received:October 22, 2015 Revised:November 25, 2015
中文摘要: RPKI(Resource Public Key Infrastructure,互联网码号资源公钥证书体系)中的签名对象由RP(Relying Party,依赖方)端同步下载后处理成IP地址块与AS(Autonomous System,自治域)号的真实授权关系,用于指导BGP路由.当前的RP使用软件rsync(Remote Sync)来同步,而rsync的同步算法并未考虑RPKI中文件(目录)的特点,导致同步效率并不理想.通过分析并结合RPKI中文件(目录)的特点,设计并实现了一种基于有序哈希树的RPKI资料库同步工具htsync.实验结果表明,与rsync相比较,htsync在同步时的数据传输量较少,同步时间较短.在设计的3种实验场景下,同步时间平均加速比分别为38.70%、30.13%和3.63%,有效地减少了同步时的时间和资源的消耗.
中文关键词: 互联网码号资源公钥证书体系 有序哈希树 数据同步
Abstract:RP(Relying Party) downloads signed objects in RPKI(Resource Public Key Infrastructure) and processes those objects into authorised relations between IP addresses and AS(Autonomous System), which is used to guide the BGP routing. The current RP uses rsync to realize the synchronization, but rsync(Remote Sync) synchronization algorithm does not take the characteristics of the files (directorise) in the RPKI into account. So the synchronization is not efficient. Through the analysis and combining with the characteristics of the files (directories) in the RPKI, this paper designs and realizes a RPKI repository synchronization tool named htsync which is based on ordered hash tree. The experimental results show that, compared with rsync, htsync transmits less data and costs less time during synchronization. In three designed experimental scenario, average speedup ratios of synchronization time are 38.70%, 30.13% and 3.63%, effectively reduce the consumption of time and resources.
keywords: RPKI ordered hash tree data synchronization
文章编号: 中图分类号: 文献标志码:
基金项目:国家自然科学基金(61163044);国家社科基金(12AZD118-4);国家科技支撑课题(2012BAH48F03,2013BAH48F04)
引用文本:
许圣明,马迪,毛伟,王伟.基于有序哈希树的RPKI资料库数据同步方法.计算机系统应用,2016,25(6):141-146
XU Sheng-Ming,MA Di,MAO Wei,WANG Wei.RPKI Repository Synchronization Method Based on Ordered Hash Tree.COMPUTER SYSTEMS APPLICATIONS,2016,25(6):141-146
许圣明,马迪,毛伟,王伟.基于有序哈希树的RPKI资料库数据同步方法.计算机系统应用,2016,25(6):141-146
XU Sheng-Ming,MA Di,MAO Wei,WANG Wei.RPKI Repository Synchronization Method Based on Ordered Hash Tree.COMPUTER SYSTEMS APPLICATIONS,2016,25(6):141-146