Sharing a Startup’s Big Data Lessons
Download as PPTX, PDF2 likes607 views
Traackr evaluated several NoSQL database options to store its heterogeneous, unstructured web data. Document databases were the best fit due to their flexibility to store variable length text like tweets and blog posts without predefined schemas. MongoDB was selected due to its maturity, adoption, and support for ad-hoc queries and batch processing needed by Traackr in early 2010.
![{
"realName": "David Chancogne",
"title": "CTO",
"description": "Web. Geek.rnTraackr: http://traackr.comrnPropz: http://propz.me",
"primaryAffiliation": "Traackr",
"email": "dchancogne@traackr.com",
"location": "Cambridge, MA, United States",
"siteReferences": [
{
"siteUrl": "http://twitter.com/dchancogne",
"metrics": [
{
"value": 216,
"name": "twitter_followers_count"
},
{
"value": 2107,
"name": "twitter_statuses_count"
}
]
},
{
"siteUrl": "http://traackr.com/blog/author/david",
"metrics": [
{
"value": 21,
"name": "google_inbound_links"
}
]
}
]
}
Influencer data as JSON](https://image.slidesharecdn.com/bigdataexperiencesfromastartup-130220094556-phpapp02/85/Sharing-a-Startup-s-Big-Data-Lessons-44-320.jpg)
![{
”_id": "770cf5c54492344ad5e45fb791ae5d52”,
"realName": "David Chancogne",
"title": "CTO",
"description": "Web. Geek.rnTraackr: http://traackr.comrnPropz: http://propz.me",
"primaryAffiliation": "Traackr",
"email": "dchancogne@traackr.com",
"location": "Cambridge, MA, United States",
"siteReferences": [
{ Embedded list of
"siteId": "b31236da306270dc2b5db34e943af88d", references to sites
"contribution": 0.25 augmented with
},
influencer-specific
{
"siteId": "602dc370945d3b3480fff4f2a541227c", site attributes (e.g.
"contribution": 1.0 percent contribution
} to content)
]
}
Modeling an influencer](https://image.slidesharecdn.com/bigdataexperiencesfromastartup-130220094556-phpapp02/85/Sharing-a-Startup-s-Big-Data-Lessons-129-320.jpg)
![{
”_id": "770cf5c54492344ad5e45fb791ae5d52”,
"realName": "David Chancogne",
"title": "CTO",
"description": "Web. Geek.rnTraackr: http://traackr.comrnPropz: http://propz.me",
"primaryAffiliation": "Traackr",
"email": "dchancogne@traackr.com",
"location": "Cambridge, MA, United States",
"siteReferences": [
{
"siteId": "b31236da306270dc2b5db34e943af88d",
"contribution": 0.25 siteId indexed for
}, “find influencers
{ connected to site X”
"siteId": "602dc370945d3b3480fff4f2a541227c",
"contribution": 1.0
}
]
}
> db.influencers.ensureIndex({siteReferences.siteId: 1});
> db.influencers.find({siteReferences.siteId: "602dc370945d3b3480fff4f2a541227c"});
Modeling an influencer](https://image.slidesharecdn.com/bigdataexperiencesfromastartup-130220094556-phpapp02/85/Sharing-a-Startup-s-Big-Data-Lessons-130-320.jpg)
![[PASS Summit 2016] Azure DocumentDB: A Deep Dive into Advanced Features](https://cdn.slidesharecdn.com/ss_thumbnails/passdocdbadvfeatures-161029023827-thumbnail.jpg?width=560&fit=bounds)
More Related Content
What's hot (20)
Viewers also liked (17)
Similar to Sharing a Startup’s Big Data Lessons (20)
Sharing a Startup’s Big Data Lessons
- 1. Sharing a Startup’s Big Data Lessons Experiences with non-RDBMS solutions at
- 2. Who we are • A search engine • A people search engine • An influencer search engine • Subscription- based
- 3. George Stathis VP Engineering 14+ years of experience building full-stack web software systems with a past focus on e-commerce and publishing. Currently responsible for building engineering capability to enable Traackr's growth goals.
- 4. What’s this talk about? • Share what we know about Big Data/NoSQL: what’s behind the buzz words? • Our reasons and method for picking a NoSQL database • Share the lessons we learned going through the process
- 5. Big Data/NoSQL: behind the buzz words
- 6. What is Big Data? • 3 Vs: – Volume – Velocity – Variety
- 7. What is Big Data? Volume + Velocity • Data sets too large or coming in at too high a velocity to process using traditional databases or desktop tools. E.g. big science Astronomy web logs atmospheric science rfid genomics sensor networks biogeochemical social networks military surveillance social data medical records internet text and documents photography archives internet search indexing video archives call detail records large-scale e-commerce
- 8. What is Big Data? Variety • Big Data is varied and unstructured Traditional static reports Analytics, exploration & experimentation
- 9. What is Big Data? • Scaling data processing cost effectively $$$$$ $$$$$$$$ $$$
- 10. What is NoSQL? • NoSQL ≠ No SQL • NoSQL ≈ Not Only SQL • NoSQL addresses RDBMS limitations, it’s not about the SQL language • RDBMS = static schema • NoSQL = schema flexibility; don’t have to know exact structure before storing
- 11. What is Distributed Computing? • Sharing the workload: divide a problem into many tasks, each of which can be solved by one or more computers • Allows computations to be accomplished in acceptable timeframes • Distributed computation approaches were developed to leverage multiple machines: MapReduce • With MapReduce, the program goes to the data since the data is too big to move
- 12. What is MapReduce? Source: developer.yahoo.com
- 13. What is MapReduce? • MapReduce = batch processing = analytical • MapReduce ≠ interactive • Therefore many NoSQL solutions don’t outright replace warehouse solutions, they complement them • RDBMS is still safe
- 14. What is Big Data? Velocity • In some instances, being able to process large amounts of data in real-time can yield a competitive advantage. E.g. – Online retailers leveraging buying history and click- though data for real-time recommendations • No time to wait for MapReduce jobs to finish • Solutions: streaming processing (e.g. Twitter Storm), pre-computing (e.g. aggregate and count analytics as data arrives), quick to read key/value stores (e.g. distributed hashes)
- 15. What is Big Data? Data Science • Emergence of Data Science • Data Scientist ≈ Statistician • Possess scientific discipline & expertise • Formulate and test hypotheses • Understand the math behind the algorithms so they can tweak when they don’t work • Can distill the results into an easy to understand story • Help businesses gain actionable insights
- 16. Big Data Landscape Source: capgemini.com
- 17. Big Data Landscape Source: capgemini.com
- 18. Big Data Landscape Source: capgemini.com
- 19. So what’s Traackr and why did we need a NoSQL DB?
- 20. Traackr: context • A cloud computing company as about to launch a new platform; how does it find the most influential IT bloggers on the web that can help bring visibility to the new product? How does it find the opinion leaders, the people that matter?
- 21. Traackr: a people search engine Up to 50 keywords per search!
- 22. Traackr: a people search engine Proprietary 3-scale ranking People as Content search aggregated results by author
- 23. Traackr: 30,000 feet Acquisition Processing Storage & Indexing Services Applications
- 24. NoSQL is usually associated with “Web Scale” (Volume & Velocity)
- 25. Do we fit the “Web scale” profile? • In terms of users/traffic?
- 31. Do we fit the “Web scale” profile? • In terms of users/traffic? • In terms of the amount of data?
- 32. PRIMARY> use traackr switched to db traackr PRIMARY> db.stats() { "db" : "traackr", "collections" : 12, "objects" : 68226121, "avgObjSize" : 2972.0800625760330, That’s a quarter of a "dataSize" : 202773493971, terabyte … "storageSize" : 221491429671, "numExtents" : 199, "indexes" : 33, "indexSize" : 27472394891, "fileSize" : 266623699968, "nsSizeMB" : 16, "ok" : 1 }
- 33. Wait! What? My Synology NAS at home can hold 2TB!
- 34. No need for us to track the entire web Influencer Content Web Content Not at scale :-)
- 35. Do we fit the “Web scale” profile? • In terms of users/traffic? • In terms of the amount of data?
- 36. Variety view of “Web Scale” Web data is: Heterogeneous Unstructured (text)
- 37. Visualization of the Internet, Nov. 23rd 2003 Source: http://www.opte.org/
- 38. Data sources are isolated islands of rich data with lose links to one another
- 39. How do we build a database that models all possible entities found on the web?
- 40. Modeling the web: the RDBMS way
- 42. or
- 44. { "realName": "David Chancogne", "title": "CTO", "description": "Web. Geek.rnTraackr: http://traackr.comrnPropz: http://propz.me", "primaryAffiliation": "Traackr", "email": "[email protected]", "location": "Cambridge, MA, United States", "siteReferences": [ { "siteUrl": "http://twitter.com/dchancogne", "metrics": [ { "value": 216, "name": "twitter_followers_count" }, { "value": 2107, "name": "twitter_statuses_count" } ] }, { "siteUrl": "http://traackr.com/blog/author/david", "metrics": [ { "value": 21, "name": "google_inbound_links" } ] } ] } Influencer data as JSON
- 45. NoSQL = schema flexibility
- 46. Do we fit the “Web scale” profile? • In terms of users/traffic? • In terms of the amount of data?
- 47. Do we fit the “Web scale” profile? • In terms of users/traffic? • In terms of the amount of data? • In terms of the variety of the data ✓
- 48. Traackr’s Datastore Requirements • Schema flexibility ✓ • Good at storing lots of variable length text • Batch processing options
- 49. Requirement: text storage Variable text length: 140 multi-page character < big variance < tweets blog posts
- 50. Requirement: text storage RDBMS’ answer to variable text length: Plan ahead for largest value CLOB/BLOB
- 51. Requirement: text storage Issues with CLOB/BLOG for us: No clue what largest value is CLOB/BLOB for tweets = wasted space
- 52. Requirement: text storage NoSQL solutions are great for text: No length requirements (automated chunking) Limited space overhead
- 53. Traackr’s Datastore Requirements • Schema flexibility ✓ • Good at storing lots of variable length text ✓ • Batch processing options
- 54. Requirement: batch processing Some NoSQL solutions come with MapReduce Source: http://code.google.com/
- 55. Requirement: batch processing MapReduce + RDBMS: Possible but proprietary solutions Usually involves exporting data from RDBMS into a NoSQL system anyway. Defeats data locality benefit of MR
- 56. Traackr’s Datastore Requirements • Schema flexibility ✓ • Good at storing lots of variable length text ✓ • Batch processing options ✓ A NoSQL option is the right fit
- 57. How did we pick a NoSQL DB?
- 58. Bewildering number of options (early 2010) Key/Value Databases Column Databases • Distributed hashtables • Spread sheet like • Designed for high load • Key is a row id • In-memory or on-disk • Attributes are columns • Eventually consistent • Columns can be grouped into families Document Databases Graph Databases • Like Key/Value • Graph Theory G=(E,V) • Value = Document • Great for modeling • Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 59. Bewildering number of options (early 2010) Key/Value Databases Column Databases • Distributed hashtables • Spread sheet like • Designed for high load • Key is a row id • In-memory or on-disk • Attributes are columns • Eventually consistent • Columns can be grouped into families Document Databases Graph Databases • Like Key/Value • Graph Theory G=(E,V) • Value = Document • Great for modeling • Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 60. Trimming options Key/Value Databases Column Databases • Distributed hashtables while•weSpread sheet like Graph Databases: can model • • Key is a row Designed for high as a graph we don’t want to id our domain load • pigeonhole ourselves into this structure. columns In-memory or on-disk • Attributes are • Eventually consistent use these tools for can be grouped We’d rather • Columns specialized data analysis but not as the into families main data store. Document Databases Graph Databases • Like Key/Value • Graph Theory G=(E,V) • Value = Document • Great for modeling • Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 61. Trimming options Key/Value Databases Column Databases Memcache: memory-based, • Distributed hashtables • Spread sheet like we need true persistence • Designed for high load • Key is a row id • In-memory or on-disk • Attributes are columns • Eventually consistent • Columns can be grouped into families Document Databases Graph Databases • Like Key/Value • Graph Theory G=(E,V) • Value = Document • Great for modeling • Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 62. Trimming options Key/Value Databases Column Databases • Distributed hashtables • Spread sheet like • Designed for high load • Key is a row id • In-memory or on-disk • Attributes are columns • Eventually consistent • Columns can be grouped Amazon SimpleDB: not willing to store our data in into families a proprietary datastore. Document Databases Graph Databases • Like Key/Value • Graph Theory G=(E,V) • Value = Document • Great for modeling • Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 63. Trimming options Key/Value Databases Column Databases • Distributed hashtables • Spread sheet like • Designed for high load • Key is a row id • In-memory or on-disk • Attributes are columns • Eventually consistent • Columns can be grouped into families Document Databases Graph Databases • Like Key/Value • Graph Theory G=(E,V) Not willing to store ourProject a Redis and LinkedIn’s data in • Value proprietary datastore. • = Document Voldermort: no query filters, Great for modeling • Document used as queues or better = JSON/BSON networks • JSON = Flexible Schema distributed caches • Great for graph-based query algorithms
- 64. Trimming options Key/Value Databases Column Databases • Distributed hashtables • Spread sheet like • Designed for high load • Key is a row id • In-memory or on-disk • Attributes are columns CouchDB: no ad-hoc queries; • Eventually consistent • Columns can us maturity in early 2010 madebe grouped into families shy away although we did try early prototypes. Document Databases Graph Databases • Like Key/Value • Graph Theory G=(E,V) • Value = Document • Great for modeling • Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 65. Trimming options Key/Value Databases Column Databases • Distributed hashtables • Spread sheet like • Designed for high load • Key is a row id • In-memory or on-disk • Attributes are columns • Eventually consistent • Columns can be grouped into families Document Databases 2010, Graph Databases Cassandra: in early • • maturity questions, no secondary Graph Theory G=(E,V) Like Key/Value • Value = Document processing Great for modeling indexes and no batch • • options (came later on). Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 66. Trimming options Key/Value Databases Column Databases • Distributed hashtables • Spread sheet like • MongoDB: in earlyis a row id Designed for high load • Key 2010, maturity • In-memory or on-disk questions, adoption questions • Attributes are columns and no batch processing options. • Eventually consistent • Columns can be grouped into families Document Databases Graph Databases • Like Key/Value • Graph Theory G=(E,V) • Value = Document • Great for modeling • Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 67. Trimming options Key/Value Databases Column Databases • Distributed hashtables • Spread sheet like • Designed for high load • Key is a row id • In-memory or on-disk • Attributes are columns • Eventually consistent • Columns can be grouped into families Document Databases Graph Databases • Like Key/Value very close but• in early 2010, Riak: Graph Theory G=(E,V) • • Great for Value = Document adoption questions. modeling we had • Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 68. Trimming options Key/Value Databases Column Databases • Distributed hashtables • Spread sheet like • Designed for high load • Key is a row id • In-memory or on-disk • Attributes are columns • Eventually consistent • Columns can be grouped into families Document Databases Graph Databases • Like Key/Value came across as•theGraphmature G=(E,V) HBase: most Theory • Value = Document with several deployments, a at the time, • Great for modeling • Document = JSON/BSON "out-of-the box" healthy community, networks secondary indexes through a contrib and • JSON = Flexible Schema • Great for graph-based support for batch processing using Hadoop/MR query algorithms .
- 69. Lessons Learned Challenges Rewards - Complexity - Choices - Missing Features - Empowering - Problem solution fit - Community - Resources - Cost
- 70. Rewards: Choices Key/Value Databases Column Databases • Distributed hashtables • Spread sheet like • Designed for high load • Key is a row id • In-memory or on-disk • Attributes are columns • Eventually consistent • Columns can be grouped into families Document Databases Graph Databases • Like Key/Value • Graph Theory G=(E,V) • Value = Document • Great for modeling • Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 71. Rewards: Choices Source: capgemini.com
- 72. Lessons Learned Challenges Rewards - Complexity - Choices - Missing Features - Empowering - Problem solution fit - Community - Resources - Cost
- 73. When Big-Data = Big Architectures Must have an odd Master/slave architecture number of means a single point of failure, Zookeeper quorum so you need to protect your nodes master. Then you can run your Hbase nodes but it’s recommended to co-locate regionservers with hadoop datanodes so you have to manage resources. Must have a Hadoop HDFS cluster of at least 2x replication factor nodes And then we also have to manage the MapReduce processes and resources in the Hadoop layer. Source: http://www.larsgeorge.com/2009/10/hbase-architecture-101-storage.html
- 75. Jokes aside, no one said open source was easy to use
- 76. To be expected • Hadoop/Hbase are designed to move mountains • If you want to move big stuff, be prepared to sometimes use big equipment
- 77. What it means to a startup Development capacity before Congrats, you are now a sysadmin… Development capacity after
- 78. Lessons Learned Challenges Rewards - Complexity - Choices - Missing Features - Empowering - Problem solution fit - Community - Resources - Cost
- 79. Mapping an saved search to a column store Name Ranks References to influencer records
- 80. Mapping an saved search to a column store “attributes” column family Unique for general “influencerId” column family key attributes for influencer ranks and foreign keys
- 81. Mapping an saved search to a column store Influencer ranks can be attribute “name” attribute names as well
- 82. Mapping an saved search to a column store Can get pretty long so needs indexing and pagination
- 83. Problem: no out-of-the-box row-based indexing and pagination
- 84. Jumping right into the code
- 85. Lessons Learned Challenges Rewards - Complexity - Choices - Missing Features - Empowering - Problem solution fit - Community - Resources - Cost
- 86. a few months later…
- 87. Need to upgrade to Hbase 0.90 • Making sure to remain on recent code base • Performance improvements • Mostly to get the latest bug fixes No thanks!
- 88. Looks like something is missing
- 90. Our DB indexes depend on this!
- 91. Let’s get this straight • Hbase no longer comes with secondary indexing out-of-the-box • It’s been moved out of the trunk to GitHub • Where only one other company besides us seems to care about it
- 92. Only one other maintainer besides us
- 93. What it means to a startup Congrats, you are now an hbase contrib maintainer… Development capacity
- 95. Lessons Learned Challenges Rewards - Complexity - Choices - Missing Features - Empowering - Problem solution fit - Community - Resources - Cost
- 96. Homegrown Hbase Indexes Row ids for Posts Rows have id prefixes that can be efficiently scanned using STARTROW and STOPROW filters
- 97. Homegrown Hbase Indexes Row ids for Posts Find posts for influencer_id_1234
- 98. Homegrown Hbase Indexes Row ids for Posts Find posts for influencer_id_5678
- 99. Homegrown Hbase Indexes • No longer depending on unmaintained code • Work with out-of-the-box Hbase installation
- 100. What it means to a startup You are back but you still need to maintain indexing logic Development capacity
- 101. a few months later…
- 102. Cracks in the data model huffingtonpost.com published under writes for http://www.huffingtonpost.com/arianna-huffington/post_1.html http://www.huffingtonpost.com/arianna-huffington/post_2.html authored by http://www.huffingtonpost.com/arianna-huffington/post_3.html huffingtonpost.com published under writes for http://www.huffingtonpost.com/shaun-donovan/post1.html http://www.huffingtonpost.com/shaun-donovan/post2.html authored by http://www.huffingtonpost.com/shaun-donovan/post3.html
- 103. Cracks in the data model huffingtonpost.com published under writes for Denormalized/duplicated for fast runtime access http://www.huffingtonpost.com/arianna-huffington/post_1.html and storage of influencer- http://www.huffingtonpost.com/arianna-huffington/post_2.html authored by http://www.huffingtonpost.com/arianna-huffington/post_3.html to-site relationship properties huffingtonpost.com published under writes for http://www.huffingtonpost.com/shaun-donovan/post1.html http://www.huffingtonpost.com/shaun-donovan/post2.html authored by http://www.huffingtonpost.com/shaun-donovan/post3.html
- 104. Cracks in the data model huffingtonpost.com published under writes for http://www.huffingtonpost.com/arianna-huffington/post_1.html http://www.huffingtonpost.com/arianna-huffington/post_2.html authored by huffingtonpost.com published under writes for http://www.huffingtonpost.com/shaun-donovan/post1.html http://www.huffingtonpost.com/shaun-donovan/post2.html authored by http://www.huffingtonpost.com/shaun-donovan/post3.html http://www.huffingtonpost.com/arianna-huffington/post_3.html Content attribution logic could sometimes mis-attribute posts because of the duplicated data.
- 105. Cracks in the data model huffingtonpost.com published under writes for http://www.huffingtonpost.com/arianna-huffington/post_1.html authored by huffingtonpost.com published under writes for http://www.huffingtonpost.com/shaun-donovan/post1.html http://www.huffingtonpost.com/shaun-donovan/post2.html authored by http://www.huffingtonpost.com/shaun-donovan/post3.html http://www.huffingtonpost.com/arianna-huffington/post_3.html http://www.huffingtonpost.com/arianna-huffington/post_2.html Exacerbated when we started tracking people’s content on a daily basis in mid- 2011
- 106. Fixing the cracks in the data model Normalize the sites http://www.huffingtonpost.com/arianna-huffington/post_1.html http://www.huffingtonpost.com/arianna-huffington/post_2.html authored by http://www.huffingtonpost.com/arianna-huffington/post_3.html writes for published under huffingtonpost.com published under writes for http://www.huffingtonpost.com/shaun-donovan/post1.html http://www.huffingtonpost.com/shaun-donovan/post2.html authored by http://www.huffingtonpost.com/shaun-donovan/post3.html
- 107. Fixing the cracks in the data model • Normalization requires stronger secondary indexing • Our application layer indexing would need revisiting…again!
- 108. What it means to a startup Psych! You are back to writing indexing code. Development capacity
- 110. Lessons Learned Challenges Rewards - Complexity - Choices - Missing Features - Empowering - Problem solution fit - Community - Resources - Cost
- 111. Traackr’s Datastore Requirements (Revisited) • Schema flexibility • Good at storing lots of variable length text • Out-of-the-box SECONDARY INDEX support! • Simple to use and administer
- 112. NoSQL picking – Round 2 (mid 2011) Key/Value Databases Column Databases • Distributed hashtables • Spread sheet like • Designed for high load • Key is a row id • In-memory or on-disk • Attributes are columns • Eventually consistent • Columns can be grouped into families Document Databases Graph Databases • Like Key/Value • Graph Theory G=(E,V) • Value = Document • Great for modeling • Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 113. NoSQL picking – Round 2 (mid 2011) Key/Value Databases Column Databases • Distributed hashtables • Spread sheet like • Designed for high load • Key is a row id • In-memory or on-disk • Attributes are columns • Eventually consistent • Columns can be grouped into families Nope! Document Databases Graph Databases • Like Key/Value • Graph Theory G=(E,V) • Value = Document • Great for modeling • Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 114. NoSQL picking – Round 2 (mid 2011) Key/Value Databases Column Databases • Distributed hashtables • Spread sheet like • Designed for high load • Key is a row id • Graph Databases:•weAttributes are columns In-memory or on-disk looked at • • Columns can Eventually consistent closer but passed again be grouped Neo4J a bit for the same reasons into families as before. Document Databases Graph Databases • Like Key/Value • Graph Theory G=(E,V) • Value = Document • Great for modeling • Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 115. NoSQL picking – Round 2 (mid 2011) Key/Value Databases Column Databases Memcache: still no • Distributed hashtables • Spread sheet like • Designed for high load • Key is a row id • In-memory or on-disk • Attributes are columns • Eventually consistent • Columns can be grouped into families Document Databases Graph Databases • Like Key/Value • Graph Theory G=(E,V) • Value = Document • Great for modeling • Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 116. NoSQL picking – Round 2 (mid 2011) Key/Value Databases Column Databases • Distributed hashtables • Spread sheet like • Designed for high load • Key is a row id • In-memory or on-disk • Attributes are columns • Eventually consistent • Columns can be grouped Amazon SimpleDB: still no. into families Document Databases Graph Databases • Like Key/Value • Graph Theory G=(E,V) • Value = Document • Great for modeling • Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 117. NoSQL picking – Round 2 (mid 2011) Key/Value Databases Column Databases • Distributed hashtables • Spread sheet like • Designed for high load • Key is a row id • In-memory or on-disk • Attributes are columns • Eventually consistent • Columns can be grouped into families Document Databases Graph Databases • Like Key/Value • Graph Theory G=(E,V) Not willing to store ourProject a Redis and LinkedIn’s data in • Value proprietary datastore. • =Voldermort: still no Document Great for modeling • Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 118. NoSQL picking – Round 2 (mid 2011) Key/Value Databases Column Databases • Distributed hashtables • Spread sheet like • Designed for high load • Key is a row id • In-memory or on-disk • Attributes are columns CouchDB: more mature but still • Eventually consistent • Columns can no ad-hoc queries. be grouped into families Document Databases Graph Databases • Like Key/Value • Graph Theory G=(E,V) • Value = Document • Great for modeling • Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 119. NoSQL picking – Round 2 (mid 2011) Key/Value Databases Column Databases • Distributed hashtables • Spread sheet like • Designed for high load • Key is a row id • In-memory or on-disk • Attributes are columns • Eventually consistent • Columns can be grouped into families Document Databasesa bit, added Cassandra: matured quite Graph Databases secondary indexes and batch processing • • Like Key/Valuerestrictive in its’ use than Graph Theory G=(E,V) options but more • • Value =solutions. After the Hbase lesson, Great for modeling other Document • Document useJSON/BSON simplicity of = was now more important. networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 120. NoSQL picking – Round 2 (mid 2011) Key/Value Databases Column Databases • Distributed hashtables • Spread sheet like • Designed for high load • Key is a row id • In-memory or on-disk • Attributes are columns • Eventually consistent • Columns can be grouped into families Document Databases Graph Databases • • Graph Theory G=(E,V) Like Key/Value strong contender still but Riak: • • Great for Value = Document questions remained. modeling adoption • Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 121. NoSQL picking – Round 2 (mid 2011) Key/Value Databasesby leaps Column Databases MongoDB: matured and bounds, increased • • Spread sheet like Distributed hashtables 10gen, advanced indexing adoption, support from • • batch processing Designed for high load as some Key is a row id out-of-the-box as well options, breeze to use, well documented and fit into • • Attributes In-memory or on-disk code base very nicely. are columns our existing • Eventually consistent • Columns can be grouped into families Document Databases Graph Databases • Like Key/Value • Graph Theory G=(E,V) • Value = Document • Great for modeling • Document = JSON/BSON networks • JSON = Flexible Schema • Great for graph-based query algorithms
- 122. Lessons Learned Challenges Rewards - Complexity - Choices - Missing Features - Empowering - Problem solution fit - Community - Resources - Cost
- 123. Immediate Benefits • No more maintaining custom application-layer secondary indexing code
- 124. What it means to a startup Yay! I’m back! Development capacity
- 125. Immediate Benefits • No more maintaining custom application-layer secondary indexing code • Single binary installation greatly simplifies administration
- 126. What it means to a startup Honestly, I thought I’d never see you guys again! Development capacity
- 127. Immediate Benefits • No more maintaining custom application-layer secondary indexing code • Single binary installation greatly simplifies administration • Our NoSQL could now support our domain model
- 129. { ”_id": "770cf5c54492344ad5e45fb791ae5d52”, "realName": "David Chancogne", "title": "CTO", "description": "Web. Geek.rnTraackr: http://traackr.comrnPropz: http://propz.me", "primaryAffiliation": "Traackr", "email": "[email protected]", "location": "Cambridge, MA, United States", "siteReferences": [ { Embedded list of "siteId": "b31236da306270dc2b5db34e943af88d", references to sites "contribution": 0.25 augmented with }, influencer-specific { "siteId": "602dc370945d3b3480fff4f2a541227c", site attributes (e.g. "contribution": 1.0 percent contribution } to content) ] } Modeling an influencer
- 130. { ”_id": "770cf5c54492344ad5e45fb791ae5d52”, "realName": "David Chancogne", "title": "CTO", "description": "Web. Geek.rnTraackr: http://traackr.comrnPropz: http://propz.me", "primaryAffiliation": "Traackr", "email": "[email protected]", "location": "Cambridge, MA, United States", "siteReferences": [ { "siteId": "b31236da306270dc2b5db34e943af88d", "contribution": 0.25 siteId indexed for }, “find influencers { connected to site X” "siteId": "602dc370945d3b3480fff4f2a541227c", "contribution": 1.0 } ] } > db.influencers.ensureIndex({siteReferences.siteId: 1}); > db.influencers.find({siteReferences.siteId: "602dc370945d3b3480fff4f2a541227c"}); Modeling an influencer
- 131. Other Benefits • Ad hoc queries and reports became easier to write with JavaScript: no need for a Java developer to write map reduce code to extract the data in a usable form like it was needed with Hbase. • Simpler backups: Hbase mostly relied on HDFS redundancy; intra- cluster replication is available but experimental and a lot more involved to setup. • Great documentation • Great adoption and community
- 132. looks like we found the right fit!
- 133. We have more of this Development capacity
- 134. And less of this Source: socialbutterflyclt.com
- 135. Recap & Final Thoughts • 3 Vs of Big Data: – Volume – Velocity – Variety Traackr • Big Data technologies are complementary to SQL and RDBMS • Until machines can think for themselves Data Science will be increasingly important
- 136. Recap & Final Thoughts • Be prepared to deal with less mature tech • Be as flexible as the data => fearless refactoring • Importance of ease of use and administration cannot be overstated for a small startup
- 137. Q&A
Editor's Notes
- #7: Big science: Large Hadron Collider (LHC)Sensor networks: forest fire detectionCall detail record, a record of a (billing) event produced by a telecommunication network element
- #8: Big science: Large Hadron Collider (LHC)Sensor networks: forest fire detectionCall detail record, a record of a (billing) event produced by a telecommunication network element
- #9: Scaling here means maintaining throughput of computation and analysis while data sizes increase: divide up the work on multiple machines
- #10: Scaling here means maintaining throughput of computation and analysis while data sizes increase: divide up the work on multiple machines
- #15: Scaling here means maintaining throughput of computation and analysis while data sizes increase: divide up the work on multiple machines
- #16: Scaling here means maintaining throughput of computation and analysis while data sizes increase: divide up the work on multiple machines
- #17: Scaling here means maintaining throughput of computation and analysis while data sizes increase: divide up the work on multiple machines
- #18: Scaling here means maintaining throughput of computation and analysis while data sizes increase: divide up the work on multiple machines
- #19: Scaling here means maintaining throughput of computation and analysis while data sizes increase: divide up the work on multiple machines
- #33: Taking a look at the amount of storage we are using as of a month ago in Mongo; this includes indexes
- #35: The point is that we don’t need to track the entire web: just the subset belonging to influencers!
- #37: There is a different perspective on “Web Scale” that has to do with the nature of the data on the web
- #44: Take the approach of using a simplifiedentity model
- #45: …withsemi-structured data storage formats like JSON:Facilitate capturing related attribute structures Enablethe flexibility of definingnew attributes as they are discovered
- #52: CLOB pre-allocated space
- #53: Sparse maps
- #55: - This is something we thought we needed back in early 2010- Traack needs to score its’ entire DB of influencers on a weekly basis to adjust the weighted averages and stats that drive the scores. This means processing north of 750K of sites, over 650K influencers and soon, millions of posts (post-level attributes)
- #61: Graph Databases: while we can model our domain as a graph we don’t want to pigeonhole ourselves into this structure. We’d rather use these tools for specialized data analysis but not as the main data store.
- #62: Memcache: memory-based,we need true persistence
- #63: Amazon SimpleDB: not willing to store our data in a proprietary datastore.
- #64: Redis and LinkedIn’s Project Voldermort: no query filters, better used as queues or distributed caches
- #65: CouchDB: no ad-hoc queries; maturity in early 2010 made us shy away although we did try early prototypes
- #66: Cassandra: in early 2010, maturity questions, no secondary indexes and no batch processing options (came later on).
- #67: MongoDB: in early 2010, maturity questions, adoption questions and no batch processing options
- #68: Riak: very close but in early 2010, we had adoption questions
- #69: HBase: came across as the most mature at the time, with several deployments, a healthy community, "out-of-the box" secondary indexes through a contrib and support for batch processing using Hadoop/MR Hadoop and its’ maturity was a big reason we picked HBase
- #74: Had to deal with a complex right from the start:- minimum number of data nodes to support replication- odd number of zookeper nodes to avoid voting deadlocks- co-locating region servers = paying close attention to JVM resources- Master = SPOF- co-locating job trackers = paying close attention to JVM resources
- #80: - Quick overview of how we modeled a list in hbase => saved searches- This is what our customers see- Let's consider the name, the ranks of the influencers and the influencer references
- #81: Each row has a unique key: the alist idWe would group general attributes under one family of columns appropriately named “attributes”. Benefit: can get Alist information without loading all the influencersWe would group the influencer references under another family of columns named “influencerIds”
- #82: Now we can see where the attributes we see on the screen are stored
- #85: - We coded the pagination and indexing features ourselves and contributed them back- Felt really good about it!
- #94: It wasn’t bad enough we had to write our own code to support our indexing needs, we now had to maintain a third-party code base that was quickly becoming outdated!
- #97: Simplified example for posts
- #104: Denormalized/duplicated for fast runtime access and storage of influencer-to-site relationship properties
- #105: Content attribution logic could sometimes mis-attribute posts because of the duplicated
- #106: Exacerbated when we started tracking people’s content on a daily basis in mid-2011
- #115: Graph Databases: we looked at Neo4J a bit closer but passed again for the same reasons as before
- #119: CouchDB: more mature but still no ad-hoc queries
- #120: Cassandra: matured quite a bit, added secondary indexes and batch processing options but more restrictive in its’ use than other solutions. After the Hbase lesson, simplicity of use was now more important
- #121: Riak: strong contender still but adoption questions
- #122: MongoDB: matured by leaps and bounds, increased adoption, support from 10gen, advanced indexing out-of-the-box as well as some batch processing options, breeze to use, well documented and fit into our existing code base very nicely.
- #130: Embedded list of references to sites augmented with influencer-specific site attributes (e.g. percent contribution to content)
- #131: siteId indexed for “find influencers connected to site X”
- #136: Big science: Large Hadron Collider (LHC)Sensor networks: forest fire detectionCall detail record, a record of a (billing) event produced by a telecommunication network element