Analyzing the tradeoff between throughput and latency in multicore scalable in-memory database systems

Hitoshi Mitake, Hiroshi Yamada, Tatsuo Nakajima

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Abstract

    In this paper, we present a tradeoff between throughput and latency in multicore scalable in-memory database systems by showing the results of a performance evaluation and analysis of Masstree, a state-of-the-art multicore scalable data structure that forms the foundation of a variety of multicore scalable database systems. The key technique to make Masstree scalable is an advanced concurrency control technique. Such a technique reduces cache line contention between cores and contributes to high throughput and scalability. However, surprisingly, the worst case latency of the Masstree-based key-value storage system was more than 10 times larger than the score of the memcached system. To detect the main source of the high latency spikes, we analyzed the concurrency control techniques of Masstree. As a result, we found that read-copy update (RCU), an important technique that enables scalability in Masstree, becomes the dominant factor in the high latency spikes. We present a straightforward approach to resolve the latency spikes. In addition, we also show the limitation of the straightforward approach and possible future directions of essential solutions.

    Original languageEnglish
    Title of host publicationProceedings of the 7th ACM SIGOPS Asia-Pacific Workshop on Systems, APSys 2016
    PublisherAssociation for Computing Machinery, Inc
    ISBN (Electronic)9781450342650
    DOIs
    Publication statusPublished - 2016 Aug 4
    Event7th ACM SIGOPS Asia-Pacific Workshop on Systems, APSys 2016 - Hong Kong, China
    Duration: 2016 Aug 42016 Aug 5

    Other

    Other7th ACM SIGOPS Asia-Pacific Workshop on Systems, APSys 2016
    CountryChina
    CityHong Kong
    Period16/8/416/8/5

    Fingerprint

    Concurrency control
    Scalability
    Throughput
    Data storage equipment
    Data structures

    ASJC Scopus subject areas

    • Computer Networks and Communications
    • Hardware and Architecture

    Cite this

    Mitake, H., Yamada, H., & Nakajima, T. (2016). Analyzing the tradeoff between throughput and latency in multicore scalable in-memory database systems. In Proceedings of the 7th ACM SIGOPS Asia-Pacific Workshop on Systems, APSys 2016 [2967361] Association for Computing Machinery, Inc. https://doi.org/10.1145/2967360.2967361

    Analyzing the tradeoff between throughput and latency in multicore scalable in-memory database systems. / Mitake, Hitoshi; Yamada, Hiroshi; Nakajima, Tatsuo.

    Proceedings of the 7th ACM SIGOPS Asia-Pacific Workshop on Systems, APSys 2016. Association for Computing Machinery, Inc, 2016. 2967361.

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Mitake, H, Yamada, H & Nakajima, T 2016, Analyzing the tradeoff between throughput and latency in multicore scalable in-memory database systems. in Proceedings of the 7th ACM SIGOPS Asia-Pacific Workshop on Systems, APSys 2016., 2967361, Association for Computing Machinery, Inc, 7th ACM SIGOPS Asia-Pacific Workshop on Systems, APSys 2016, Hong Kong, China, 16/8/4. https://doi.org/10.1145/2967360.2967361
    Mitake H, Yamada H, Nakajima T. Analyzing the tradeoff between throughput and latency in multicore scalable in-memory database systems. In Proceedings of the 7th ACM SIGOPS Asia-Pacific Workshop on Systems, APSys 2016. Association for Computing Machinery, Inc. 2016. 2967361 https://doi.org/10.1145/2967360.2967361
    Mitake, Hitoshi ; Yamada, Hiroshi ; Nakajima, Tatsuo. / Analyzing the tradeoff between throughput and latency in multicore scalable in-memory database systems. Proceedings of the 7th ACM SIGOPS Asia-Pacific Workshop on Systems, APSys 2016. Association for Computing Machinery, Inc, 2016.
    @inproceedings{cae0cc9750b5464e95a4d98567b20e0a,
    title = "Analyzing the tradeoff between throughput and latency in multicore scalable in-memory database systems",
    abstract = "In this paper, we present a tradeoff between throughput and latency in multicore scalable in-memory database systems by showing the results of a performance evaluation and analysis of Masstree, a state-of-the-art multicore scalable data structure that forms the foundation of a variety of multicore scalable database systems. The key technique to make Masstree scalable is an advanced concurrency control technique. Such a technique reduces cache line contention between cores and contributes to high throughput and scalability. However, surprisingly, the worst case latency of the Masstree-based key-value storage system was more than 10 times larger than the score of the memcached system. To detect the main source of the high latency spikes, we analyzed the concurrency control techniques of Masstree. As a result, we found that read-copy update (RCU), an important technique that enables scalability in Masstree, becomes the dominant factor in the high latency spikes. We present a straightforward approach to resolve the latency spikes. In addition, we also show the limitation of the straightforward approach and possible future directions of essential solutions.",
    author = "Hitoshi Mitake and Hiroshi Yamada and Tatsuo Nakajima",
    year = "2016",
    month = "8",
    day = "4",
    doi = "10.1145/2967360.2967361",
    language = "English",
    booktitle = "Proceedings of the 7th ACM SIGOPS Asia-Pacific Workshop on Systems, APSys 2016",
    publisher = "Association for Computing Machinery, Inc",

    }

    TY - GEN

    T1 - Analyzing the tradeoff between throughput and latency in multicore scalable in-memory database systems

    AU - Mitake, Hitoshi

    AU - Yamada, Hiroshi

    AU - Nakajima, Tatsuo

    PY - 2016/8/4

    Y1 - 2016/8/4

    N2 - In this paper, we present a tradeoff between throughput and latency in multicore scalable in-memory database systems by showing the results of a performance evaluation and analysis of Masstree, a state-of-the-art multicore scalable data structure that forms the foundation of a variety of multicore scalable database systems. The key technique to make Masstree scalable is an advanced concurrency control technique. Such a technique reduces cache line contention between cores and contributes to high throughput and scalability. However, surprisingly, the worst case latency of the Masstree-based key-value storage system was more than 10 times larger than the score of the memcached system. To detect the main source of the high latency spikes, we analyzed the concurrency control techniques of Masstree. As a result, we found that read-copy update (RCU), an important technique that enables scalability in Masstree, becomes the dominant factor in the high latency spikes. We present a straightforward approach to resolve the latency spikes. In addition, we also show the limitation of the straightforward approach and possible future directions of essential solutions.

    AB - In this paper, we present a tradeoff between throughput and latency in multicore scalable in-memory database systems by showing the results of a performance evaluation and analysis of Masstree, a state-of-the-art multicore scalable data structure that forms the foundation of a variety of multicore scalable database systems. The key technique to make Masstree scalable is an advanced concurrency control technique. Such a technique reduces cache line contention between cores and contributes to high throughput and scalability. However, surprisingly, the worst case latency of the Masstree-based key-value storage system was more than 10 times larger than the score of the memcached system. To detect the main source of the high latency spikes, we analyzed the concurrency control techniques of Masstree. As a result, we found that read-copy update (RCU), an important technique that enables scalability in Masstree, becomes the dominant factor in the high latency spikes. We present a straightforward approach to resolve the latency spikes. In addition, we also show the limitation of the straightforward approach and possible future directions of essential solutions.

    UR - http://www.scopus.com/inward/record.url?scp=84986598806&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84986598806&partnerID=8YFLogxK

    U2 - 10.1145/2967360.2967361

    DO - 10.1145/2967360.2967361

    M3 - Conference contribution

    BT - Proceedings of the 7th ACM SIGOPS Asia-Pacific Workshop on Systems, APSys 2016

    PB - Association for Computing Machinery, Inc

    ER -