MicroHash:An Efficient Index Structure for Flash-Based Sensor Devices

1. Demetris Zeinalipour [ zeinalipour@ouc.ac.cy ] School of Pure and Applied Sciences Open University of Cyprus MicroHash:An Efficient Index Structure for Flash-Based Sensor Devices IBM Research, Zurich, Switzerland, Dec. 12th, 2008 http://www.cs.ucy.ac.cy/~dzeina/

2. Presentation Goals • To provide an overview of Wireless Sensor Networks and related Data Acquisition Frameworks. • To highlight some important storage and retrieval challenges that arise in this context.

3. Acknowledgements • This is a joint work with my collaborators at the University of California – Riverside. • Our results were presented in the following papers: • "MicroHash: An Efficient Index Structure for Flash-Based Sensor Devices", D. Zeinalipour-Yazti, S. Lin, V. Kalogeraki, D. Gunopulos and W. Najjar, The 4th USENIX Conference on File and Storage Technologies (FAST’05), San Francisco, USA, December, 2005. • " Efficient Indexing Data Structures for Flash-Based Sensor Devices", S. Lin, D. Zeinalipour-Yazti, V. Kalogeraki, D. Gunopulos, W. Najjar, ACM Transactions on Storage (TOS), ACM Press, Vol.2, No. 4, pp. 468-503, November 2006.

4. Presentation Outline • Overview of Wireless Sensor Networks • Overview of Data Acquisition Frameworks • The MicroHash Index Structure. • MicroHash Experimental Evaluation • Conclusions and Future Work

5. Wireless Sensor Devices • Resource constrained devices utilized for monitoring and researching the physical world at a high fidelity. UC-Berkeley’s Mica2dot Xbow’s Mica UC-Riverside’s RISE Xbow’s TelosB Xbow’s i-mote2

6. Wireless Sensor Device Sensors: Numeric readings in a limited range (e.g., temperature -40F..+250F with one decimal point precision) at a high frequency (2-2000Hz) Storage Radio, used for transmitting the acquired data to some storage site (SINK) (9.6Kbps-250Kbps

7. Wireless Sensor Network 7

8. Wireless Sensor Networks • Applications have already emerged in: • Environmental and habitant monitoring • Seismic and Structural monitoring • Understanding Animal Migrations & Interactions between species • Automation, Tracking, Hazard Monitoring Scenarios, Urban Monitoring etc Great Duck Island – Maine (Temperature, Humidity etc). Golden Gate – SF, Vibration and Displacement of the bridge structure Zebranet (Kenya) GPS trajectory

9. Wireless Sensor Networks The Great Duck Island Study (Maine, USA) • Large-Scale deployment by Intel Research, Berkeley in 2002-2003 (Maine USA). • Focuses on monitoring microclimate in and around the nests of endangered species which are sensitive to disturbance. • They deployed more than 166 motes installed in remote locations (such as 1000 feets in the forest)

10. Wireless Sensor Networks WebServer

11. Wireless Sensor Networks The James Reserve Project, CA, USA Available at: http://dms.jamesreserve.edu/

12. Wireless Sensor Networks Microsoft’s SenseWeb/SensorMap Technology SenseWeb:A peer-produced sensor network that consists of sensors deployed by contributors across the globe SensorMap:A mashup of SenseWeb’s data on a map interface Swiss Experiment (SwissEx) (6 sites on the Swiss Alps) Chicago (Traffic, CCTV Cameras, Temperature, etc.) Available at: http://research.microsoft.com/nec/SenseWeb/

13. Characteristics • The Energy Source is limited. Energy source: AA batteries, Solar Panels 2. Local Processing is cheaper than transmitting over the radio. Transmitting 1 Byte over the Radio consumes as much energy as ~1200 CPU instructions. 3. Local Storage is cheaper than transmitting over the radio. Transmitting 512B over a single-hop 9.6Kbps (915MHz) radio requires 82,000μJ, while writing to local flash only 760μJ.

14. Presentation Outline • Overview of Wireless Sensor Networks (WSN) • Overview of Data Acquisition Frameworks • The MicroHash Index Structure • MicroHash Experimental Evaluation • Conclusions and Future Work

15. Centralized Storage • A Database that collects readings from many Sensors. • Centralized: Storage, Indexing, Query Processing, Triggers, etc.

16. Centralized Storage I • Crossbow’s MoteViewsoftware • NO in-network Aggregation/Filtering • NO in-Network Storage Available at: http://www.xbow.com/

17. Centralized Storage II • TinyDB - A Declarative Interface for Data Acquisition in Sensor Networks. • In-Network Aggregation/Filtering • Limited In-Network Storage (No Indexes) e.g., SELECT MAX(temp) FROM sensors Available at: http://telegraph.cs.berkeley.edu/tinydb/

18. Centralized Storage: Conclusions • Frameworks such as TinyDB: • - Are suitable for continuous queries. • - Push aggregation in the network but keep much of the processing at the sink. • New Challenges: • Many applications DON’T require the continuous evaluation of user queries (e.g., historic query: Find the average temperature for the last 6 months) • In many applications there is no sink (e.g., remote deployments and mobile sensor networks) • Local Storage on sensor devices keeps growing. • RISE supports a 1GB external SD Card • I-Mote-2 support 32MB Flash/32MB SRAM

19. Our Model: In-Situ Data Storage • Data remains In-situ (at the generating site) in a sliding window fashion. • When required, users conduct on-demand queries to retrieve information of interest. A Network of Sensor Databases

20. In-Situ Data Storage: Motivation Center for Conservation Biology@ UCR: Research of Soil-Organisms • A set of sensors monitor the CO2 levels in the soil over a large window of time. • Not a real-time application. • Most acquired values are not of particular interest. • D. Zeinalipour-Yazti, S. Neema, D. Gunopulos, V. Kalogeraki and W. Najjar, • "Data Acquision in Sensor Networks with Large Memories", IEEE Intl. Workshop on NetworkingMeets Databases NetDB (ICDE'2005), Tokyo, Japan, 2005.

21. Presentation Outline • Overview of Wireless Sensor Networks • Overview of Data Acquisition Frameworks • The MicroHash Index Structure • MicroHash Experimental Evaluation • Conclusions and Future Work

22. Flash Memory at a Glance • The most prevalent storage medium used for Sensor Devices is Flash Memory (NAND Flash) • Fastest growing memory market (‘05 $8.7B, ‘06:$11B) • (NAND) Flash Advantages • Simple Cell Architecture (high capacity in a small surface) => Economical Reproduction • Fast Random Access (50-80 μs) compared to 10-20ms in Disks • Shock Resistant • Power Efficient Surface mount NAND flash Removable NAND Devices

23. Flash Memory at a Glance • Write-Constrain: Writing can only be performed at a page granularity (256B~512B) to an empty page (if occupied we need to delete its content). • Delete-Constrain: Erasure of a page can only be performed at a block granularity (i.e. 8KB~64KB) • Wear-Constrain: Each page can only be written a limited number of times (typically 10,000-100,000) Measurements using RISE Energy(Page Size = 512 B) Read = 24 μJ Write =763μJ Block Erase =425μJ Asymmetric Read/Write Energy Cost!

24. MicroHash Index Objectives • General Objectives • Provide efficient access to any record stored on flash by timestamp or value. • Execute a wide spectrum of queries based on our index, similarly to generic DB indexes. • Design Objectives (Adhere to Flash Constrains): • Avoid wearing out specific pages. • Minimize random access deletions of pages. • Minimize main memory (SRAM) structures • SRAM is extremely limited (8KB-64KB). • Small memory-footprint => quick initialization.

25. Main Structures • 4 Page Types: a) Root Page, b) Directory Page, c) Index Page and d) Data Page • 4 Phases of Operation: a) Initialization, b) Growing, c) Repartition and d) Garbage Collect.

26. Growing the MicroHash Index • Collect data in an SRAM buffer page Pwrite • When Pwrite is full flush it out to flash media • Next create index records for each data record in Pwrite • If SRAM gets full, Index pages are forced out to flash media by an LRU policy. Buffer Buffer Pwrite Pwrite (ts, 74F) 40 50 60 Index 70 x 80 90 Index Pages Directory

27. Growing the MicroHash Index A populated Flash Media Flash Media idx: next empty page

28. Garbage Collection in MicroHash • When the media gets full some pages need to be deleted => delete the oldest pages. • Oldest Block? The next block following the idx pointer. • Note: • This might create invalid index records. • This will be handled by our search algorithm

29. Directory Repartition in MicroHash • MicroHash starts out with a directory that is segmented into equiwidth buckets • e.g., divide the temperature range [0,100] into c buckets) • Not efficient as certain buckets will never be utilized • Consider the first few or last few buckets below.

30. Directory Repartition in MicroHash • If bucket A links to more thanτ index pages, evict the least used bucket B and segment the full bucket A into A and A’ • We want to avoid bucket reassignments of old records as this would be very expensive >> Add(18) << A: _ _ _ Example: τ=3 C: #entries since last split S: timestamp of last addition

31. Searching in MicroHash • Searching by value “Find the timestamp (s) on which the temperature was 100F” • Simple operation in MicroHash • We simply find the right Directory Bucket, from there the respective index page and then data record (page-by-page) • Searching by timestamp “Find the temperature of some sensor on a given timestamp tq” • Problem: Index pages are mixed together with data pages. • Solutions: 1. Binary Search (O(logn), 18 pages for a 128MB media) 2. LBSearch (less than 10 pages for a 128MB media) 3. ScaleSearch (better than LBSearch, ~4.5 pages for a 128MB media)

32. LBSearch and ScaleSearch Solutions to the Search-by-timestamp problem: • LBSearch: We recursively create a lower bound on the position of tq until the given timestamp is located. • ScaleSearch: Quite similar to LBSearch, however in the first step we proceed more aggressively (by exploiting data distribution) tq=300 tq=350 Query tq=500 tq=420 tq=490 tq=500

33. Searching Bottlenecks • Index Pages written on flash might not be fully occupied • When we access these pages we transfer a lot of emptybytes (padding) between the flash media and SRAM. • Proposed Solutions: • Solution 1: Two-Phase Page Reads • Solution 2: ELF-like Chaining of Index Pages

34. Improving Search Performance • Solution 1: Utilize Two-Phase Page Reads. • Reads the 8B header from the flash media. • Then read the correct payload in the next phase.

35. Improving Search Performance • Solution 2: Avoid non-full index pages using ELF*. • ELF: a linked list in which each page, other than the last page, is completely full. • keeps copying the last non-full page into a newer page, when new records are requested to be added. *Dai et. al.,Efficient Log Structured Flash File System, SenSys 2004

36. Presentation Outline • Overview of Wireless Sensor Networks • Overview of Data Acquisition Frameworks • The MicroHash Index Structure • MicroHash Experimental Evaluation • Conclusions and Future Work

37. Experimental Evaluation • Implemented MicroHash in nesC. • We tested it using TinyOS along with a trace-driven experimental methodology. • Datasets: • Washington State Climate • 268MB dataset contains readings in 2000-2005. • Great Duck Island • 97,000 readings between October and November 2002. • Evaluation Parameters: i) Space Overhead, ii) Energy Overhead, iii) Search Performance 37

38. Space Overhead of Index • Measure: IndexPages/(DataPages+IndexPages) • Two Index page layouts • Offset, an index record has the following form {datapageid,offset} • NoOffset, in which an index record has the form {datapageid} • 128 MB flash media (256,000 pages) • Conclusions: • Space Overhead is minimized with: • A Larger Buffer • No-Offset • Pressure Data

39. Space Overhead of Index Bitmap Representations of the Flash Media 2.5KB Buffer Index | Data Pages Index Occupancy Black denotes the index pages 10KB Buffer Index | Data Pages Index Occupancy Increasing the Buffer Decreases the Index Overhead

40. Search Performance • Measure: # of page reads to find a record by timestamp • 2 Index page layouts (128MB flash, varying SRAM) • Anchor: Index Pages store the last known timestamp • No Anchor: Timestamp is only stored in Data Pages • Conclusions: • Search Performance is increased with: • Larger Write Buffer during Indexing • Anchors • ScaleSearch

41. Indexing the Great Duck Island Trace • Used 3KB index buffer and a 4MB flash card to store all the 97,000 20-byte data readings. • The index never requires more than 28% additional space • Indexing the records has only a small increase in energy demand: the energy cost of storing the records on flash without an index is 3042mJ • We were able to find any record by its timestamp with 4.75 page reads on average

42. Presentation Outline • Overview of Wireless Sensor Networks • Overview of Data Acquisition Frameworks • The MicroHash Index Structure • MicroHash Experimental Evaluation • Conclusions and Future Work

43. Conclusions & Future Work • We proposed the MicroHash index, which is an efficient external memory hash index for sensor devices that addresses the distinct characteristics of flash memory • Our experimental evaluation shows that the structure we propose is both efficient and practical • Future work: • Develop a complete library of indexes and data structures (stacks, queues, b+trees, etc.) • Buffer optimizations and Online Compression • Support Range Queries

44. Demetris Zeinalipour • Thank you! • Questions? • Related Publications • "MicroHash: An Efficient Index Structure for Flash-Based Sensor Devices", D. Zeinalipour,S. Lin, V. Kalogeraki, D. Gunopulos, W. Najjar, In USENIX FAST’05. • " Efficient Indexing Data Structures for Flash-Based Sensor Devices", • ACM Transactions on Storage (TOS), November 2006. MicroHash:An Efficient Index Structure for Flash-Based Sensor Devices Presentation and publications available at: http://www.cs.ucy.ac.cy/~dzeina/

45. Backup Slides

46. The Anatomy of a Sensor Device • Processor, in various(sleep, idle, active) modes • Power source AA or Coin batteries, Solar Panels • SRAM used for the program code and for in-memory buffering. • LEDs used for debugging • Radio, used for transmitting the acquired data to some storage site (SINK) (9.6Kbps-250Kbps) Storage • Sensors: Numeric readings in a limited range (e.g. temperature -40F..+250F with one decimal point precision) at a high frequency (2-2000Hz)

47. Sensor Devices & Capabilities Sensing Capabilities • Light • Temperature • Humidity • Pressure • Tone Detection • Wind Speed • Soil Moisture • Location (GPS) • etc…

48. In-Network Storage: Data Centric Storage • Outline • Data is stored on specific nodes in the network (e.g., humidity on node A and temperature on node B) • Locating Data can be performed without flooding (e.g., temperature-related data is stored on node B). Q: SELECT nodeid where temp=100F Temperature Store

49. The Programming Cycle • The Operating System TinyOS (UC-Berkeley): Component-based architecture that allows programmers to wire together the minimum required components in order to minimize code size and energy consumption (The operating system is really a number of libraries that can be statically linked to the sensor binary at compile time) • The Programming Language nesC (Intel Research, Berkeley): an event-based C-variant optimized for programming sensor devices event result_t Clock.fire() { state = !state; if (state) call Leds.redOn(); else call Leds.redOff(); } “Hello World”: Blinking the red LED!

50. The Programming Cycle The Testing Environment • Debugging code directly on a sensor device is a tedious procedure • nesC allows programmers to compile their code to • A Binary File that is burnt to the sensor • A Binary File that runs on a PC • TOSSIM (TinyOS Simulation) is the environment which allows programmers to simulate the PC binary directly on a PC. • This enables accurate simulations, fine grained energy modeling (with PowerTOSSIM) and visualization (TinyViz)