{"title":"基于管道策略的时间戳连续数据保护方法","authors":"Jingning Liu, Tianming Yang, Zuoheng Li, Ke Zhou","doi":"10.1109/FCST.2008.27","DOIUrl":null,"url":null,"abstract":"Traditional cold backup technology can not fight against data loss between two consecutive backups which usually has a time span of 24 hours. Hot backup technology, such as snapshot decreases the amount of potential data loss, but still cannot completely prevent data loss. More recently, a new technology called continuous data protection (CDP) is emerging to provide more powerful data protection. CDP captures every change to the data volume hence can provide any point-in-time data views and enable timely recovery to any point-in-time of user data. However, existing implementations of CDP usually suffer from poor write performance and cannot meet the data protection demand of some high performance application environments. How to improve the CDP performance in terms of write throughput and response time is one of the most important considerations in implementing CDP. This paper presents a time-stamp pipeline strategy continuous data protection approach called TSPSCDP which uses pipeline strategy in write operation of CDP to improve data processing capability. Our experiments demonstrated that TSPSCDP can not only recover data to any point-in-time upon a failure but also improve the CDP performance in terms of write throughput and respond time upon the ordinary CDP by a factor of 2 and about 25% respectively.","PeriodicalId":206207,"journal":{"name":"2008 Japan-China Joint Workshop on Frontier of Computer Science and Technology","volume":"133 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2008-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"5","resultStr":"{\"title\":\"TSPSCDP: A Time-Stamp Continuous Data Protection Approach Based on Pipeline Strategy\",\"authors\":\"Jingning Liu, Tianming Yang, Zuoheng Li, Ke Zhou\",\"doi\":\"10.1109/FCST.2008.27\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Traditional cold backup technology can not fight against data loss between two consecutive backups which usually has a time span of 24 hours. Hot backup technology, such as snapshot decreases the amount of potential data loss, but still cannot completely prevent data loss. More recently, a new technology called continuous data protection (CDP) is emerging to provide more powerful data protection. CDP captures every change to the data volume hence can provide any point-in-time data views and enable timely recovery to any point-in-time of user data. However, existing implementations of CDP usually suffer from poor write performance and cannot meet the data protection demand of some high performance application environments. How to improve the CDP performance in terms of write throughput and response time is one of the most important considerations in implementing CDP. This paper presents a time-stamp pipeline strategy continuous data protection approach called TSPSCDP which uses pipeline strategy in write operation of CDP to improve data processing capability. Our experiments demonstrated that TSPSCDP can not only recover data to any point-in-time upon a failure but also improve the CDP performance in terms of write throughput and respond time upon the ordinary CDP by a factor of 2 and about 25% respectively.\",\"PeriodicalId\":206207,\"journal\":{\"name\":\"2008 Japan-China Joint Workshop on Frontier of Computer Science and Technology\",\"volume\":\"133 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2008-12-27\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"5\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2008 Japan-China Joint Workshop on Frontier of Computer Science and Technology\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/FCST.2008.27\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2008 Japan-China Joint Workshop on Frontier of Computer Science and Technology","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/FCST.2008.27","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
TSPSCDP: A Time-Stamp Continuous Data Protection Approach Based on Pipeline Strategy
Traditional cold backup technology can not fight against data loss between two consecutive backups which usually has a time span of 24 hours. Hot backup technology, such as snapshot decreases the amount of potential data loss, but still cannot completely prevent data loss. More recently, a new technology called continuous data protection (CDP) is emerging to provide more powerful data protection. CDP captures every change to the data volume hence can provide any point-in-time data views and enable timely recovery to any point-in-time of user data. However, existing implementations of CDP usually suffer from poor write performance and cannot meet the data protection demand of some high performance application environments. How to improve the CDP performance in terms of write throughput and response time is one of the most important considerations in implementing CDP. This paper presents a time-stamp pipeline strategy continuous data protection approach called TSPSCDP which uses pipeline strategy in write operation of CDP to improve data processing capability. Our experiments demonstrated that TSPSCDP can not only recover data to any point-in-time upon a failure but also improve the CDP performance in terms of write throughput and respond time upon the ordinary CDP by a factor of 2 and about 25% respectively.