MongoDB, Hadoop and humongous data at MongoSV 2012

This presentation given at MongoSV 2012 focuses on data processing when using MongoDB as your primary database including integration with Hadoop & the new MongoDB aggregation framework. Learn how to integrate MongoDB with Hadoop for large-scale distributed data processing. Using tools like MapReduce, Pig and Streaming you will learn how to do analytics and ETL on large datasets with the ability to load and save data against MongoDB. With Hadoop MapReduce, Java and Scala programmers will find a native solution for using MapReduce to process their data with MongoDB. Programmers of all kinds will find a new way to work with ETL using Pig to extract and analyze large datasets and persist the results to MongoDB. Python and Ruby Programmers can rejoice as well in a new way to write native Mongo MapReduce using the Hadoop Streaming interfaces.

MongoDB, Hadoop and humongous data – MongoSV 2012 from Steve Francia

MongoDB, Hadoop and humongous data – MongoSV 2012 — Presentation Transcript

  1. MongoDB, Hadoop & humongous data
  2. Talking about. What is Humongous Data. Humongous Data & You. MongoDB & Data processing. Future of Humongous Data
  3. @spf13 AKA Steve Francia. 15+ years building the internet. Father, husband, skateboarder. Chief Solutions Architect @ 10gen responsible for drivers, integrations, web & docs
  4. What is humongous data ?
  5. 2000 Google Inc. Today announced it has released the largest search engine on the Internet.Google’s new index, comprisingmore than 1 billion URLs
  6. 2008 Our indexing system for processing links indicates that we now count 1 trillion unique URLs (and the number of individual webpages out there is growing by several billion pages per day).
  7. An unprecedented amount of data is being created and isaccessible
  8. Data Growth 1,0001000 750 500 500 250 250 120 55 4 10 24 1 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 Millions of URLs
  9. Truly Exponential Growth Is hard for people to grasp A BBC reporter recently: “Your current PCis more powerful than the computer they had on board the first flight to the moon”.
  10. Moore’s Law Applies to more than just CPUs Boiled down it is that things double at regular intervals. It’s exponential growth.. and applies to big data
  11. How BIG is it?
  12. How BIG is it?2008
  13. How BIG is it? 20072008 2005 2006 2003 2004 2001 2002
  14. Why all this talk about BIG Data now?
  15. In the past few years open source software emerged enabling ‘us’ to handle BIG Data
  16. The Big Data Story
  17. Is actually two stories
  18. Doers & Tellers talking about different things http://www.slideshare.net/siliconangle/trendconnect-big-data-report-september
  19. Tellers
  20. Doers
  21. Doers talk a lot more about actual solutions
  22. They know it’s a two sided story Storage Processing
  23. Take aways MongoDB and Hadoop MongoDB for storage & operations Hadoop for processing & analytics
  24. MongoDB & Data Processing
  25. Applications have complex needs. MongoDB ideal operational database MongoDB ideal for BIG data. Not a data processing engine, but provides processing functionality
  26. Many options for Processing Data • Process in MongoDB using Map Reduce • Process in MongoDB using Aggregation Framework • Process outside MongoDB (using Hadoop)
  27. MongoDB Map Reduce Map() MongoDB Data Group(k) emit(k,v) map iterates on documents Document is $this Sort(k) 1 at time per shard Reduce(k,values) k,v Finalize(k,v) Input matches output k,v Can run multiple times
  28. MongoDB Map Reduce MongoDB map reduce quite capable… but with limits- Javascript not best language for processing map reduce- Javascript limited in external data processing libraries- Adds load to data store
  29. MongoDB Aggregation Most uses of MongoDB Map Reduce were for aggregationAggregation Framework optimized for aggregate queriesRealtime aggregation similar to SQL GroupBy
  30. MongoDB & Hadoop same as Mongos Many map operationsMongoDB shard chunks (64mb) 1 at time per input split Creates a list each split Map (k1,1v1,1ctx) Runs on same of Input Splits Map (k ,1v ,1ctx) thread as map each split Map (k , v , ctx)single server orsharded cluster (InputFormat) each split ctx.write(k2,v2)2 ctx.write(k2,v )2 Combiner(k2,values2)2 RecordReader ctx.write(k2,v ) Combiner(k2,values )2 Combiner(k2,values ) k2, 2v3 3 k , 2v 3 k ,v Partitioner(k2)2 Partitioner(k )2 Partitioner(k ) Sort(keys2) Sort(k2)2 Sort(k )MongoDB Reducer threads Reduce(k2,values3) Output Format Runs once per key kf,vf
  31. DEMOTIME
  32. DEMO Install Hadoop MongoDB Plugin Import tweets from twitter Write mapper in Python using Hadoop streamingWrite reducer in Python using Hadoop streaming Call myself a data scientist
  33. Installing Mongo-hadoop https://gist.github.com/1887726hadoop_version 0.23 hadoop_path=”/usr/local/Cellar/hadoop/$hadoop_version.0/libexec/lib”git clone git://github.com/mongodb/mongo-hadoop.gitcd mongo-hadoopsed -i “s/default/$hadoop_version/g” build.sbtcd streaming./build.sh
  34. Groking Twitter curl https://stream.twitter.com/1/statuses/sample.json -u<login>:<password> | mongoimport -d test -c live … let it run for about 2 hours
  35. DEMO 1
  36. Map Hashtags in Python#!/usr/bin/env pythonimport syssys.path.append(“.”)from pymongo_hadoop import BSONMapperdef mapper(documents): for doc in documents: for hashtag in doc[entities][hashtags]: yield {_id: hashtag[text], count: 1}BSONMapper(mapper)print >> sys.stderr, “Done Mapping.”
  37. Reduce hashtags in Python#!/usr/bin/env pythonimport syssys.path.append(“.”)from pymongo_hadoop import BSONReducerdef reducer(key, values): print >> sys.stderr, “Hashtag %s” % key.encode(utf8) _count = 0 for v in values: _count += v[count] return {_id: key.encode(utf8), count: _count}BSONReducer(reducer)
  38. All together hadoop jar target/mongo-hadoop-streaming-assembly-1.0.0-rc0.jar -mapper examples/twitter/twit_hashtag_map.py -reducer examples/twitter/twit_hashtag_reduce.py -inputURI mongodb://127.0.0.1/test.live -outputURI mongodb://127.0.0.1/test.twit_reduction -file examples/twitter/twit_hashtag_map.py -file examples/twitter/twit_hashtag_reduce.py
  39. Popular Hash Tags db.twit_hashtags.find().sort( {count : -1 }){ “_id” : “YouKnowYoureInLoveIf”, “count” : 287 }{ “_id” : “teamfollowback”, “count” : 200 }{ “_id” : “RT”, “count” : 150 }{ “_id” : “Arsenal”, “count” : 148 }{ “_id” : “milars”, “count” : 145 }{ “_id” : “sanremo”, “count” : 145 }{ “_id” : “LoseMyNumberIf”, “count” : 139 }{ “_id” : “RelationshipsShould”, “count” : 137 }{ “_id” : “Bahrain”, “count” : 129 }{ “_id” : “bahrain”, “count” : 125 }{ “_id” : “oomf”, “count” : 117 }{ “_id” : “BabyKillerOcalan”, “count” : 106 }{ “_id” : “TeamFollowBack”, “count” : 105 }{ “_id” : “WhyDoPeopleThink”, “count” : 102 }{ “_id” : “np”, “count” : 100 }
  40. DEMO 2
  41. Aggregation in Mongo 2.1 db.live.aggregate( { $unwind : “$entities.hashtags” } , { $match : { “entities.hashtags.text” : { $exists : true } } } , { $group : { _id : “$entities.hashtags.text”, count : { $sum : 1 } } } , { $sort : { count : -1 } }, { $limit : 10 })
  42. Popular Hash Tags db.twit_hashtags.aggregate(a){ “result” : [ { “_id” : “YouKnowYoureInLoveIf”, “count” : 287 }, { “_id” : “teamfollowback”, “count” : 200 }, { “_id” : “RT”, “count” : 150 }, { “_id” : “Arsenal”, “count” : 148 }, { “_id” : “milars”, “count” : 145 }, { “_id” : “sanremo”,“count” : 145 }, { “_id” : “LoseMyNumberIf”, “count” : 139 }, { “_id” : “RelationshipsShould”, “count” : 137 }, { “_id” : “Bahrain”, “count” : 129 }, { “_id” : “bahrain”, “count” : 125 } ],”ok” : 1}
  43. The Future of humongous data
  44. What is BIG? BIG today is normal tomorrow
  45. Data Growth 9,00090006750 4,4004500 2,1502250 1,000 500 55 120 250 1 4 10 24 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Millions of URLs
  46. Data Growth 9,000 9000 6750 4,4004500 2,1502250 1,000 500 55 120 250 1 4 10 24 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Millions of URLs
  47. 2012 Generating over 250 Millions of tweets per day
  48. MongoDB enables us to scale with the redefinition of BIG. New processing tools like Hadoop & Storm are enabling us to process the new BIG.
  49. Hadoop is our first step
  50. MongoDB iscommitted to working with best data tools including Hadoop, Storm, Disco, Spark & more
  51. http://spf13.com http://github.com/spf13 @spf13 Questions? download at github.com/mongodb/mongo-hadoop