SEC-NoSQL：实现NoSQL数据库的高性能安全即服务
SEC-NoSQL: Towards Implementing High Performance Security-as-a-Service for NoSQL Databases
摘要：
在过去的几年中，大数据的爆炸式增长促使云基础设施提供基于云的数据库服务，作为存储和处理大量数据的经济、高效和可扩展的解决方案。因此，NoSQL数据库变得越来越受欢迎，因为它们与其他关系数据库相比具有更好的性能和高可扩展性的固有特性。然而，在这种部署架构中，信息存储在公共云中，保护敏感数据仍然是一个主要问题。由于数据所有者在基于云的数据库解决方案中无法完全控制其敏感数据，因此许多组织不愿意采用数据库即服务（DBaaS）解决方案。最近的一些工作通过引入额外的层来提供加密数据的加密机制来解决这个问题，但是，这些方法更具体，需要对其进行适当的评估，以确保在基于云的生产环境中处理大量数据时，它们是否可以实现高性能和可扩展性。本文提出了一个实用的系统设计和实现，提供安全的NoSQL数据库（SEC-NoSQL）的服务，同时支持执行加密数据的查询，保证系统性能水平。提出了几种不同的实现模型，并使用YCSB基准考虑大量的客户端同时处理的性能进行评估。实验结果表明，我们的设计适合加密数据，同时保持了高性能和可扩展性。此外，为了将我们的解决方案部署为基于云的服务，还包括建立服务级别协议（SLA）的实用指南。
英文摘要：
During the last few years, the explosion of Big Data has prompted cloud infrastructures to provide cloud-based database services as cost effective, efficient and scalable solutions to store and process large volume of data. Hence, NoSQL databases became more and more popular because of their inherent features of better performance and high scalability compared to other relational databases. However, with this deployment architecture where the information is stored in a public cloud, protection against the sensitive data is still being a major concern. Since the data owner does not have the full control over his sensitive data in a cloud-based database solution, many organizations are reluctant to move forward with Database-as-a-Service (DBaaS) solutions. Some of the recent work addressed this issue by introducing additional layers to provide encryption mechanisms to encrypt data, however, these approaches are more application specific and they need to be properly evaluated to ensure whether they can achieve high performance with the scalability when it comes to large volume of data in a cloud-based production environment. This paper proposes a practical system design and implementation to provide Security-as-a-Service for NoSQL databases (SEC-NoSQL) while supporting the execution of query over encrypted data with guaranteed level of system performance. Several different models of implementations are proposed, and their performance is evaluated using YCSB benchmark considering large number of clients processing simultaneously. Experimental results show that our design fits well on encrypted data while maintaining the high performance and scalability. Moreover, to deploy our solution as a cloud-based service, a practical guide establishing Service Level Agreement (SLA) is also included.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
