Bigtable: A Distributed Storage System for Structured Data

Size: px
Start display at page:

Download "Bigtable: A Distributed Storage System for Structured Data"

Transcription

1 Bigtable: A Distributed Storage System for Structured Data Fay Chang, Jeffrey Dean, Sanjay Ghemawat, Wilson C. Hsieh, Deborah A. Wallach, Mike Burrows, Tushar Chandra, Andrew Fikes, Robert E. Gruber ~Harshvardhan

2 IntroducJon BigTable: A distributed storage system for managing structured data Designed to scale to a very large size: PetaBytes of data across 10 3 servers Projects using BigTable: Web Indexing, Google Earth, Personalized Search, Google AnalyJcs Different demands on BigTable: Data size: URLs, to webpages to satelite imagery Latency: backend bulk processing to real Jme data serving Provide flexible, high performance solujon for all Goals: Wide Applicability, Scalability, High Performance, High Availability

3 IntroducJon Paper describes simple data model provided by BigTable: Resembles a DataBase Different interface and does not provide a full relajonal data model Simple data model and dynamic control over data layout and format Data indexed using row and column names: arbitrary strings Clients control locality of data through schema choices Schema parameters let clients dynamically control to serve data out of memory of disk PresentaJon Outline: 1. Data Model 2. Client API 3. Underlying Google Infrastructure 4. BigTable ImplementaJon 5. Refinements 6. Performance 7. Examples 8. Conclusions

4 QuesJons? End of 5min PresentaJon

5 1. Data Model 2. Client API 3. Underlying Google Infrastructure 4. BigTable ImplementaJon 5. Refinements 6. Performance 7. Examples 8. Conclusions

6 Data Model BigTable is a sparse, distributed, persistent mulj dimensional sorted map Map indexed by: Row key, Column key, and Timestamp Each value in map is an uninterpreted array of Bytes (row:string, column:string, time:int64) string

7 Data Model Example: WebTable Want to keep a copy of a large collecjon of web pages and related informajon Could be used by many different projects Use URLs as row keys Various aspects of web pages as column names Store contents of web pages in the Contents: column under Jmestamps when they were fetched

8 Data Model Example: WebTable Want to keep a copy of a large collecjon of web pages and related informajon Could be used by many different projects Use URLs as row keys Various aspects of web pages as column names Store contents of web pages in the Contents: column under Jmestamps when they were fetched

9 Data Model: Rows Rows: Row keys are arbitrary strings: up to 64KB in size; ~ Bytes typically Each read or write of data in single row is atomic: helps in concurrency BigTable maintains data in lexicographic order by row key Tablets: Row range for a table is dynamically par22oned Each row range is called a tablet Tablet is the unit of distribu2on and load balancing Reads of short row ranges are efficient and require less communica2on Properly selected row keys can get good locality for data access Eg: in WebTable, pages in same domain are conjguous due to reverse hostnames as keys Storing pages from same domain together makes some host and domain analyses more efficient

10 Data Model: Column Families Column Families: Column keys are grouped into sets called column families Form the basic unit of access control Intend to have small # of disjnct column families in a table Rarely change during operajon In contrast, table may have an unbounded number of columns Named in the following syntax: family:qualifier Family names must be printable, qualifiers may be arbitrary strings WebTable Example: We can have two useful column families: language: contains only one column key; the language of the web page anchor: muljple column keys: qualifier is name of referring site; cell contents is the link text

11 Data Model: Timestamps Timestamps: Each cell can contain muljple versions of the same data Versions indexed by Jmestamps Timestamps are 64 bit integers Different versions stored in decreasing Jmestamp order Most recent versions can be read first Can garbage collect cell versions automajcally: Can specify either that only the last n versions be kept Or only new enough versions be kept e.g., only keep values that were wrimen in the last seven days WebTable Example: Timestamps of crawled pages stored in the contents: column Time at which the page version was crawled Garbage collecjon to keep only the most recent three versions

12 1. Data Model 2. Client API 3. Underlying Google Infrastructure 4. BigTable ImplementaJon 5. Refinements 6. Performance 7. Examples 8. Conclusions

13 Client API API: Provides funcjons for: CreaJng and delejng tables and column families Changing cluster, table, and column family metadata, access control, etc. ApplicaJons can: Write or delete values in BigTable Look up values from individual rows Iterate over a subset of the data in a table

14 Client API API: Supports single row transacjons Can perform atomic read modify write sequences on data in a single row Supports execujon of client supplied scripts in the server address spaces Scripts wrimen in Sawzall: Language developed at Google for data processing Allows various forms of data transformajons, filtering based on arbitrary expressions, and summarizajon via a variety of operators Can be used with MapReduce: Framework for running large scale parallel computajons developed at Google Used both as input source and output target for MapReduce

15 API: Client API: Example

16 1. Data Model 2. Client API 3. Underlying Google Infrastructure 4. BigTable ImplementaJon 5. Refinements 6. Performance 7. Examples 8. Conclusions

17 Building Blocks Built on several other pieces of Google infrastructure: Uses Google File System (GFS) to store and log data files Depends on a cluster management system for: Uses Google SSTable file format internally to store data Relies on Chubby

18 Building Blocks Built on several other pieces of Google infrastructure: Uses Google File System (GFS) to store and log data files Depends on a cluster management system for: Uses Google SSTable file format internally to store data Relies on Chubby

19 Building Blocks Built on several other pieces of Google infrastructure Uses Google File System (GFS) to store and log data files Depends on a cluster management system for: Scheduling jobs Managing resources on shared machines Dealing with machine failures Monitoring machine status Uses Google SSTable file format internally to store data Relies on Chubby

20 Building Blocks Built on several other pieces of Google infrastructure: Uses Google File System (GFS) to store and log data files Depends on a cluster management system for: Uses Google SSTable file format internally to store data Relies on Chubby

21 Building Blocks Built on several other pieces of Google infrastructure Uses Google File System (GFS) to store and log data files Depends on a cluster management system for: Uses Google SSTable file format internally to store data Provides a persistent, ordered immutable map from keys to values Provides operajons for: Look up the value associated with a specified key Iterate over all key/value pairs in a specified key range Relies on Chubby

22 Building Blocks Built on several other pieces of Google infrastructure: Uses Google File System (GFS) to store and log data files Depends on a cluster management system for: Uses Google SSTable file format internally to store data Relies on Chubby

23 Building Blocks Built on several other pieces of Google infrastructure Uses Google File System (GFS) to store and log data files Depends on a cluster management system for: Uses Google SSTable file format internally to store data Relies on Chubby Highly available and persistent distributed lock service Used for: Ensuring at most one acjve master at any Jme Storing bootstrap locajon of BigTable data Discovering tablet servers and finalize tablet server deaths Storing schema info and access control lists

24 1. Data Model 2. Client API 3. Underlying Google Infrastructure 4. BigTable ImplementaJon 5. Refinements 6. Performance 7. Examples 8. Conclusions

25 ImplementaJon Three major components: Library linked into every client Many Tablet servers Master server Tablet: BigTable cluster stores a number of tables Each table consists of a set of tablets Each tablet contains all data associated with a row range IniJally, each table consists of one tablet As table grows, it is automajcally split into muljple tablets Each approx MB in size by default

26 ImplementaJon Three major components: Library linked into every client Caches tablet locajons If locajon unknown or incorrect, get informajon Requires 3 6 network round trips 3 for unknown, 6 for stale Many Tablet servers Master server

27 ImplementaJon Three major components: Library linked into every client Many Tablet servers Master server

28 ImplementaJon Three major components: Library linked into every client Many Tablet servers Manages a set of tablets (typ tablets/server) Handles read and write requests to the tablets that it has loaded Splits tablets that have grown too large Can be dynamically added or removed from a cluster Master server

29 ImplementaJon Three major components: Library linked into every client Many Tablet servers Master server

30 ImplementaJon Three major components: Library linked into every client Many Tablet servers Master server Responsible for: Assigning tablets to tablet servers DetecJng the addijon and expirajon of tablet servers Balancing tablet server load Garbage collecjon of files in GFS Handles schema changes such as table and column family creajons Client data does not move through the master: Clients communicate directly with tablet servers Most clients never communicate with the master Therefore, master is lightly loaded in pracjce

31 ImplementaJon: Details Tablet LocaJon Tablet Assignment Tablet Serving

32 ImplementaJon: Details Tablet LocaJon Tablet Assignment Tablet Serving

33 ImplementaJon: Details Tablet LocaJon Use a three level hierarchy similar to a B + tree to store tablet locajon info. First level contains locajon of root table root tablet contains locajon of all tablets contained in a METADATA table Each METADATA tablet contains locajon of a set of user tables Each METADATA row stores approx. 1KB of data in memory With 128M METADATA tablets, can store 2 61 Bytes Tablet Assignment Tablet Serving

34 ImplementaJon: Details Tablet LocaJon Tablet Assignment Each tablet assigned to one tablet server at a Jme Master keeps track of set of live tablet servers and current assignments When tablet is unassigned, master assigns to tablet server with enough room Uses Chubby to keep track of tablet and acquire/release locks Tablet Serving

35 ImplementaJon: Details Tablet LocaJon Tablet Assignment Tablet Serving Tablet stored in GFS Updates commimed to commit log that stores redo records Recent commits stored in memory in sorted buffer (memtable) Older commits stored in SSTables Read request: Get metadata from METADATA table Metadata contains list of SSTables comprising a tablet + redo points Read SSTables in memory and reconstruct memtable by applying updates since redo points Write request: Check if request correct, sender authorizajons, etc. Write valid mutajons to commit log, memtable

36 1. Data Model 2. Client API 3. Underlying Google Infrastructure 4. BigTable ImplementaJon 5. Refinements 6. Performance 7. Examples 8. Conclusions

37 Refinements The implementajon required a number of refinements to achieve goals. Some are given below: Locality Groups Can group muljple column families together in locality groups Separate SSTable for each locality group in tablet Exploits locality for Read/Write efficiency Compression Clients control compression of SSTables for a locality group Also which compression format is used, opjmized for speed 10 to 1 reducjon can be achieved, compared to 4 to 1 for Gzip on HTML Caching for read performance Two level caching: Scan Cache is higher level, caches key value pairs from SSTables to tablet server Block Cache is lower level, caches SSTable blocks read from GFS Useful for sequenjal reads or random read of different columns in same locality group

38 1. Data Model 2. Client API 3. Underlying Google Infrastructure 4. BigTable ImplementaJon 5. Refinements 6. Performance 7. Examples 8. Conclusions

39 Performance EvaluaJon Setup: BigTable cluster with N tablet servers 1 GB RAM, write to GFS cell consisjng of 1786 machines, 2X 400GB HDDs each N client machines for generajng BigTable load for tests Same number to prevent bomleneck from client side 2X dual core Opteron 2GHz chips, single gigabit Ethernet link GBPS aggregate bandwidth available at the root, 2 level tree All machines in the same hosjng facility, so RTT < 1 msec

40 Performance EvaluaJon Results: Aggregate throughput increases dramajcally, by over a factor of a hundred From 1 to 500 servers Scaling: Does not scale well. Performance does not increase linearly Significant drop in per server throughput when going from 1 50 servers Caused by imbalance in load in muljple server configurajons Processors contending for CPU and network Load balancing algorithm can not do a perfect job Rebalancing is thromled to reduce # of tablet movements Load shits around as test progresses

41 1. Data Model 2. Client API 3. Underlying Google Infrastructure 4. BigTable ImplementaJon 5. Refinements 6. Performance 7. Examples 8. Conclusions

42 Real ApplicaJons Real World Deployment As of August 2006, there were 388 non test BigTable clusters 24,500 tablet servers

43 Real ApplicaJons Google AnalyJcs Google AnalyJcs is a service that helps webmasters analyze traffic pamerns Provides aggregate stajsjcs Unique visitors per day, site tracking reports, page views, etc. Enabled by embedding a small JavaScript program in the web page Program invoked when page visited, records informajon to AnalyJcs User id, info. about the page, etc. AnalyJcs summarizes data to webmasters Uses two tables: Raw Click Table (~200TB) Maintains a row for each end user session Row name is tuple containing website name and Jmestamp Ensures that sessions for same website are conjguous, sorted Compresses to 14% of original size Summary Table (~20TB) Contains predefined summaries for each website Generated from Raw Click Table by periodic MapReduce jobs MapReduce job extracts recent session data from raw click table Compresses to 29% of original size

44 1. Data Model 2. Client API 3. Underlying Google Infrastructure 4. BigTable ImplementaJon 5. Refinements 6. Performance 7. Examples 8. Conclusions

45 Conclusions BigTable Distributed system for storing structured data at Google Clusters have been in producjon use since April 2005 Design and ImplementaJon took 7 person years By August 2006, more than 60 projects using BigTable High Performance, High Availability Unusual interface: difficult to adapt? New users uncertain, parjcularly from relajonal DB background Works well in pracjce though, demonstrated by many projects using it Future work: Support for secondary indices Replicated across data centers, muljple masters

46 Questions?

Bigtable. Presenter: Yijun Hou, Yixiao Peng

Bigtable. Presenter: Yijun Hou, Yixiao Peng Bigtable Fay Chang, Jeffrey Dean, Sanjay Ghemawat, Wilson C. Hsieh, Deborah A. Wallach Mike Burrows, Tushar Chandra, Andrew Fikes, Robert E. Gruber Google, Inc. OSDI 06 Presenter: Yijun Hou, Yixiao Peng

More information

Introduction Data Model API Building Blocks SSTable Implementation Tablet Location Tablet Assingment Tablet Serving Compactions Refinements

Introduction Data Model API Building Blocks SSTable Implementation Tablet Location Tablet Assingment Tablet Serving Compactions Refinements Fay Chang, Jeffrey Dean, Sanjay Ghemawat, Wilson C. Hsieh, Deborah A. Wallach Mike Burrows, Tushar Chandra, Andrew Fikes, Robert E. Gruber Google, Inc. M. Burak ÖZTÜRK 1 Introduction Data Model API Building

More information

Bigtable: A Distributed Storage System for Structured Data. Andrew Hon, Phyllis Lau, Justin Ng

Bigtable: A Distributed Storage System for Structured Data. Andrew Hon, Phyllis Lau, Justin Ng Bigtable: A Distributed Storage System for Structured Data Andrew Hon, Phyllis Lau, Justin Ng What is Bigtable? - A storage system for managing structured data - Used in 60+ Google services - Motivation:

More information

Bigtable: A Distributed Storage System for Structured Data by Google SUNNIE CHUNG CIS 612

Bigtable: A Distributed Storage System for Structured Data by Google SUNNIE CHUNG CIS 612 Bigtable: A Distributed Storage System for Structured Data by Google SUNNIE CHUNG CIS 612 Google Bigtable 2 A distributed storage system for managing structured data that is designed to scale to a very

More information

Big Table. Google s Storage Choice for Structured Data. Presented by Group E - Dawei Yang - Grace Ramamoorthy - Patrick O Sullivan - Rohan Singla

Big Table. Google s Storage Choice for Structured Data. Presented by Group E - Dawei Yang - Grace Ramamoorthy - Patrick O Sullivan - Rohan Singla Big Table Google s Storage Choice for Structured Data Presented by Group E - Dawei Yang - Grace Ramamoorthy - Patrick O Sullivan - Rohan Singla Bigtable: Introduction Resembles a database. Does not support

More information

Bigtable: A Distributed Storage System for Structured Data By Fay Chang, et al. OSDI Presented by Xiang Gao

Bigtable: A Distributed Storage System for Structured Data By Fay Chang, et al. OSDI Presented by Xiang Gao Bigtable: A Distributed Storage System for Structured Data By Fay Chang, et al. OSDI 2006 Presented by Xiang Gao 2014-11-05 Outline Motivation Data Model APIs Building Blocks Implementation Refinement

More information

CSE 444: Database Internals. Lectures 26 NoSQL: Extensible Record Stores

CSE 444: Database Internals. Lectures 26 NoSQL: Extensible Record Stores CSE 444: Database Internals Lectures 26 NoSQL: Extensible Record Stores CSE 444 - Spring 2014 1 References Scalable SQL and NoSQL Data Stores, Rick Cattell, SIGMOD Record, December 2010 (Vol. 39, No. 4)

More information

CS November 2017

CS November 2017 Bigtable Highly available distributed storage Distributed Systems 18. Bigtable Built with semi-structured data in mind URLs: content, metadata, links, anchors, page rank User data: preferences, account

More information

Distributed Systems [Fall 2012]

Distributed Systems [Fall 2012] Distributed Systems [Fall 2012] Lec 20: Bigtable (cont ed) Slide acks: Mohsen Taheriyan (http://www-scf.usc.edu/~csci572/2011spring/presentations/taheriyan.pptx) 1 Chubby (Reminder) Lock service with a

More information

Bigtable. A Distributed Storage System for Structured Data. Presenter: Yunming Zhang Conglong Li. Saturday, September 21, 13

Bigtable. A Distributed Storage System for Structured Data. Presenter: Yunming Zhang Conglong Li. Saturday, September 21, 13 Bigtable A Distributed Storage System for Structured Data Presenter: Yunming Zhang Conglong Li References SOCC 2010 Key Note Slides Jeff Dean Google Introduction to Distributed Computing, Winter 2008 University

More information

CS November 2018

CS November 2018 Bigtable Highly available distributed storage Distributed Systems 19. Bigtable Built with semi-structured data in mind URLs: content, metadata, links, anchors, page rank User data: preferences, account

More information

CSE-E5430 Scalable Cloud Computing Lecture 9

CSE-E5430 Scalable Cloud Computing Lecture 9 CSE-E5430 Scalable Cloud Computing Lecture 9 Keijo Heljanko Department of Computer Science School of Science Aalto University keijo.heljanko@aalto.fi 15.11-2015 1/24 BigTable Described in the paper: Fay

More information

ΕΠΛ 602:Foundations of Internet Technologies. Cloud Computing

ΕΠΛ 602:Foundations of Internet Technologies. Cloud Computing ΕΠΛ 602:Foundations of Internet Technologies Cloud Computing 1 Outline Bigtable(data component of cloud) Web search basedonch13of thewebdatabook 2 What is Cloud Computing? ACloudis an infrastructure, transparent

More information

References. What is Bigtable? Bigtable Data Model. Outline. Key Features. CSE 444: Database Internals

References. What is Bigtable? Bigtable Data Model. Outline. Key Features. CSE 444: Database Internals References CSE 444: Database Internals Scalable SQL and NoSQL Data Stores, Rick Cattell, SIGMOD Record, December 2010 (Vol 39, No 4) Lectures 26 NoSQL: Extensible Record Stores Bigtable: A Distributed

More information

BigTable. CSE-291 (Cloud Computing) Fall 2016

BigTable. CSE-291 (Cloud Computing) Fall 2016 BigTable CSE-291 (Cloud Computing) Fall 2016 Data Model Sparse, distributed persistent, multi-dimensional sorted map Indexed by a row key, column key, and timestamp Values are uninterpreted arrays of bytes

More information

Bigtable: A Distributed Storage System for Structured Data

Bigtable: A Distributed Storage System for Structured Data 4 Bigtable: A Distributed Storage System for Structured Data FAY CHANG, JEFFREY DEAN, SANJAY GHEMAWAT, WILSON C. HSIEH, DEBORAH A. WALLACH, MIKE BURROWS, TUSHAR CHANDRA, ANDREW FIKES, and ROBERT E. GRUBER

More information

Bigtable: A Distributed Storage System for Structured Data

Bigtable: A Distributed Storage System for Structured Data Bigtable: A Distributed Storage System for Structured Data Fay Chang, Jeffrey Dean, Sanjay Ghemawat, Wilson C. Hsieh, Deborah A. Wallach Mike Burrows, Tushar Chandra, Andrew Fikes, Robert E. Gruber {fay,jeff,sanjay,wilsonh,kerr,m3b,tushar,fikes,gruber}@google.com

More information

Big Data Infrastructure CS 489/698 Big Data Infrastructure (Winter 2017)

Big Data Infrastructure CS 489/698 Big Data Infrastructure (Winter 2017) Big Data Infrastructure CS 489/698 Big Data Infrastructure (Winter 2017) Week 10: Mutable State (1/2) March 14, 2017 Jimmy Lin David R. Cheriton School of Computer Science University of Waterloo These

More information

BigTable: A Distributed Storage System for Structured Data (2006) Slides adapted by Tyler Davis

BigTable: A Distributed Storage System for Structured Data (2006) Slides adapted by Tyler Davis BigTable: A Distributed Storage System for Structured Data (2006) Slides adapted by Tyler Davis Motivation Lots of (semi-)structured data at Google URLs: Contents, crawl metadata, links, anchors, pagerank,

More information

Structured Big Data 1: Google Bigtable & HBase Shiow-yang Wu ( 吳秀陽 ) CSIE, NDHU, Taiwan, ROC

Structured Big Data 1: Google Bigtable & HBase Shiow-yang Wu ( 吳秀陽 ) CSIE, NDHU, Taiwan, ROC Structured Big Data 1: Google Bigtable & HBase Shiow-yang Wu ( 吳秀陽 ) CSIE, NDHU, Taiwan, ROC Lecture material is mostly home-grown, partly taken with permission and courtesy from Professor Shih-Wei Liao

More information

BigTable A System for Distributed Structured Storage

BigTable A System for Distributed Structured Storage BigTable A System for Distributed Structured Storage Fay Chang, Jeffrey Dean, Sanjay Ghemawat, Wilson C. Hsieh, Deborah A. Wallach, Mike Burrows, Tushar Chandra, Andrew Fikes, and Robert E. Gruber Adapted

More information

CSE 544 Principles of Database Management Systems. Magdalena Balazinska Winter 2009 Lecture 12 Google Bigtable

CSE 544 Principles of Database Management Systems. Magdalena Balazinska Winter 2009 Lecture 12 Google Bigtable CSE 544 Principles of Database Management Systems Magdalena Balazinska Winter 2009 Lecture 12 Google Bigtable References Bigtable: A Distributed Storage System for Structured Data. Fay Chang et. al. OSDI

More information

Design & Implementation of Cloud Big table

Design & Implementation of Cloud Big table Design & Implementation of Cloud Big table M.Swathi 1,A.Sujitha 2, G.Sai Sudha 3, T.Swathi 4 M.Swathi Assistant Professor in Department of CSE Sri indu College of Engineering &Technolohy,Sheriguda,Ibrahimptnam

More information

BigTable: A Distributed Storage System for Structured Data

BigTable: A Distributed Storage System for Structured Data BigTable: A Distributed Storage System for Structured Data Amir H. Payberah amir@sics.se Amirkabir University of Technology (Tehran Polytechnic) Amir H. Payberah (Tehran Polytechnic) BigTable 1393/7/26

More information

Distributed File Systems II

Distributed File Systems II Distributed File Systems II To do q Very-large scale: Google FS, Hadoop FS, BigTable q Next time: Naming things GFS A radically new environment NFS, etc. Independence Small Scale Variety of workloads Cooperation

More information

Big Data Infrastructure CS 489/698 Big Data Infrastructure (Winter 2016)

Big Data Infrastructure CS 489/698 Big Data Infrastructure (Winter 2016) Big Data Infrastructure CS 489/698 Big Data Infrastructure (Winter 2016) Week 10: Mutable State (1/2) March 15, 2016 Jimmy Lin David R. Cheriton School of Computer Science University of Waterloo These

More information

BigTable: A System for Distributed Structured Storage

BigTable: A System for Distributed Structured Storage BigTable: A System for Distributed Structured Storage Jeff Dean Joint work with: Mike Burrows, Tushar Chandra, Fay Chang, Mike Epstein, Andrew Fikes, Sanjay Ghemawat, Robert Griesemer, Bob Gruber, Wilson

More information

Distributed Database Case Study on Google s Big Tables

Distributed Database Case Study on Google s Big Tables Distributed Database Case Study on Google s Big Tables Anjali diwakar dwivedi 1, Usha sadanand patil 2 and Vinayak D.Shinde 3 1,2,3 Computer Engineering, Shree l.r.tiwari college of engineering Abstract-

More information

CA485 Ray Walshe NoSQL

CA485 Ray Walshe NoSQL NoSQL BASE vs ACID Summary Traditional relational database management systems (RDBMS) do not scale because they adhere to ACID. A strong movement within cloud computing is to utilize non-traditional data

More information

BigTable. Chubby. BigTable. Chubby. Why Chubby? How to do consensus as a service

BigTable. Chubby. BigTable. Chubby. Why Chubby? How to do consensus as a service BigTable BigTable Doug Woos and Tom Anderson In the early 2000s, Google had way more than anybody else did Traditional bases couldn t scale Want something better than a filesystem () BigTable optimized

More information

big picture parallel db (one data center) mix of OLTP and batch analysis lots of data, high r/w rates, 1000s of cheap boxes thus many failures

big picture parallel db (one data center) mix of OLTP and batch analysis lots of data, high r/w rates, 1000s of cheap boxes thus many failures Lecture 20 -- 11/20/2017 BigTable big picture parallel db (one data center) mix of OLTP and batch analysis lots of data, high r/w rates, 1000s of cheap boxes thus many failures what does paper say Google

More information

Lecture: The Google Bigtable

Lecture: The Google Bigtable Lecture: The Google Bigtable h#p://research.google.com/archive/bigtable.html 10/09/2014 Romain Jaco3n romain.jaco7n@orange.fr Agenda Introduc3on Data model API Building blocks Implementa7on Refinements

More information

Lessons Learned While Building Infrastructure Software at Google

Lessons Learned While Building Infrastructure Software at Google Lessons Learned While Building Infrastructure Software at Google Jeff Dean jeff@google.com Google Circa 1997 (google.stanford.edu) Corkboards (1999) Google Data Center (2000) Google Data Center (2000)

More information

Extreme Computing. NoSQL.

Extreme Computing. NoSQL. Extreme Computing NoSQL PREVIOUSLY: BATCH Query most/all data Results Eventually NOW: ON DEMAND Single Data Points Latency Matters One problem, three ideas We want to keep track of mutable state in a scalable

More information

7680: Distributed Systems

7680: Distributed Systems Cristina Nita-Rotaru 7680: Distributed Systems BigTable. Hbase.Spanner. 1: BigTable Acknowledgement } Slides based on material from course at UMichigan, U Washington, and the authors of BigTable and Spanner.

More information

Outline. Spanner Mo/va/on. Tom Anderson

Outline. Spanner Mo/va/on. Tom Anderson Spanner Mo/va/on Tom Anderson Outline Last week: Chubby: coordina/on service BigTable: scalable storage of structured data GFS: large- scale storage for bulk data Today/Friday: Lessons from GFS/BigTable

More information

Google File System and BigTable. and tiny bits of HDFS (Hadoop File System) and Chubby. Not in textbook; additional information

Google File System and BigTable. and tiny bits of HDFS (Hadoop File System) and Chubby. Not in textbook; additional information Subject 10 Fall 2015 Google File System and BigTable and tiny bits of HDFS (Hadoop File System) and Chubby Not in textbook; additional information Disclaimer: These abbreviated notes DO NOT substitute

More information

MapReduce & BigTable

MapReduce & BigTable CPSC 426/526 MapReduce & BigTable Ennan Zhai Computer Science Department Yale University Lecture Roadmap Cloud Computing Overview Challenges in the Clouds Distributed File Systems: GFS Data Process & Analysis:

More information

18-hdfs-gfs.txt Thu Oct 27 10:05: Notes on Parallel File Systems: HDFS & GFS , Fall 2011 Carnegie Mellon University Randal E.

18-hdfs-gfs.txt Thu Oct 27 10:05: Notes on Parallel File Systems: HDFS & GFS , Fall 2011 Carnegie Mellon University Randal E. 18-hdfs-gfs.txt Thu Oct 27 10:05:07 2011 1 Notes on Parallel File Systems: HDFS & GFS 15-440, Fall 2011 Carnegie Mellon University Randal E. Bryant References: Ghemawat, Gobioff, Leung, "The Google File

More information

Google Data Management

Google Data Management Google Data Management Vera Goebel Department of Informatics, University of Oslo 2009 Google Technology Kaizan: continuous developments and improvements Grid computing: Google data centers and messages

More information

W b b 2.0. = = Data Ex E pl p o l s o io i n

W b b 2.0. = = Data Ex E pl p o l s o io i n Hypertable Doug Judd Zvents, Inc. Background Web 2.0 = Data Explosion Web 2.0 Mt. Web 2.0 Traditional Tools Don t Scale Well Designed for a single machine Typical scaling solutions ad-hoc manual/static

More information

Cloud Programming. Programming Environment Oct 29, 2015 Osamu Tatebe

Cloud Programming. Programming Environment Oct 29, 2015 Osamu Tatebe Cloud Programming Programming Environment Oct 29, 2015 Osamu Tatebe Cloud Computing Only required amount of CPU and storage can be used anytime from anywhere via network Availability, throughput, reliability

More information

Programming model and implementation for processing and. Programs can be automatically parallelized and executed on a large cluster of machines

Programming model and implementation for processing and. Programs can be automatically parallelized and executed on a large cluster of machines A programming model in Cloud: MapReduce Programming model and implementation for processing and generating large data sets Users specify a map function to generate a set of intermediate key/value pairs

More information

The Google File System

The Google File System The Google File System Sanjay Ghemawat, Howard Gobioff, and Shun-Tak Leung Google SOSP 03, October 19 22, 2003, New York, USA Hyeon-Gyu Lee, and Yeong-Jae Woo Memory & Storage Architecture Lab. School

More information

18-hdfs-gfs.txt Thu Nov 01 09:53: Notes on Parallel File Systems: HDFS & GFS , Fall 2012 Carnegie Mellon University Randal E.

18-hdfs-gfs.txt Thu Nov 01 09:53: Notes on Parallel File Systems: HDFS & GFS , Fall 2012 Carnegie Mellon University Randal E. 18-hdfs-gfs.txt Thu Nov 01 09:53:32 2012 1 Notes on Parallel File Systems: HDFS & GFS 15-440, Fall 2012 Carnegie Mellon University Randal E. Bryant References: Ghemawat, Gobioff, Leung, "The Google File

More information

The Google File System

The Google File System The Google File System Sanjay Ghemawat, Howard Gobioff, and Shun-Tak Leung SOSP 2003 presented by Kun Suo Outline GFS Background, Concepts and Key words Example of GFS Operations Some optimizations in

More information

Infrastructure system services

Infrastructure system services Infrastructure system services Badri Nath Rutgers University badri@cs.rutgers.edu Processing lots of data O(B) web pages; each O(K) bytes to O(M) bytes gives you O(T) to O(P) bytes of data Disk Bandwidth

More information

Google File System. Sanjay Ghemawat, Howard Gobioff, and Shun-Tak Leung Google fall DIP Heerak lim, Donghun Koo

Google File System. Sanjay Ghemawat, Howard Gobioff, and Shun-Tak Leung Google fall DIP Heerak lim, Donghun Koo Google File System Sanjay Ghemawat, Howard Gobioff, and Shun-Tak Leung Google 2017 fall DIP Heerak lim, Donghun Koo 1 Agenda Introduction Design overview Systems interactions Master operation Fault tolerance

More information

PoS(ISGC 2011 & OGF 31)075

PoS(ISGC 2011 & OGF 31)075 1 GIS Center, Feng Chia University 100 Wen0Hwa Rd., Taichung, Taiwan E-mail: cool@gis.tw Tien-Yin Chou GIS Center, Feng Chia University 100 Wen0Hwa Rd., Taichung, Taiwan E-mail: jimmy@gis.tw Lan-Kun Chung

More information

Distributed computing: index building and use

Distributed computing: index building and use Distributed computing: index building and use Distributed computing Goals Distributing computation across several machines to Do one computation faster - latency Do more computations in given time - throughput

More information

Data Storage in the Cloud

Data Storage in the Cloud Data Storage in the Cloud KHALID ELGAZZAR GOODWIN 531 ELGAZZAR@CS.QUEENSU.CA Outline 1. Distributed File Systems 1.1. Google File System (GFS) 2. NoSQL Data Store 2.1. BigTable Elgazzar - CISC 886 - Fall

More information

CA485 Ray Walshe Google File System

CA485 Ray Walshe Google File System Google File System Overview Google File System is scalable, distributed file system on inexpensive commodity hardware that provides: Fault Tolerance File system runs on hundreds or thousands of storage

More information

A Cloud Storage Adaptable to Read-Intensive and Write-Intensive Workload

A Cloud Storage Adaptable to Read-Intensive and Write-Intensive Workload DEIM Forum 2011 C3-3 152-8552 2-12-1 E-mail: {nakamur6,shudo}@is.titech.ac.jp.,., MyCassandra, Cassandra MySQL, 41.4%, 49.4%.,, Abstract A Cloud Storage Adaptable to Read-Intensive and Write-Intensive

More information

Distributed Data Management. Christoph Lofi Institut für Informationssysteme Technische Universität Braunschweig

Distributed Data Management. Christoph Lofi Institut für Informationssysteme Technische Universität Braunschweig Distributed Data Management Christoph Lofi Institut für Informationssysteme Technische Universität Braunschweig http://www.ifis.cs.tu-bs.de Exams 25 minutes oral examination 20.-24.02.2012 19.-23.03.2012

More information

11 Storage at Google Google Google Google Google 7/2/2010. Distributed Data Management

11 Storage at Google Google Google Google Google 7/2/2010. Distributed Data Management 11 Storage at Google Distributed Data Management 11.1 Google Bigtable 11.2 Google File System 11. Bigtable Implementation Wolf-Tilo Balke Christoph Lofi Institut für Informationssysteme Technische Universität

More information

Distributed Filesystem

Distributed Filesystem Distributed Filesystem 1 How do we get data to the workers? NAS Compute Nodes SAN 2 Distributing Code! Don t move data to workers move workers to the data! - Store data on the local disks of nodes in the

More information

CS /29/18. Paul Krzyzanowski 1. Question 1 (Bigtable) Distributed Systems 2018 Pre-exam 3 review Selected questions from past exams

CS /29/18. Paul Krzyzanowski 1. Question 1 (Bigtable) Distributed Systems 2018 Pre-exam 3 review Selected questions from past exams Question 1 (Bigtable) What is an SSTable in Bigtable? Distributed Systems 2018 Pre-exam 3 review Selected questions from past exams It is the internal file format used to store Bigtable data. It maps keys

More information

CS5412: DIVING IN: INSIDE THE DATA CENTER

CS5412: DIVING IN: INSIDE THE DATA CENTER 1 CS5412: DIVING IN: INSIDE THE DATA CENTER Lecture V Ken Birman We ve seen one cloud service 2 Inside a cloud, Dynamo is an example of a service used to make sure that cloud-hosted applications can scale

More information

Data Informatics. Seon Ho Kim, Ph.D.

Data Informatics. Seon Ho Kim, Ph.D. Data Informatics Seon Ho Kim, Ph.D. seonkim@usc.edu HBase HBase is.. A distributed data store that can scale horizontally to 1,000s of commodity servers and petabytes of indexed storage. Designed to operate

More information

Distributed Systems Pre-exam 3 review Selected questions from past exams. David Domingo Paul Krzyzanowski Rutgers University Fall 2018

Distributed Systems Pre-exam 3 review Selected questions from past exams. David Domingo Paul Krzyzanowski Rutgers University Fall 2018 Distributed Systems 2018 Pre-exam 3 review Selected questions from past exams David Domingo Paul Krzyzanowski Rutgers University Fall 2018 November 28, 2018 1 Question 1 (Bigtable) What is an SSTable in

More information

Islands of Splunk. MulJple Splunk as a Service Architecture and ImplementaJon

Islands of Splunk. MulJple Splunk as a Service Architecture and ImplementaJon Copyright 2014 Splunk Inc. Islands of Splunk MulJple Splunk as a Service Architecture and ImplementaJon Michael de Buin, Schuberg Philis mdebruin@schubergphilis.com Gert Kremer, Schuberg Philis gkremer@schubergphilis.com

More information

CS5412: OTHER DATA CENTER SERVICES

CS5412: OTHER DATA CENTER SERVICES 1 CS5412: OTHER DATA CENTER SERVICES Lecture V Ken Birman Tier two and Inner Tiers 2 If tier one faces the user and constructs responses, what lives in tier two? Caching services are very common (many

More information

Cassandra- A Distributed Database

Cassandra- A Distributed Database Cassandra- A Distributed Database Tulika Gupta Department of Information Technology Poornima Institute of Engineering and Technology Jaipur, Rajasthan, India Abstract- A relational database is a traditional

More information

The MapReduce Abstraction

The MapReduce Abstraction The MapReduce Abstraction Parallel Computing at Google Leverages multiple technologies to simplify large-scale parallel computations Proprietary computing clusters Map/Reduce software library Lots of other

More information

CS5412: DIVING IN: INSIDE THE DATA CENTER

CS5412: DIVING IN: INSIDE THE DATA CENTER 1 CS5412: DIVING IN: INSIDE THE DATA CENTER Lecture V Ken Birman Data centers 2 Once traffic reaches a data center it tunnels in First passes through a filter that blocks attacks Next, a router that directs

More information

The Google File System

The Google File System The Google File System Sanjay Ghemawat, Howard Gobioff, and Shun-Tak Leung December 2003 ACM symposium on Operating systems principles Publisher: ACM Nov. 26, 2008 OUTLINE INTRODUCTION DESIGN OVERVIEW

More information

Big Data Management and NoSQL Databases

Big Data Management and NoSQL Databases NDBI040 Big Data Management and NoSQL Databases Lecture 2. MapReduce Doc. RNDr. Irena Holubova, Ph.D. holubova@ksi.mff.cuni.cz http://www.ksi.mff.cuni.cz/~holubova/ndbi040/ Framework A programming model

More information

CS 345A Data Mining. MapReduce

CS 345A Data Mining. MapReduce CS 345A Data Mining MapReduce Single-node architecture CPU Machine Learning, Statistics Memory Classical Data Mining Disk Commodity Clusters Web data sets can be very large Tens to hundreds of terabytes

More information

The Google File System

The Google File System The Google File System Sanjay Ghemawat, Howard Gobioff, and Shun-Tak Leung Google* 정학수, 최주영 1 Outline Introduction Design Overview System Interactions Master Operation Fault Tolerance and Diagnosis Conclusions

More information

Distributed Systems. 05r. Case study: Google Cluster Architecture. Paul Krzyzanowski. Rutgers University. Fall 2016

Distributed Systems. 05r. Case study: Google Cluster Architecture. Paul Krzyzanowski. Rutgers University. Fall 2016 Distributed Systems 05r. Case study: Google Cluster Architecture Paul Krzyzanowski Rutgers University Fall 2016 1 A note about relevancy This describes the Google search cluster architecture in the mid

More information

CSE 124: Networked Services Lecture-16

CSE 124: Networked Services Lecture-16 Fall 2010 CSE 124: Networked Services Lecture-16 Instructor: B. S. Manoj, Ph.D http://cseweb.ucsd.edu/classes/fa10/cse124 11/23/2010 CSE 124 Networked Services Fall 2010 1 Updates PlanetLab experiments

More information

Yuval Carmel Tel-Aviv University "Advanced Topics in Storage Systems" - Spring 2013

Yuval Carmel Tel-Aviv University Advanced Topics in Storage Systems - Spring 2013 Yuval Carmel Tel-Aviv University "Advanced Topics in About & Keywords Motivation & Purpose Assumptions Architecture overview & Comparison Measurements How does it fit in? The Future 2 About & Keywords

More information

Authors : Sanjay Ghemawat, Howard Gobioff, Shun-Tak Leung Presentation by: Vijay Kumar Chalasani

Authors : Sanjay Ghemawat, Howard Gobioff, Shun-Tak Leung Presentation by: Vijay Kumar Chalasani The Authors : Sanjay Ghemawat, Howard Gobioff, Shun-Tak Leung Presentation by: Vijay Kumar Chalasani CS5204 Operating Systems 1 Introduction GFS is a scalable distributed file system for large data intensive

More information

The Chubby lock service for loosely- coupled distributed systems

The Chubby lock service for loosely- coupled distributed systems The Chubby lock service for loosely- coupled distributed systems Mike Burrows, Google Inc. Presented By- Rahul Malviya Outline Paxos Algorithm The Chubby Lock Service What is Paxos Help to implemenjng

More information

MapReduce. U of Toronto, 2014

MapReduce. U of Toronto, 2014 MapReduce U of Toronto, 2014 http://www.google.org/flutrends/ca/ (2012) Average Searches Per Day: 5,134,000,000 2 Motivation Process lots of data Google processed about 24 petabytes of data per day in

More information

Parallel Computing: MapReduce Jin, Hai

Parallel Computing: MapReduce Jin, Hai Parallel Computing: MapReduce Jin, Hai School of Computer Science and Technology Huazhong University of Science and Technology ! MapReduce is a distributed/parallel computing framework introduced by Google

More information

Distributed computing: index building and use

Distributed computing: index building and use Distributed computing: index building and use Distributed computing Goals Distributing computation across several machines to Do one computation faster - latency Do more computations in given time - throughput

More information

FAQs Snapshots and locks Vector Clock

FAQs Snapshots and locks Vector Clock //08 CS5 Introduction to Big - FALL 08 W.B.0.0 CS5 Introduction to Big //08 CS5 Introduction to Big - FALL 08 W.B. FAQs Snapshots and locks Vector Clock PART. LARGE SCALE DATA STORAGE SYSTEMS NO SQL DATA

More information

CLOUD-SCALE FILE SYSTEMS

CLOUD-SCALE FILE SYSTEMS Data Management in the Cloud CLOUD-SCALE FILE SYSTEMS 92 Google File System (GFS) Designing a file system for the Cloud design assumptions design choices Architecture GFS Master GFS Chunkservers GFS Clients

More information

Crossing the Chasm: Sneaking a parallel file system into Hadoop

Crossing the Chasm: Sneaking a parallel file system into Hadoop Crossing the Chasm: Sneaking a parallel file system into Hadoop Wittawat Tantisiriroj Swapnil Patil, Garth Gibson PARALLEL DATA LABORATORY Carnegie Mellon University In this work Compare and contrast large

More information

Distributed Systems 16. Distributed File Systems II

Distributed Systems 16. Distributed File Systems II Distributed Systems 16. Distributed File Systems II Paul Krzyzanowski pxk@cs.rutgers.edu 1 Review NFS RPC-based access AFS Long-term caching CODA Read/write replication & disconnected operation DFS AFS

More information

CSE 124: Networked Services Fall 2009 Lecture-19

CSE 124: Networked Services Fall 2009 Lecture-19 CSE 124: Networked Services Fall 2009 Lecture-19 Instructor: B. S. Manoj, Ph.D http://cseweb.ucsd.edu/classes/fa09/cse124 Some of these slides are adapted from various sources/individuals including but

More information

Google big data techniques (2)

Google big data techniques (2) Google big data techniques (2) Lecturer: Jiaheng Lu Fall 2016 10.12.2016 1 Outline Google File System and HDFS Relational DB V.S. Big data system Google Bigtable and NoSQL databases 2016/12/10 3 The Google

More information

! Design constraints. " Component failures are the norm. " Files are huge by traditional standards. ! POSIX-like

! Design constraints.  Component failures are the norm.  Files are huge by traditional standards. ! POSIX-like Cloud background Google File System! Warehouse scale systems " 10K-100K nodes " 50MW (1 MW = 1,000 houses) " Power efficient! Located near cheap power! Passive cooling! Power Usage Effectiveness = Total

More information

Demand- Paged Virtual Memory

Demand- Paged Virtual Memory Demand- Paged Virtual Memory Main Points Can we provide the illusion of near infinite memory in limited physical memory? Demand- paged virtual memory Memory- mapped files How do we choose which page to

More information

The Google File System

The Google File System The Google File System Sanjay Ghemawat, Howard Gobioff and Shun Tak Leung Google* Shivesh Kumar Sharma fl4164@wayne.edu Fall 2015 004395771 Overview Google file system is a scalable distributed file system

More information

Indexing Large-Scale Data

Indexing Large-Scale Data Indexing Large-Scale Data Serge Abiteboul Ioana Manolescu Philippe Rigaux Marie-Christine Rousset Pierre Senellart Web Data Management and Distribution http://webdam.inria.fr/textbook November 16, 2010

More information

Scaling Up HBase. Duen Horng (Polo) Chau Assistant Professor Associate Director, MS Analytics Georgia Tech. CSE6242 / CX4242: Data & Visual Analytics

Scaling Up HBase. Duen Horng (Polo) Chau Assistant Professor Associate Director, MS Analytics Georgia Tech. CSE6242 / CX4242: Data & Visual Analytics http://poloclub.gatech.edu/cse6242 CSE6242 / CX4242: Data & Visual Analytics Scaling Up HBase Duen Horng (Polo) Chau Assistant Professor Associate Director, MS Analytics Georgia Tech Partly based on materials

More information

BigData and Map Reduce VITMAC03

BigData and Map Reduce VITMAC03 BigData and Map Reduce VITMAC03 1 Motivation Process lots of data Google processed about 24 petabytes of data per day in 2009. A single machine cannot serve all the data You need a distributed system to

More information

The Google File System

The Google File System October 13, 2010 Based on: S. Ghemawat, H. Gobioff, and S.-T. Leung: The Google file system, in Proceedings ACM SOSP 2003, Lake George, NY, USA, October 2003. 1 Assumptions Interface Architecture Single

More information

A Fast and High Throughput SQL Query System for Big Data

A Fast and High Throughput SQL Query System for Big Data A Fast and High Throughput SQL Query System for Big Data Feng Zhu, Jie Liu, and Lijie Xu Technology Center of Software Engineering, Institute of Software, Chinese Academy of Sciences, Beijing, China 100190

More information

Percolator. Large-Scale Incremental Processing using Distributed Transactions and Notifications. D. Peng & F. Dabek

Percolator. Large-Scale Incremental Processing using Distributed Transactions and Notifications. D. Peng & F. Dabek Percolator Large-Scale Incremental Processing using Distributed Transactions and Notifications D. Peng & F. Dabek Motivation Built to maintain the Google web search index Need to maintain a large repository,

More information

DIVING IN: INSIDE THE DATA CENTER

DIVING IN: INSIDE THE DATA CENTER 1 DIVING IN: INSIDE THE DATA CENTER Anwar Alhenshiri Data centers 2 Once traffic reaches a data center it tunnels in First passes through a filter that blocks attacks Next, a router that directs it to

More information

NPTEL Course Jan K. Gopinath Indian Institute of Science

NPTEL Course Jan K. Gopinath Indian Institute of Science Storage Systems NPTEL Course Jan 2012 (Lecture 39) K. Gopinath Indian Institute of Science Google File System Non-Posix scalable distr file system for large distr dataintensive applications performance,

More information

Staggeringly Large Filesystems

Staggeringly Large Filesystems Staggeringly Large Filesystems Evan Danaher CS 6410 - October 27, 2009 Outline 1 Large Filesystems 2 GFS 3 Pond Outline 1 Large Filesystems 2 GFS 3 Pond Internet Scale Web 2.0 GFS Thousands of machines

More information

The Google File System

The Google File System The Google File System By Ghemawat, Gobioff and Leung Outline Overview Assumption Design of GFS System Interactions Master Operations Fault Tolerance Measurements Overview GFS: Scalable distributed file

More information

Google is Really Different.

Google is Really Different. COMP 790-088 -- Distributed File Systems Google File System 7 Google is Really Different. Huge Datacenters in 5+ Worldwide Locations Datacenters house multiple server clusters Coming soon to Lenior, NC

More information

Big Data Processing Technologies. Chentao Wu Associate Professor Dept. of Computer Science and Engineering

Big Data Processing Technologies. Chentao Wu Associate Professor Dept. of Computer Science and Engineering Big Data Processing Technologies Chentao Wu Associate Professor Dept. of Computer Science and Engineering wuct@cs.sjtu.edu.cn Schedule (1) Storage system part (first eight weeks) lec1: Introduction on

More information

Federated Array of Bricks Y Saito et al HP Labs. CS 6464 Presented by Avinash Kulkarni

Federated Array of Bricks Y Saito et al HP Labs. CS 6464 Presented by Avinash Kulkarni Federated Array of Bricks Y Saito et al HP Labs CS 6464 Presented by Avinash Kulkarni Agenda Motivation Current Approaches FAB Design Protocols, Implementation, Optimizations Evaluation SSDs in enterprise

More information

The Google File System. Alexandru Costan

The Google File System. Alexandru Costan 1 The Google File System Alexandru Costan Actions on Big Data 2 Storage Analysis Acquisition Handling the data stream Data structured unstructured semi-structured Results Transactions Outline File systems

More information