预测索引
Predictive Indexing
摘要：
在数据库管理系统（DBMS）中，对自动索引调优进行了大量的研究。但是，这些解决方案中的大多数都是通过回顾性地一次性进行计算代价高昂的物理设计更改来调整索引配置的。这样的改变降低了DBMS在处理期间的性能，并且由于工作负载转移和相关联的改变之间的延迟而降低了在随后的查询处理期间的实用性。一种更好的方法是生成随时间调整物理设计的小更改，预测这些更改的实用性，并提前应用它们以最大限度地发挥其影响。本文介绍了预测索引，不断提高数据库的物理设计使用轻量级的物理设计的变化。它使用机器学习模型来预测这些更改的效用，并不断改进数据库的索引配置以处理不断变化的工作负载。我们引入了一个轻量级的混合扫描操作，DBMS可以利用部分建立的索引进行查询处理。我们的评估表明，预测索引提高了3.5- 5.2倍的数据库管理系统的吞吐量相比，其他国家的最先进的索引方法。我们证明了预测索引与其他轻量级自动化物理设计调优方法无缝配合。
英文摘要：
There has been considerable research on automated index tuning in database management systems (DBMSs). But the majority of these solutions tune the index configuration by retrospectively making computationally expensive physical design changes all at once. Such changes degrade the DBMS's performance during the process, and have reduced utility during subsequent query processing due to the delay between a workload shift and the associated change. A better approach is to generate small changes that tune the physical design over time, forecast the utility of these changes, and apply them ahead of time to maximize their impact. This paper presents predictive indexing that continuously improves a database's physical design using lightweight physical design changes. It uses a machine learning model to forecast the utility of these changes, and continuously refines the index configuration of the database to handle evolving workloads. We introduce a lightweight hybrid scan operator with which a DBMS can make use of partially-built indexes for query processing. Our evaluation shows that predictive indexing improves the throughput of a DBMS by 3.5--5.2x compared to other state-of-the-art indexing approaches. We demonstrate that predictive indexing works seamlessly with other lightweight automated physical design tuning methods.
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。
