SFM: Mitigating Read/Write Amplification Problem of LSM-Tree-Based Key-Value Stores

被引:1
作者
Lee, Hoyoung [1 ]
Lee, Minho [2 ]
Eom, Young Ik [3 ]
机构
[1] Sungkyunkwan Univ, Dept Comp Sci & Engn, Suwon 16419, South Korea
[2] Sungkyunkwan Univ, Dept Elect & Comp Engn, Suwon 16419, South Korea
[3] Sungkyunkwan Univ, Dept Elect & Comp Engn, Coll Comp & Informat, Suwon 16419, South Korea
基金
新加坡国家研究基金会;
关键词
Persistent key-value stores; LSM-tree; storage management; database management systems (DBMS); compaction; merge policy; read/write amplification problem;
D O I
10.1109/ACCESS.2021.3098736
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Persistent key-value stores have been widely adopted as storage engines for modern IT infrastructures because they provide high performance with simple design principles. Moreover, many key-value stores commonly employ LSM-tree as their index structure due to its attractive features such as high write throughput and storage space efficiency. Unfortunately, LSM-tree has critical drawbacks in that it leads to write/read amplification problem. One of the prevalent solutions for remedying the write amplification problem is the tiering merge policy that reduces the number of rewrites by delaying merge operations. However, in spite of this advantage, the tiering merge policy may lead to a side-effect that induces high read amplification, increasing search/scan cost for upcoming read operations. In this paper, we concentrate on mitigating the high read amplification problem of the tiering merge policy, while maintaining its low write amplification. To achieve this, we propose a novel LSM-tree scheme, called Spatially Fragmented LSM-tree (SFM), which delays merge operations only for the non-read-intensive key-spaces. For this, SFM identifies the read intensity of each key-spaces by dynamically estimating their read/write hotness. We have implemented SFM based on PebblesDB and evaluated the performance benefits of our scheme under real-world workloads of Facebook. Experimental results clearly show that our scheme improves throughput by up to 1.67x compared with the conventional schemes while maintaining low write amplification, and also indicate that its latency is lowered by up to 41.41% on average by mitigating the read amplification problem of the existing schemes, by up to 43.68%.
引用
收藏
页码:103153 / 103166
页数:14
相关论文
共 43 条
  • [1] AsterixDB: A Scalable, Open Source BDMS
    Alsubaiee, Sattam
    Altowim, Yasser
    Altwaijry, Hotham
    Behm, Alexander
    Borkar, Vinayak
    Bu, Yingyi
    Carey, Michael
    Cetindil, Inci
    Cheelangi, Madhusudan
    Faraaz, Khurram
    Gabrielova, Eugenia
    Grover, Raman
    Heilbron, Zachary
    Kim, Young-Seok
    Chen Li
    Li, Guangqiang
    Ok, Ji Mahn
    Onose, Nicola
    Pirzadeh, Pouria
    Tsotras, Vassilis
    Vernica, Rares
    Wen, Jian
    Westmann, Till
    [J]. PROCEEDINGS OF THE VLDB ENDOWMENT, 2014, 7 (14): : 1905 - 1916
  • [2] Amur H., 2013, 193 CERCS 193 CERCS, P1
  • [3] [Anonymous], 2015, MXNET FLEXIBLE EFFIC
  • [4] [Anonymous], 2013, P ACM SIGMOD INT C M, DOI DOI 10.1145/2463676.2465296
  • [5] Balmau O, 2019, PROCEEDINGS OF THE 2019 USENIX ANNUAL TECHNICAL CONFERENCE, P753
  • [6] Balmau O, 2017, 2017 USENIX ANNUAL TECHNICAL CONFERENCE (USENIX ATC '17), P363
  • [7] Cao Z., 2020, ROCKSDB TRACE REPLAY
  • [8] Cao ZC, 2020, PROCEEDINGS OF THE 18TH USENIX CONFERENCE ON FILE AND STORAGE TECHNOLOGIES, P209
  • [9] Bigtable: A distributed storage system for structured data
    Chang, Fay
    Dean, Jeffrey
    Ghemawat, Sanjay
    Hsieh, Wilson C.
    Wallach, Deborah A.
    Burrows, Mike
    Chandra, Tushar
    Fikes, Andrew
    Gruber, Robert E.
    [J]. ACM TRANSACTIONS ON COMPUTER SYSTEMS, 2008, 26 (02):
  • [10] Chazelle B., 2004, P 15 ANN ACM SIAM S, P30