智能存储NDP操作结果集管理

Proceedings of the 18th International Workshop on Data Management on New Hardware Pub Date : 2022-06-12 DOI:10.1145/3533737.3535097

Tobias Vinçon, Christian Knoedler, Arthur Bernhardt, Leonardo Solis-Vasquez, Lukas Weber, Andreas Koch, Ilia Petrov

{"title":"智能存储NDP操作结果集管理","authors":"Tobias Vinçon, Christian Knoedler, Arthur Bernhardt, Leonardo Solis-Vasquez, Lukas Weber, Andreas Koch, Ilia Petrov","doi":"10.1145/3533737.3535097","DOIUrl":null,"url":null,"abstract":"Current data-intensive systems suffer from scalability as they transfer massive amounts of data to the host DBMS to process it there. Novel near-data processing (NDP) DBMS architectures and smart storage can provably reduce the impact of raw data movement. However, transferring the result-set of an NDP operation may increase the data movement, and thus, the performance overhead. In this paper, we introduce a set of in-situ NDP result-set management techniques, such as spilling, materialization, and reuse. Our evaluation indicates a performance improvement of 1.13 × to 400 ×.","PeriodicalId":381503,"journal":{"name":"Proceedings of the 18th International Workshop on Data Management on New Hardware","volume":"516 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2022-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Result-Set Management for NDP Operations on Smart Storage\",\"authors\":\"Tobias Vinçon, Christian Knoedler, Arthur Bernhardt, Leonardo Solis-Vasquez, Lukas Weber, Andreas Koch, Ilia Petrov\",\"doi\":\"10.1145/3533737.3535097\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Current data-intensive systems suffer from scalability as they transfer massive amounts of data to the host DBMS to process it there. Novel near-data processing (NDP) DBMS architectures and smart storage can provably reduce the impact of raw data movement. However, transferring the result-set of an NDP operation may increase the data movement, and thus, the performance overhead. In this paper, we introduce a set of in-situ NDP result-set management techniques, such as spilling, materialization, and reuse. Our evaluation indicates a performance improvement of 1.13 × to 400 ×.\",\"PeriodicalId\":381503,\"journal\":{\"name\":\"Proceedings of the 18th International Workshop on Data Management on New Hardware\",\"volume\":\"516 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2022-06-12\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Proceedings of the 18th International Workshop on Data Management on New Hardware\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1145/3533737.3535097\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the 18th International Workshop on Data Management on New Hardware","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3533737.3535097","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 0

摘要

当前的数据密集型系统在可伸缩性方面存在问题，因为它们需要将大量数据传输到主机DBMS进行处理。新颖的近数据处理(NDP) DBMS体系结构和智能存储可以有效地减少原始数据移动的影响。但是，传输NDP操作的结果集可能会增加数据移动，从而增加性能开销。在本文中，我们介绍了一套原位NDP结果集管理技术，如溢出、物化和重用。我们的评估表明性能提高了1.13倍到400倍。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

Result-Set Management for NDP Operations on Smart Storage

Current data-intensive systems suffer from scalability as they transfer massive amounts of data to the host DBMS to process it there. Novel near-data processing (NDP) DBMS architectures and smart storage can provably reduce the impact of raw data movement. However, transferring the result-set of an NDP operation may increase the data movement, and thus, the performance overhead. In this paper, we introduce a set of in-situ NDP result-set management techniques, such as spilling, materialization, and reuse. Our evaluation indicates a performance improvement of 1.13 × to 400 ×.

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

Proceedings of the 18th International Workshop on Data Management on New Hardware

自引率

0.00%

发文量