Hadoop at eBay
Download as PPT, PDF7 likes2,092 views
The document summarizes eBay's Hadoop analytics platform called Athena: 1. Athena is eBay's large Hadoop cluster with over 500 nodes that was built in 2010 to power advanced analytics on complex data. 2. The platform utilizes Hadoop core components like HDFS, MapReduce, HBase and tools like Hive, Pig and Mobius to access and analyze clickstream, transaction, and other data from eBay. 3. Athena supports a variety of use cases including search relevance, data mining, and analytics to enhance eBay's operations and customer experience.
Hadoop at eBay
- 1. @eBay Anil Madan amadan@ebay.com Analytics Platform Development
- 2. 2007 Research Team Builds a 4 node Cluster Subset of Click Stream and EDW data Innovation with Mobius Query Language Visualization and Click Path analysis 2009 Sept Search Clusters Machine Learning Ranking cluster of 28 nodes Search relevance cluster of 10 nodes Subset of Click Stream and EDW Data 2010 May – Athena* Exploratory Cluster of 532 nodes Platform Teams join hands with Search/Research to build a larger cluster . Build it as a core competency for advanced insights for complex data Rapid build-out with timelines pulled in by couple of months * Athena, is the goddess of civilization, wisdom, strength, strategy, craft, justice and skill in Greek mythology MIT's Athena ushered the world in a new era of distributed systems when it started in the mid 80s.
- 3. Infrastructure Enterprise Nodes Sun 64bit , Red Hat Linux 2 Quad Core Nehalem, 72GB RAM, 4TB Servers NameNode(s) Job Tracker Zookeeper HBaseMaster Ganglia Server eBay (Cloudera) HUE Data Nodes SGI-Rackables, Cent OS, 1U , 5.3PB 2 Quad Core Nehalem, 36GB RAM, 10TB Hbase on 20 nodes Network TOR 1Gbps Core Switches uplink 40Gbps
- 4. Ecosystem Hadoop Core (HDFS,Common) MapReduce (Java, Streaming, Pipes,Scala) Data Access (Hbase, Pig, Hive) Tools & Libraries (HUE,UC4,Oozie.Mobius,Mahout) Monitoring & Alerting (Ganglia, Nagios) MapReduce Sourcing data primarily Java Applications using Perl, Scala, Python… Data Access Frameworks Hbase - for EDWdata Pig – data piplelines Hive – Adhoc queries MQL – Mobius Query Language Monitoring & Alerting Ganglia, Nagios Tools HUE/Mobius – lifecycle of user jobs UC4 - scheduling Oozie – user workflow and data pipelines Mahout – data mining
- 5. Administration Groups Built to support multiple groups Job invocation uses the group name Fair Scheduler Allocations based on investment Weights Minimum share of mappers and reducers poolMaxJobsDefault userMaxJobsDefault defaultMinSharePreemptionTimeout fairSharePreemptionTimeout Auth & Auth HUE – custom module to use corp. credentials CLI*– PAM custom module Security* - Implement token interface to replace Kerberos with SAML. * Work in Progress
- 6. Data Sourcing Patterns Click Stream EDW Images Search Indices Analytics Reporting Algorithmic Models Acquisition Description Source Preparation Format Pattern Click Stream Session Event Session Container Session/Event Streamed as LZO/Text SessionContainer generate Sequence Files Session/Event Data Build an index and use LzoTextInputFormat for splits based on the work done by Johan Oskarsson/Twitter Session Container ‘Value to Type Conversion’ Pattern Secondary sort with reduce side join EDW Item Transaction User Feedback Bids Streamed as GZIP/Text Generate SequenceFile/ Hbase snapshot with previous day snapshot and current day data. Hive StorageHandlers to point to SequenceFile/Hbase snapshot TotalOrderPartitoner with RandomSamplers to identify partition ranges for reducers. Create Hbase regions using Hfile Update RegionServers using ruby script loadtable.rb Concerns - Hbase append performance, Hfile flush HBASE-1923
- 7. Search Use Case – Machine Learned Ranking ClickStream Items Users Feedback Classifiers Ranking Function Great Search Results Goal Enhance search relevance for eBay’s items. Hadoop Usage Build a ranking function that takes multiple factors into account like price, listing format, seller track record, relevance. Ability to add new factors to validate hypothesis .
- 8. Research Use Case – Description Data Mining Goal Extend catalog coverage Hadoop Usage Leverage data mining/machine learning techniques to create inventory into name value pairs in an completely unsupervised way BARBIE 1999 "PREMIERE NIGHT" Home Shopping Special Edition Gorgeous Doll With Beautiful Blond Hair / In A Gown Of Purple And Silver New / Never Removed From Box / Doll Is In Mint Condition / Remember This Beauty Is 11 Years Old Free Shipping To US Only / Will Ship International / Please E-mail For Cost Feel Free To Ask Me Any Questions Or Concerns Smoke - Free Environment Free Shipping Year: 1999 Model: premiere night Edition: home shopping special Hair: blond Gown: purple and silver Condition: new / never removed from box / mint
- 9. Platform Details Metrics Job Statistics, System/Disk Consumption, Utilization Infrastructure Publish/Subscribe ETL tools, low latency data movement Development Tools, Environment, IDE, Architecture Schemas, Metadata, Governance, Policies Operations Administration, Configuration, Monitoring Reporting Visualization, BI Generation, Information delivery Security User & Group Management, Auth & Auth Clusters Details Exploratory Strategic investment 1000-5000 nodes Production Site facing, low latency, high availability Use Case Specific Advertising, Trust & Safety , Merchandizing
- 10. Acknowledgments Athena Team Cloudera Inc. Community
Editor's Notes
- #7: 09/08/10