Chronos: a NoSQL system on flash memory for industrial process data

详细信息    查看全文

• 作者：Brice Chardin ; Jean-Marc Lacombe ; Jean-Marc Petit
• 关键词：Database ; Flash memory ; NoSQL system ; Data historian
• 刊名：Distributed and Parallel Databases
• 出版年：2016
• 出版时间：September 2016
• 年：2016
• 卷：34
• 期：3
• 页码：293-319
• 全文大小：1,620 KB
• 参考文献：1.Pasteur, O., Léger, S.: Results of the use of MySQL free DBMS as a data historian. EDF internal technical repport, H-P1D-2007-02670-FR (2007)
  2.Bouganim, L., Jónsson, B.T., Bonnet, P.: uFLIP: Understanding flash IO patterns. In: CIDR’09: 4th Biennial Conference on Innovative Data Systems Research, Asilomar, USA (2009)
  3.Wang, Y., Goda, K., Kitsuregawa, M.: Evaluating non-in-place update techniques for flash-based transaction processing systems. In: DEXA’09: 20th International Conference on Database and Expert Systems Applications, Linz, Austria, pp. 777–791 (2009)
  4.Birrell, A., Isard, M., Thacker, C., Wobber, T.: A design for high-performance flash disks. Oper. Syst. Rev. 41(2), 88–93 (2007)CrossRef
  5.Chardin, B., Pasteur, O., Petit, J.-M.: An FTL-agnostic layer to improve random write on flash memory. In: FlashDB’11: 1st International Workshop on Flash-Based Database Systems, Hong Kong, China, pp. 214–225 (2011)
  6.Arpaci-Dusseau, A.C., Arpaci-Dusseau, R.H., Prabhakaran, V.: Removing the costs of indirection in flash-based SSDs with nameless writes. In: HotStorage’10: 2nd Workshop on Hot Topics in Storage and File Systems, Boston, USA, pp. 1–5 (2010)
  7.Graefe, G.: Modern B-tree techniques. Found. Trends Databases 3(4), 203–402 (2011)CrossRef
  8.Olson, M.A., Bostic, K., Seltzer, M.I., Berkeley, D.B.: In: FREENIX’99: 1999 USENIX Annual Technical Conference. FREENIX Track, Monterey, USA, pp. 183–191 (1999)
  9.Transaction Processing Performance Council: TPC Benchmark C Standard Specification (2007)
  10.Transaction Processing Performance Council: TPC Benchmark H Standard Specification (2008)
  11.Arasu, A., Cherniack, M., Galvez, E., Maier, D., Maskey, A.S., Ryvkina, E., Stonebraker, M., Tibbetts, R.: Linear road: a stream data management benchmark. In: VLDB’04: 30th International Conference on Very Large Data Bases, Toronto, Canada, pp. 480–491 (2004)
  12.Pasteur, O.: Overview of the long-term stored data in power generation units. EDF internal technical repport, H-P1D-2007-01076-FR (2007)
  13.Chardin, B., Lacombe, J.-M., Petit, J.-M.: Data historians in the data management landscape. In: TPCTC’12: 4th TPC Technology Conference on Performance Evaluation & Benchmarking, Istanbul, Turkey (2012)
  14.Shafer, I., Sambasivan, R.R., Rowe, A., Ganger, G.R.: Specialized storage for big numeric time series. In: HotStorage’13: 5th Workshop on Hot Topics in Storage and File Systems, pp. 1–5. (2013)
  15.Sumbaly, R., Kreps, J., Gao, L., Feinberg, A., Soman, C., Shah, S.: Serving large-scale batch computed data with project voldemort. In: Proceedings of the 10th USENIX Conference on File and Storage Technologies, FAST’12, pp. 18–18 (2012)
  16.Woodhouse, D.: Jffs : The Journalling Flash File System. http://​sources.​redhat.​com/​jffs2/​jffs2 (2001)
  17.Konishi, R., Amagai, Y., Sato, K., Hifumi, H., Kihara, S., Moriai, S.: The linux implementation of a log-structured file system. SIGOPS Oper. Syst. Rev. 40(3), 102–107 (2006)CrossRef
  18.Stoica, R., Athanassoulis, M., Johnson, R., Ailamaki, A.: Evaluating and repairing write performance on flash devices. In: DaMoN’09: 5th International Workshop on Data Management on New Hardware, Providence, USA, pp. 9–14 (2009)
  19.Lee, S.-W., Moon, B.: Design of flash-based DBMS: an in-page logging approach. In: SIGMOD’07: 33rd International Conference on Management of Data, Beijing, China, pp. 55–66 (2007)
  20.Kim, Y.-R., Whang, K.-Y., Song, I.-Y.: Page-differential logging: an efficient and DBMS-independent approach for storing data into flash memory. In: SIGMOD’10: 36th International Conference on Management of Data, Indianapolis, USA, pp. 363–374 (2010)
  21.Li, Y., Hey, B., Luo, Q., Yi, K.: Tree indexing on flash disks. In: ICDE’09: 25th International Conference on Data Engineering, Shanghai, China, pp. 1303–1306 (2009)
  22.Agrawal, D., Ganesan, D., Sitaraman, R., Diao, Y., Singh, S.: Lazy-adaptive tree: an optimized index structure for flash devices. Proc. VLDB Endow. 2(1), 361–372 (2009)CrossRef
  23.Dobreva, V., Albutiu, M.-C., Brunel, R., Neumann, T., Kemper, A.: Get tracked: a triple store for rfid traceability data. In: Advances in Databases and Information Systems, vol. 7503 of Lecture Notes in Computer Science, Springer, Berlin, pp. 167–180 (2012)
  24.Agrawal, N., Prabhakaran, V., Wobber, T., Davis, J.D., Manasse, M., Panigrahy, R.: Design tradeoffs for SSD performance. In: USENIX’08: 2008 USENIX Annual Technical Conference, Boston, USA, pp. 57–70 (2008)
  25.Lee, S.-W., Moon, B., Park, C.: Advances in flash memory SSD technology for enterprise database applications. In: SIGMOD’09: 35th International Conference on Management of Data, Providence, USA, pp. 863–870 (2009)
  26.Jung, M., Kandemir, M.: Revisiting widely held SSD expectations and rethinking system-level implications. SIGMETRICS Perform. Eval. Rev. 41(1), 203–216 (2013)CrossRef
  27.Bjorling, M., Bonnet, P., Bouganim, L., Dayan, N.: The necessary death of the block device interface. In: CIDR’13: 6th Biennial Conference on Innovative Data Systems Research, Asilomar, USA (2013)
  
• 作者单位：Brice Chardin (1)
  Jean-Marc Lacombe (2)
  Jean-Marc Petit (3)
  
  1. LIAS/ISAE-ENSMA, Université de Poitiers, 1 avenue Clément Ader, 86961, Chasseneuil, France
  2. EDF R&D, 6 Quai Watier, 78401, Chatou, France
  3. Université de Lyon, CNRS, LIRIS, INSA-Lyon, 7 av Jean Capelle, 69621, Villeurbanne, France
  
• 刊物类别：Computer Science
• 刊物主题：Database Management
  Data Structures
  Information Systems Applications and The Internet
  Operating Systems
  Memory Structures
  
• 出版者：Springer Netherlands
• ISSN：1573-7578

文摘

Within Électricité de France (EDF) hydroelectric power stations, IGCBoxes are industrial mini PCs dedicated to industrial process data archiving. These equipments expose distinctive features, mainly on their storage system based exclusively on flash memory due to environmental constraints. This type of memory had notable consequences on data acquisition performance, with a substantial drop compared with hard disk drives. In this setting, we have designed Chronos, an open-source NoSQL system for sensor data management on flash memories. Chronos includes an efficient quasi-sequential write pattern along with an index management technique adapted for process data management. As a result, Chronos supports a higher velocity for inserted data, with acquisition rates improved by a factor of 20–54 over different solutions, therefore solving a practical bottleneck for EDF.