Jovian Data Amazon Final Version
Download as PPTX, PDF0 likes528 views
- JovianDATA provides a cloud-based analytics platform that transforms large volumes of advertising, search, and sales data into actionable insights at low cost. - The platform uses role-based temporary clusters on AWS EC2 to reduce capex, with dynamic provisioning and selective data replication for on-demand high performance analytics at a fraction of the cost of traditional architectures. - It further reduces costs through application isolation techniques like hibernating unused applications on Amazon S3/EBS and provisioning them in parallel on EC2 when needed, saving up to 100x on infrastructure costs compared to always-on architectures.
Jovian Data Amazon Final Version
- 1. Analytics at the Speed of ThoughtSatya RamachandranVice President of EngineeringAnupam SinghChief Technology OfficerApril 14, 20102460 North First Street, Suite 170, San Jose, CA 95131 408-433-9383 www.joviandata.com
- 2. JovianDATA MissionTechnology platform to optimize your conversion funnel at the lowest cost
- 3. Why move to the cloud?Considering AWS actively but not sure about Cap Ex benefits
- 4. Current stack’s cloud readiness
- 5. Application provisioning challengesIntroducing JovianDATAAd Server Data, Search Engine Data+Sales/Conversion DataActionable InsightsBillions of Impressions, Clicks & Conversions (100’s of TB)+Site/Web Analytics DataNo samplingMulti-dimensional analytics+In-FlightCustomer/3rd Party Data Extremely low TCOFast Time-to-Value SaaS+Other Data Sources
- 6. Transforming Data to Actionable InsightsCampaign Heat MapTimeFully Materialized Data CubeIncremental updates
- 9. AgendaJovianDATA Company OverviewJovianInsights – The Power of AnalyticsJovianDATA Cube StorageInnovations in Advanced Analytics using commodity clustersAnalytics Lifecycle ManagementInnovations in Cloud Infrastructure Management
- 10. Avoiding Expensive Data ProcessingUsage based Automatic View MaterializationAvoid Network I/O Multi-Dimensional PartitioningReduce Disk I/OBy Materializing Expensive Groups
- 11. Why move to the cloud?
- 12. AgendaReducing CapexDynamic ProvisioningApplication Isolation
- 13. Managing CapEx with Role Based ClustersSINGLECLUSTER FOR DATA CLEANSING, LOAD AND QUERY15TB100 NODESMonthly Cost = $28,800
- 14. Managing Cap-Ex with Role Based ClustersUIAd Server Data, Search Engine Data2 hours daily for load on 10 nodes8 hours daily for query on 5 nodesMonthly Cost = $2,052DATA CLEANSINGQUERYLOAD MODELHIBERNATE MODEL
- 15. AgendaReducing CapexDynamic ProvisioningApplication Isolation
- 16. Selective Replication for on demand perfPower analyst needs to perform complex, heavy number-crunching query that typically take 8 - 10 hours
- 17. Solution
- 18. FlexRestoreTM
- 19. Adds two new temporary nodes (Temp1, Temp2)
- 20. Creates new replicas for hot partitions and redistributes across nodesWith Replication Factor = 1Site Section Analytics = 10 minutesNode1Node2Node3Node4P34P12P1P1P22P3P3P12With Replication Factor = 10Site Section Analytics = 30 secondsP12P22P34P22Nodeset1P3P34P1P3Temp1Temp2P1P34P22P12
- 21. Reduce replication to maintain cost When the analysis is done and the extra performance is not needed, the SLA Controller brings down the two temporary nodes (and the extra replicas)
- 22. Benefits
- 23. High performance computing power when you need it
- 24. But only when you need it to hold down operating costsNode1Node2Node3Node4P34P1P12P1P22P3P3P12P12P22P34P22Nodeset1P3P34P1P3Temp1Temp2P34P1P22P12
- 25. AgendaReducing CapexDynamic ProvisioningApplication Isolation
- 26. Provision Tera Scale Applications in MinutesWithout Application IsolationData for all advertisers is kept‘live’ on 50 nodes Campaign Manager needs to runheavy duty reports for a Big Advertiser50 live nodes per month=$14, 400FUNNEL ANALYSIS FOR CLIENT
- 27. Provision Tera Scale Applications in MinutesApplication is provisioned in parallel from S3/EBS into EC2Campaign Manager requestsApplication Provisioning for a Specific Advertiser50 nodes for fortnightly analysis=$320FUNNEL ANALYSIS FOR CLIENTHIBERNATED MODEL
- 28. SummaryReducing CapEx with Role based Temporary Clusters on EC210x Cost Savings with EC2 usageDynamic Provisioning with Selective Replication on EC210x Performance on EC2 replicationApplication Isolation with Application Hibernation on S3/EBS100x Cost Savings with EC2-S3
- 29. Thank You