Data Domain Cloud Tier: Backup Here, Backup There, Deduplicated Everywhere!

Abhinav Duggal, Fani Jenkins,Philip Shilane, Ramprasad Chinthekindi, Ritesh Shah,Mahesh Kamat

PROCEEDINGS OF THE 2019 USENIX ANNUAL TECHNICAL CONFERENCE(2019)

引用 24|浏览19
暂无评分
摘要
Data Domain has added a cloud tier capability to its on-premises storage appliance, allowing clients to achieve the cost benefits of deduplication in the cloud. While there were many architectural changes necessary to support a cloud tier in a mature storage product, in this paper, we focus on innovations needed to support key functionality for customers. Consider typical customer interactions: First, a customer determines which files to migrate to the cloud by estimating how much space will be freed on the on-premises Data Domain appliance. Second, a customer transfers selected files to the cloud and later restores files back. Finally, a customer deletes a file in the cloud when its retention period has expired. Each of these operations requires significant architectural changes and new algorithms to address both the impact of deduplicated storage and the latency and expense of cloud object storage. We also present analysis from deployed cloud tier systems. As an example, some customers have moved more than 20PB of logical data to the cloud tier and achieved a total compression factor (deduplication * local compression) of 40x or more, resulting in millions of dollars of cost savings.
更多
查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要