Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Cloud Native Cost Optimization UCC

6,463 views

Published on

Summary of fast development and cloud native architecture along with cost optimization techniques. Presented as opening keynote at the Utility and Cloud Computing 2014 as part of the Cloud Control Workshop.

Published in: Technology

Cloud Native Cost Optimization UCC

  1. 1. Cloud Native Cost Optimization Adrian Cockcroft @adrianco Technology Fellow - Battery Ventures UCC2014 Cloud Control Workshop - December 2014
  2. 2. @adrianco What does everyone want?
  3. 3. @adrianco Less Time Less Cost
  4. 4. @adrianco Faster ! See talks by @adrianco Speed and Scale - QCon New York Fast Delivery - GOTO Copenhagen
  5. 5. @adrianco Cheaper ! This talk: How to use Cloud Native architecture to reduce cost without slowing down
  6. 6. Speeding up Development Cloud Native Applications Cost Optimization
  7. 7. Why am I here? %*&!” By Simon Wardley http://enterpriseitadoption.com/
  8. 8. Why am I here? %*&!” 2009 By Simon Wardley http://enterpriseitadoption.com/
  9. 9. Why am I here? %*&!” 2009 By Simon Wardley http://enterpriseitadoption.com/
  10. 10. Why am I here? @adrianco’s job at the intersection of cloud and Enterprise IT, looking for disruption and opportunities. %*&!” By Simon Wardley http://enterpriseitadoption.com/ 2014 2009
  11. 11. Why am I here? @adrianco’s job at the intersection of cloud and Enterprise IT, looking for disruption and opportunities. %*&!” By Simon Wardley http://enterpriseitadoption.com/ 2014 2009 Example: Docker wasn’t on anyone’s roadmap for 2014. It’s on everyone’s roadmap for 2015.
  12. 12. What does @adrianco do? Maintain Relationship with Cloud Vendors @adrianco Technology Due Diligence on Deals Presentations at Conferences Presentations at Companies Technical Advice for Portfolio Companies Program Committee for Conferences Networking with Tinkering with Interesting People Technologies
  13. 13. Speeding Up Development
  14. 14. Observe Orient Act Continuous Delivery Decide
  15. 15. Land grab opportunity Competitive Observe Orient Decide Act Move Customer Pain Point Measure Customers Continuous Delivery
  16. 16. Land grab opportunity Competitive Observe Orient Decide Act Move Customer Pain Point INNOVATION Measure Customers Continuous Delivery
  17. 17. Land grab opportunity Competitive Observe Orient Decide Act Move Customer Pain Point Analysis Model Hypotheses INNOVATION Measure Customers Continuous Delivery
  18. 18. Land grab opportunity Competitive Observe Orient Decide Act Move Customer Pain Point Analysis BIG DATA Model Hypotheses INNOVATION Measure Customers Continuous Delivery
  19. 19. Land grab opportunity Competitive Observe Orient Decide Act Move Customer Pain Point Analysis BIG DATA Plan Response JFDI Share Plans Model Hypotheses INNOVATION Measure Customers Continuous Delivery
  20. 20. Land grab opportunity Competitive Observe Orient Decide Act Move Customer Pain Point Analysis BIG DATA Plan Response JFDI Share Plans Model Hypotheses INNOVATION CULTURE Measure Customers Continuous Delivery
  21. 21. Land grab opportunity Competitive Observe Orient Decide Act Move Customer Pain Point Analysis BIG DATA Plan Response JFDI Share Plans Launch AB Test Automatic Deploy Incremental Features Model Hypotheses INNOVATION CULTURE Measure Customers Continuous Delivery
  22. 22. Land grab opportunity Competitive Observe Orient Decide Measure Customers Act Move Customer Pain Point Analysis BIG DATA Plan Response JFDI Share Plans Launch AB Test Automatic Deploy Incremental Features Model Hypotheses INNOVATION CULTURE CLOUD Continuous Delivery
  23. 23. Land grab opportunity Competitive Observe Orient Decide Measure Customers Act Move Customer Pain Point Analysis BIG DATA Plan Response JFDI Share Plans Launch AB Test Automatic Deploy Incremental Features Model Hypotheses INNOVATION CULTURE CLOUD Continuous Delivery
  24. 24. Land grab opportunity Competitive Observe Orient Decide Measure Customers Act Move Customer Pain Point Analysis BIG DATA Plan Response JFDI Share Plans Launch AB Test Automatic Deploy Incremental Features Model Hypotheses INNOVATION CULTURE CLOUD Continuous Delivery
  25. 25. Release Plan Developer Developer Developer Developer Developer QA Release Integration Ops Replace Old With New Release Monolithic service updates Works well with a small number of developers and a single language like php, java or ruby
  26. 26. Release Plan Developer Developer Developer Developer Developer Monolithic service updates QA Release Integration Ops Replace Old With New Release Bugs Works well with a small number of developers and a single language like php, java or ruby
  27. 27. Release Plan Developer Developer Developer Developer Developer Monolithic service updates QA Release Integration Ops Replace Old With New Release Bugs Bugs Works well with a small number of developers and a single language like php, java or ruby
  28. 28. @adrianco Breaking Down the SILOs
  29. 29. UX Dev Adm Prod Mgr @adrianco Breaking Down the SILOs QA DBA Sys Adm Net Adm SAN
  30. 30. UX Dev Adm Prod Mgr @adrianco Breaking Down the SILOs QA DBA Sys Adm Net Adm SAN Product Team Using Monolithic Delivery Product Team Using Monolithic Delivery
  31. 31. Product Team Using Monolithic Delivery Product Team Using Monolithic Delivery UX Dev Adm Prod Mgr @adrianco Breaking Down the SILOs QA DBA Sys Adm Net Adm SAN Product Team Using Microservices Product Team Using Microservices Product Team Using Microservices
  32. 32. Product Team Using Monolithic Delivery Product Team Using Monolithic Delivery UX Dev Adm Prod Mgr @adrianco Breaking Down the SILOs QA DBA Sys Adm Net Adm SAN Product Team Using Microservices Product Team Using Microservices Platform Team Product Team Using Microservices
  33. 33. Product Team Using Monolithic Delivery Product Team Using Monolithic Delivery UX Dev Adm Prod Mgr @adrianco Breaking Down the SILOs QA DBA Sys Adm Net Adm SAN Product Team Using Microservices Platform Team A P I Product Team Using Microservices Product Team Using Microservices
  34. 34. Product Team Using Monolithic Delivery Product Team Using Monolithic Delivery UX Dev Adm Prod Mgr @adrianco Breaking Down the SILOs QA DBA Sys Adm Net Adm SAN Product Team Using Microservices Platform Team A P I Product Team Using Microservices Product Team Using Microservices DevOps is a Re-Org!
  35. 35. Developer Developer Developer Developer Developer Old Release Still Running Release Plan Release Plan Release Plan Release Plan Immutable microservice deployment scales, is faster with large teams and diverse platform components
  36. 36. Developer Developer Developer Developer Developer Immutable microservice deployment scales, is faster with large teams and diverse platform components Old Release Still Running Release Plan Release Plan Release Plan Release Plan Deploy Feature to Production Deploy Feature to Production Deploy Feature to Production Deploy Feature to Production
  37. 37. Developer Developer Developer Developer Developer Immutable microservice deployment scales, is faster with large teams and diverse platform components Old Release Still Running Release Plan Release Plan Release Plan Release Plan Deploy Feature to Production Deploy Feature to Production Deploy Feature to Production Deploy Feature to Production Bugs
  38. 38. Developer Developer Developer Developer Developer Immutable microservice deployment scales, is faster with large teams and diverse platform components Old Release Still Running Release Plan Release Plan Release Plan Release Plan Deploy Feature to Production Deploy Feature to Production Deploy Feature to Production Deploy Feature to Production Bugs Deploy Feature to Production
  39. 39. Configure Configure Developer Developer Developer Release Plan Release Plan Release Plan Standardized portable container deployment saves time and effort Deploy Standardized Services https://hub.docker.com
  40. 40. Configure Configure Developer Developer Developer Release Plan Release Plan Release Plan Deploy Standardized Services Deploy Feature to Production Deploy Feature to Production Deploy Feature to Production Bugs Deploy Feature to Production Standardized portable container deployment saves time and effort https://hub.docker.com
  41. 41. @adrianco Developing at the Speed of Docker Developers • Compile/Build • Seconds Extend container • Package dependencies • Seconds PaaS deploy Containers • Docker startup • Seconds
  42. 42. @adrianco Developing at the Speed of Docker Developers • Compile/Build • Seconds Extend container • Package dependencies • Seconds PaaS deploy Containers • Docker startup • Seconds Speed is addictive, hard to go back to taking much longer to get things done
  43. 43. @adrianco What Happened? Rate of change increased Cost and size and risk of change reduced
  44. 44. Cloud Native Applications
  45. 45. Cloud Native A new engineering challenge Construct a highly agile and highly available service from ephemeral and assumed broken components
  46. 46. Inspiration
  47. 47. Inspiration
  48. 48. Inspiration
  49. 49. Inspiration
  50. 50. Inspiration
  51. 51. Inspiration
  52. 52. Inspiration
  53. 53. Inspiration
  54. 54. Inspiration
  55. 55. Inspiration
  56. 56. State of the Art in Cloud Native Microservice Architectures AWS Re:Invent : Asgard to Zuul https://www.youtube.com/watch?v=p7ysHhs5hl0 Resiliency at Massive Scale https://www.youtube.com/watch?v=ZfYJHtVL1_w Microservice Architecture https://www.youtube.com/watch?v=CriDUYtfrjs http://www.infoq.com/presentations/scale-gilt http://www.slideshare.net/mcculloughsean/itier-breaking-up-the-monolith-philly-ete http://www.infoq.com/presentations/Twitter-Timeline-Scalability http://www.infoq.com/presentations/twitter-soa http://www.infoq.com/presentations/Zipkin https://speakerdeck.com/mattheath/scaling-micro-services-in-go-highload-plus-plus-2014
  57. 57. @adrianco Traditional vs. Cloud Native Storage Business Logic Database Master Fabric Storage Arrays Database Slave Fabric Storage Arrays
  58. 58. @adrianco Traditional vs. Cloud Native Storage Business Logic Database Master Fabric Storage Arrays Database Slave Fabric Storage Arrays Business Logic Cassandra Zone A nodes Cassandra Zone B nodes Cassandra Zone C nodes Cloud Object Store Backups
  59. 59. @adrianco Traditional vs. Cloud Native Storage Business Logic Database Master Fabric Storage Arrays Database Slave Fabric Storage Arrays Business Logic Cassandra Zone A nodes Cassandra Zone B nodes Cassandra Zone C nodes Cloud Object Store Backups SSDs inside arrays disrupt incumbent suppliers
  60. 60. @adrianco Traditional vs. Cloud Native Storage Business Logic Database Master Fabric Storage Arrays Database Slave Fabric Storage Arrays Business Logic Cassandra Zone A nodes Cassandra Zone B nodes Cassandra Zone C nodes Cloud Object Store Backups SSDs inside ephemeral instances disrupt an entire industry SSDs inside arrays disrupt incumbent suppliers
  61. 61. @adrianco Traditional vs. Cloud Native Storage Business Logic Database Master Fabric Storage Arrays Database Slave Fabric Storage Arrays Business Logic Cassandra Zone A nodes Cassandra Zone B nodes Cassandra Zone C nodes Cloud Object Store Backups SSDs inside ephemeral instances disrupt an entire industry SSDs inside arrays disrupt incumbent suppliers See also discussions about “Hyper-Converged” storage
  62. 62. ● Edda - the “black box flight recorder” for configuration state ● Chaos Monkey - enforcing stateless business logic ● Chaos Gorilla - enforcing zone isolation/replication ● Chaos Kong - enforcing region isolation/replication ● Security Monkey - watching for insecure configuration settings ● See over 40 NetflixOSS projects at netflix.github.com ● Get “Technical Indigestion” trying to keep up with techblog.netflix.com @adrianco Trust with Verification
  63. 63. Netflix Cloud Native Benchmark
  64. 64. Autoscaled Ephemeral Instances at Netflix Largest services use autoscaled red/black code pushes Average lifetime of an instance is 36 hours Push Autoscale Up Autoscale Down
  65. 65. Netflix Automatic Code Deployment Canary Bad Signature Implemented by Simon Tuffs
  66. 66. Netflix Automatic Code Deployment Canary Bad Signature Implemented by Simon Tuffs
  67. 67. @adrianco Happy Canary Signature
  68. 68. @adrianco Speeding Up The Platform Datacenter Snowflakes • Deploy in months • Live for years
  69. 69. @adrianco Speeding Up The Platform Datacenter Snowflakes • Deploy in months • Live for years Virtualized and Cloud • Deploy in minutes • Live for weeks
  70. 70. @adrianco Speeding Up The Platform Datacenter Snowflakes • Deploy in months • Live for years Virtualized and Cloud • Deploy in minutes • Live for weeks Docker Containers • Deploy in seconds • Live for minutes/hours
  71. 71. @adrianco Speeding Up The Platform Datacenter Snowflakes • Deploy in months • Live for years Virtualized and Cloud • Deploy in minutes • Live for weeks Docker Containers • Deploy in seconds • Live for minutes/hours AWS Lambda • Deploy in milliseconds • Live for seconds
  72. 72. AWS Lambda • Deploy in milliseconds • Live for seconds @adrianco Speeding Up The Platform Datacenter Snowflakes • Deploy in months • Live for years Virtualized and Cloud • Deploy in minutes • Live for weeks Docker Containers • Deploy in seconds • Live for minutes/hours Speed enables and encourages new microservice architectures
  73. 73. With AWS Lambda compute resources are charged by the 100ms, not the hour First 1,000,000 node.js executions/month are free First 400,000 GB-seconds of RAM-CPU are free
  74. 74. Monitoring Requirements Metric resolution microseconds Metric update rate 1 second Metric to display latency less than human attention span (<10s)
  75. 75. @adrianco Low Latency SaaS Based Monitors www.vividcortex.com and www.boundary.com
  76. 76. Adrian’s Tinkering Projects Model and visualize microservices Simulate interesting architectures ! See github.com/adrianco/spigo Simulate Protocol Interactions in Go ! See github.com/adrianco/d3grow Dynamic visualization
  77. 77. Cost Optimization
  78. 78. Capacity Optimization for a Single System Bottleneck Upper Spec Limit ! When demand probability exceeds USL by 4.0 sigma scale up resource to maintain low latency Lower Spec Limit ! When demand probability is below USL by 3.0 sigma scale down resource to save money Documentation on Capability Plots To get accurate high dynamic range histograms see http://hdrhistogram.org/ Slideshare: 2003 Presentation on Capacity Planning Methods See US Patent: 7467291
  79. 79. But interesting systems don’t have a single bottleneck nowadays…
  80. 80. But interesting systems don’t have a single bottleneck nowadays…
  81. 81. @adrianco How is Cost Measured?
  82. 82. 1 Bottom Up 2 Product 3 Top Down
  83. 83. @adrianco Bottom Up Costs Add up the cost to buy and operate every component 1
  84. 84. @adrianco Product Cost Cost of delivering and maintaining each product 2
  85. 85. @adrianco Top Down Costs Divide total budget by the number of components 3
  86. 86. @adrianco Top Down vs. Bottom Up Will never match! Hidden Subsidies vs. Hidden Costs 3 1
  87. 87. @adrianco Agile Team Product Profit Value minus costs Time to value ROI, NPV, MMF Profit center 2 $
  88. 88. @adrianco What about cloud costs?
  89. 89. @adrianco Cloud Native Cost Optimization Optimize for speed first Turn it off! Capacity on demand Consolidate and Reserve Plan for price cuts FOSS tooling $$ $
  90. 90. @adrianco The Capacity Planning Problem
  91. 91. @adrianco Best Case Waste Product(Launch(Agility(2(Rightsized( Pre2Launch( Build2out( Tes9ng( Launch( Growth( Growth( Demand( Cloud( Datacenter( $( Cloud capacity used is maybe half average DC capacity
  92. 92. @adrianco Failure to Launch Product(Launch(-(Under-es1mated( Pre-Launch Build-out Testing Launch Growth Growth Mad scramble to add more DC capacity during launch phase outages
  93. 93. @adrianco Over the Top Losses Product(Launch(Agility(–(Over6es8mated( Pre-Launch Build-out Testing Launch Growth Growth $ Capacity wasted on failed launch magnifies the losses
  94. 94. @adrianco Turning off Capacity Off-peak production Test environments Dev out of hours When%you%turn%off%your%cloud%resources,% you%actually%stop%paying%for%them% Dormant Data Science
  95. 95. @adrianco Containerize Test Environments Snapshot or freeze Fast restart needed Persistent storage 40 of 168 hrs/wk Bin-packed containers shippable.com saved 70%
  96. 96. @adrianco Seasonal Savings 50% Savings 1 5 9 13 17 21 25 29 33 37 41 45 49 Web Servers Week Optimize during a year Weekly&CPU&Load&
  97. 97. @adrianco Autoscale the Costs Away 50%+%Cost%Saving% Scale%up/down% by%70%+% Move%to%Load=Based%Scaling%
  98. 98. @adrianco Daily Duty Cycle Instances' Business'Throughput' Reactive Autoscaling saves around 50% Predictive Autoscaling saves around 70% See Scryer on Netflix Tech Blog
  99. 99. @adrianco Underutilized and Unused AWS$Support$–$Trusted$Advisor$–$ Your$personal$cloud$assistant$
  100. 100. @adrianco Clean Up the Crud Other&simple&op-miza-on&-ps& • Don’t&forget&to…& – Disassociate&unused&EIPs& – Delete&unassociated&Amazon& EBS&volumes& – Delete&older&Amazon&EBS& snapshots& – Leverage&Amazon&S3&Object& Expira-on& & Janitor&Monkey&cleans&up&unused&resources&
  101. 101. @adrianco Total Cost of Oranges When%Comparing%TCO…! Make!sure!that! you!are!including! all!the!cost!factors! into!considera4on! Place% Power% Pipes% People% Pa6erns%
  102. 102. @adrianco Total Cost of Oranges When%Comparing%TCO…! Make!sure!that! you!are!including! all!the!cost!factors! into!considera4on! Place% Power% Pipes% People% Pa6erns% How much does datacenter automation software and support cost per instance?
  103. 103. @adrianco When Do You Pay? bill Run My Stuff Now Next Month Datacenter Up Front Costs Lease Building Ages Ago Install AC etc Rack & Stack Private Cloud SW
  104. 104. Cost Model Comparisons AWS has most complex model • Both highest and lowest cost options! CPU/Memory Ratios Vary • Can’t get same config everywhere Features Vary • Local SSD included on some vendors, not others • Network and storage charges also vary
  105. 105. @adrianco Digital Ocean Flat Pricing Hourly Price ($0.06/hr) Monthly Price ($40/mo) $ No Upfront $ No Upfront $0.060/hr $0.056/hr $1555/36mo $1440/36mo Savings 7% Prices on Dec 7th, for 2 Core, 4G RAM, SSD, purely to show typical savings
  106. 106. @adrianco Google Sustained Usage Full Price Without Sustained Usage Typical Sustained Usage Each Month Full Sustained Usage Each Month $ No Upfront $ No Upfront $ No Upfront $0.063/hr $0.049/hr $0.045/hr $1633/36mo $1270/36mo $1166/36mo Savings 22% 29% Prices on Dec 7th, for n1.standard-1 (1 vCPU, 3.75G RAM, no disk) purely to show typical savings
  107. 107. @adrianco AWS Reservations On Demand No Upfront 1 year Partial Upfront 3 year All Upfront 3 year $ No Upfront $No Upfront $337 Upfront $687 Upfront $0.070/hr $0.050/hr $0.0278/hr $0.00/hr $1840/36mo $1314/36mo $731/36mo $687/36mo Savings 29% 60% 63% Prices on Dec 7th, for m3.medium (1 vCPU, 3.75G RAM, SSD) purely to show typical savings
  108. 108. @adrianco Blended Benefits Mix$and$Match$Reserved$Types$and$On4Demand$ Instances( On8Demand Light&RI Light&RI Light&RI Light&RI Heavy&Utilization&Reserved Instances Days(of(Month( 12 10 8 6 4 2 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 On Demand Partial Upfront All Upfront
  109. 109. @adrianco Consolidated Cost Cutting Consolidated+Billing:+Single+payer+for+a+group+of+ accounts+ • One$Bill+for+mul7ple+accounts+ • Easy$Tracking$of+account+ charges+(e.g.,+download+CSV+of+ cost+data)+ • Volume$Discounts+can+be+ reached+faster+with+combined+ usage+ • Reserved$Instances$are+shared+ across+accounts+(including+RDS+ Reserved+DBs)+
  110. 110. @adrianco Production Priority Over%Reserve(the(Produc0on(Environment( Produc0on(Env.( Total#Capacity# Account( 100#Reserved# QA/Staging(Env.( Account( 0#Reserved## Perf(Tes0ng(Env.( Account( 0#Reserved## Development(Env.( Account( 0#Reserved# Storage(Account( 0#Reserved#
  111. 111. @adrianco Actual Reservation Usage Consolidated+Billing+Borrows+Unused+Reserva4ons+ Produc4on+Env.+ Total#Capacity# Account+ 68#Used# QA/Staging+Env.+ Account+ 10#Borrowed# Perf+Tes4ng+Env.+ Account+ 6#Borrowed## Development+Env.+ Account+ 12#Borrowed# Storage+Account+ 4#Borrowed#
  112. 112. @adrianco Consolidated Reservations Burst capacity guarantee Higher availability with lower cost Other accounts soak up any extra Monthly billing roll-up Capitalize upfront charges! But: Fixed location and instance type
  113. 113. @adrianco Re-sell Reservations Reserved'Instance'Marketplace' Buy$a$smaller$term$instance$ Buy$instance$with$different$OS$or$type$ Buy$a$Reserved$instance$in$different$region$ Sell$your$unused$Reserved$Instance$ Sell$unwanted$or$over;bought$capacity$ Further$reduce$costs$by$op>mizing$
  114. 114. @adrianco Use EC2 Spot Instances Cloud native dynamic autoscaled spot instances ! Real world total savings up to 50%
  115. 115. @adrianco Right Sizing Instances Fit the instance size to the workload
  116. 116. @adrianco Technology Refresh Older m1 and m2 families • Slower CPUs • Higher response times • Smaller caches (6MB) • Oldest m1.xlarge • 15G/8.5ECU/35c 23ECU/$ • Old m2.xlarge • 17G/6.5ECU/25c 26ECU/$ New m3 family • Faster CPUs • Lower response times • Larger caches (20MB) • Java perf ratio > ECU • New m3.xlarge • 15G/13ECU/28c 46ECU/$ • 77% better ECU/$ • Deploy fewer instances
  117. 117. @adrianco Data Science for Free Follow%the%Customer%(Run%web%servers)%during%the%day% No.%of%Reserved% Instances% Follow%the%Money%(Run%Hadoop%clusters)%at%night% 16 14 12 10 8 6 4 2 0 Mon Tue Wed Thur Fri Sat Sun No#of#Instances#Running# Week Auto7Scaling7Servers Hadoop7Servers
  118. 118. @adrianco Six Ways to Cut Costs Building#Cost>Aware#Cloud#Architectures# #1#Business#Agility#by#Rapid#Experimenta8on#=#Profit# #2#Business>driven#Auto#Scaling#Architectures#=#Savings## #3#Mix#and#Match#Reserved#Instances#with#On>Demand#=#Savings# #4#Consolidated#Billing#and#Shared#Reserva8ons#=#Savings# #5#Always>on#Instance#Type#Op8miza8on#=#Recurring#Savings# #6#Follow#the#Customer#(Run#web#servers)#during#the#day# Follow#the#Money#(Run#Hadoop#clusters)#at#night# Credit to Jinesh Varia of AWS for this summary
  119. 119. Data shown is purely illustrative @adrianco Expect Prices to Drop Three Years Halving Every 18mo = maybe 40% over-all savings 100 75 50 25 0 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
  120. 120. @adrianco Compounded Savings
  121. 121. 100 Traditional application using AWS heavy use reservations 30 30 25 @adrianco Lift and Shift Compounding 100 75 50 25 0 70 70 70 Base Price Rightsized Seasonal Daily Scaling Reserved Tech Refresh Price Cuts Base price is for capacity bought up-front
  122. 122. 100 Traditional application using AWS heavy use reservations 30 30 25 @adrianco Lift and Shift Compounding 100 75 50 25 0 70 70 70 Base Price Rightsized Seasonal Daily Scaling Reserved Tech Refresh Price Cuts Base price is for capacity bought up-front
  123. 123. 100 Traditional application using AWS heavy use reservations 30 30 25 @adrianco Lift and Shift Compounding 100 75 50 25 0 70 70 70 Base Price Rightsized Seasonal Daily Scaling Reserved Tech Refresh Price Cuts Base price is for capacity bought up-front
  124. 124. 100 Traditional application using AWS heavy use reservations 30 30 25 @adrianco Lift and Shift Compounding 100 75 50 25 0 70 70 70 Base Price Rightsized Seasonal Daily Scaling Reserved Tech Refresh Price Cuts Base price is for capacity bought up-front
  125. 125. 100 Cloud native application partially optimized light use reservations 20 15 25 @adrianco Conservative Compounding 100 75 50 25 0 35 50 70 Base Price Rightsized Seasonal Daily Scaling Reserved Tech Refresh Price Cuts
  126. 126. 100 Cloud native application partially optimized light use reservations 20 15 25 @adrianco Conservative Compounding 100 75 50 25 0 35 50 70 Base Price Rightsized Seasonal Daily Scaling Reserved Tech Refresh Price Cuts
  127. 127. 100 Cloud native application partially optimized light use reservations 20 15 25 @adrianco Conservative Compounding 100 75 50 25 0 35 50 70 Base Price Rightsized Seasonal Daily Scaling Reserved Tech Refresh Price Cuts
  128. 128. 100 Cloud native application partially optimized light use reservations 20 15 25 @adrianco Conservative Compounding 100 75 50 25 0 35 50 70 Base Price Rightsized Seasonal Daily Scaling Reserved Tech Refresh Price Cuts
  129. 129. 100 Cloud native application partially optimized light use reservations 20 15 25 @adrianco Conservative Compounding 100 75 50 25 0 35 50 70 Base Price Rightsized Seasonal Daily Scaling Reserved Tech Refresh Price Cuts
  130. 130. 100 Cloud native application fully optimized autoscaling mixed reservation use costs 4% of base price over three years! 12 8 6 4 @adrianco Agressive Compounding 100 75 50 25 0 25 50 Base Price Rightsized Seasonal Daily Scaling Reserved Tech Refresh Price Cuts
  131. 131. @adrianco Cost Monitoring and Optimization
  132. 132. @adrianco Final Thoughts Turn off idle instances Clean up unused stuff Optimize for pricing model Assume prices will go down Go cloud native to be fast and save Complex dynamic control issues!
  133. 133. @adrianco Further Reading ● Battery Ventures Blog http://www.battery.com/powered ● Adrian’s Blog http://perfcap.blogspot.com and Twitter @adrianco ● Slideshare http://slideshare.com/adriancockcroft ! ● Monitorama Opening Keynote Portland OR - May 7th, 2014 - Video available ● GOTO Chicago Opening Keynote May 20th, 2014 - Video available ● Qcon New York – Speed and Scale - June 11th, 2014 - Video available ● Structure - Cloud Trends - San Francisco - June 19th, 2014 - Video available ● GOTO Copenhagen/Aarhus – Denmark – Sept 25th, 2014 ● DevOps Enterprise Summit - San Francisco - Oct 21-23rd, 2014 - Videos available ● GOTO Berlin - Germany - Nov 6th, 2014 ● Dockercon Europe - Amsterdam - Microservices - December 4th, 2014 See www.battery.com for a list of portfolio investments

×