Se ha denunciado esta presentación.
Utilizamos tu perfil de LinkedIn y tus datos de actividad para personalizar los anuncios y mostrarte publicidad más relevante. Puedes cambiar tus preferencias de publicidad en cualquier momento.

AWS Summit Kuala Lumpur Keynote with Stephen Orban - Head of Enterprise Strategy

1.946 visualizaciones

Publicado el

AWS Summit Kuala Lumpur, Keynote address with Stephen Orban - Head of Enterprise Strategy, Amazon Web Services.
May 21st 2015

Publicado en: Tecnología

AWS Summit Kuala Lumpur Keynote with Stephen Orban - Head of Enterprise Strategy

  1. 1. Welcome @stephenorban! Stephen Orban, Head of Enterprise Strategy, AWS!
  2. 2. How Is The AWS Business Progressing?
  3. 3. 2012 2013 2014 Amazon S3 Usage 102% YoY increase in data transfer to and from S3 (Q4 2014 vs Q4 2013, not including Amazon use)
  4. 4. 2011 2012 2013 2014 Amazon EC2 Usage 93% YoY increase in EC2 instance usage (Q4 2014 vs Q4 2013, not including Amazon use)
  5. 5. 2008 2009 2010 2011 2012 2013 2014 Over 1 Million Active Customers “Active customer” is defined as a non-Amazon customer with AWS account usage activity in the past month, including the free tier
  6. 6. Startup Customers Meerkat!
  7. 7. Enterprise Customers
  8. 8. Public Sector Customers
  9. 9. Premier Consulting Partners
  10. 10. ISV Partners
  11. 11. Gartner  “Magic  Quadrant  for  Cloud  Infrastructure  as  a  Service,”  Lydia  Leong,  Douglas  Toombs,  Bob  Gill,  Gregor  Petri,  Tiny  Haynes,  May  28,  2014.  This  Magic  Quadrant  graphic  was  published  by  Gartner,  Inc.  as  part  of  a  larger  research  note  and  should  be  evaluated  in  the  context  of  the  enMre  report.  The  Gartner  report  is  available    at  hNp://aws.amazon.com/resources/analyst-­‐ reports/.  Gartner  does  not  endorse  any  vendor,  product  or  service  depicted  in  its  research  publicaMons,  and  does  not  advise  technology  users  to  select  only  those  vendors  with  the  highest  raMngs.  Gartner  research  publicaMons  consist  of  the  opinions  of  Gartner's  research  organizaMon  and  should  not  be  construed  as  statements  of  fact.  Gartner  disclaims  all  warranMes,  expressed   or  implied,  with  respect  to  this  research,  including  any  warranMes  of  merchantability  or  fitness  for  a  parMcular  purpose.   Gartner Magic Quadrant Cloud Infrastructure as a Service
  12. 12. The fastest growing multi-billion dollar enterprise IT company in the world Amazon Web Services > 40% Salesforce 26% VMware 15% Google 15% CDW 12% Microsoft 8% Cisco 7% SAP 7% EMC 6% Oracle 0% Teradata -1% CA -3% Symantec -4% HP -5% IBM -12% Multi-billion dollar enterprise IT vendors Most recent YoY revenue growth rate
  13. 13. Cloud Has Become The New Normal
  14. 14. What Are The Patterns Of This New Normal?
  15. 15. Speed Is Not Just For Start-ups: Companies of All Sizes Will Move Faster Than Ever Before 1
  16. 16. It’s Hard To Stay Competitive Today Without The Cloud
  17. 17. Old World Large upfront capital investment Basic compute and storage only Responsible for feature upgrades Slow to get new capabilities Low, variable cost Broad and deep platform New features arrive daily Ready to use
  18. 18. 24 48 61 82 159 280 516 2008 2009 2010 2011 2012 2013 2014 AWS Rapid Pace Of Innovation 516 major new features and services launched in 2014
  19. 19. Enterprises Are Using The Cloud For New Apps & Digital Transformation StatCast App platform Fast Moving Consumer Goods E-commerce Digital Personal Finance Web Telecommunications KTC
  20. 20. DEXTER LAU Group Executive Director KIM TECK CHEONG CONSOLIDATED SDN BHD!
  21. 21. ²  Founded in 1938 | 76 Years of experience! ! ²  One of the largest Consumer Package Goods (CPG) Distributor! in East Malaysia with a market size of RM6.0 billion! ²  Business presence in:! o  Sabah - Kota Kinabalu, Labuan, Tawau, Sandakan! o  Sarawak – Sibu, Kuching, Miri! ² 3769 sales and distribution points covering over 80 districts ! A LITTLE ABOUT US!
  22. 22. BUSINESS FOCUS! Provision of market access and coverage of CPG! Manufacturing of bakery products! Distribution and warehousing of third party brands of CPG! Distribution and warehousing of own brands of CPG! Manufacturing and distribution of own brand of bakery products! Customer Base! Wholesalers !Retailers! ! Food Service Operators ! Distributors! WHAT WE DO!
  23. 23. ²  Leading international brands! ! ²  227 brands ! ! ²  7957 products! F&B ! Personal Care! OTC drugs & health ! supplements!Baby Care !Household!
  24. 24. ²  Frozen and dry goods! ²  Bakery products! ! ²  Producing approximately 35,000 CREAMOS ! bread rolls daily! ! ! OWN HOUSEHOLD BRANDS!
  25. 25. CHALLENGES ! ! ! Ø  Business expansion ! Ø  Unacceptable service delay from the local data centre! ! Ø  High cost to manage the servers in-house and storage! Ø  Data back-up! Ø  Advancing technology and e-commerce trend ! ! Traditional v/s Technology!
  26. 26. ! Ø  Provision of flexibility for ongoing business growth! ! Ø  Reliability and security ! ! Ø  Support routine back-ups and disaster recovery! Ø  Increase productivity ! Ø  Technical and operational ! Ø  Cost management and efficiency! SEEKING FOR SOLUTIONS! ?!
  27. 27. MIGRATION TO CLOUD! Security! Scalability! Service reliability! Agility! Secure & protect ! business application! Support routine back-ups and disaster recovery!
  28. 28. THANK YOU KIM TECK CHEONG CONSOLIDATED SDN BHD!
  29. 29. Start-ups Build Businesses From Scratch In The Cloud 2
  30. 30. No legacy Lower cost structureNo dependencies S Move quickly Building All Applications In The Cloud
  31. 31. Disrupt Long Standing Industries, Quickly Hotels Storage Gaming Groceries Taxi Booking
  32. 32. A Startup’s Journey in the Cloud Wei Zhu CTO
  33. 33. •  Location based smartphone ride booking and dispatching service •  Our vision is to revamp the South East Asian public transportation •  Founded by 2 HBS Malaysian grads •  $330M USD in funding to date
  34. 34. Old world vs New World Infrastructure
  35. 35. Rush hour at
  36. 36. 7 bookings every second
  37. 37. ●  6 countries comprising of 20 cities in Southeast Asia ●  80k drivers active ●  iOS and Android apps have been downloaded 4.4 Million times to date 100% running on AWS Cloud
  38. 38. Not Enough Engineers? http://grab.careers
  39. 39. What are the next steps on AWS / vision of cloud Database scaling technology is a common bottleneck. Aurora will let us scale faster and meet growth demand. Data from millions of mobile devices will let us match supply and demand, in real time. This is achievable with deep integration with Kinesis and EMR. Experiments on task isolation and predictive load management with containerized environments with EC2 Container Service. Amazon Kinesis! Amazon Elastic MapReduce!
  40. 40. Thank you
  41. 41. Customers Want Access To The Complete Package 3
  42. 42. Key Components Of Agility + = Quick to provision Don’t have to reinvent the wheel Vast infrastructure technology platform
  43. 43. More Functionality Than Any Other Infrastructure Provider Enterprise Applications Administration & Security Core Services Platform Services Infrastructure AWS Marketplace
  44. 44. More Functionality Than Any Other Infrastructure Provider Enterprise Applications Administration & Security Core Services Platform Services Infrastructure AWS Marketplace
  45. 45. “IT organizations cannot treat cloud IaaS providers like commodities” Lydia Leong
  46. 46. Customers Want To Use The Leading Platform Largest customer & partner community Most functionality Longest experience
  47. 47. Customers Want To Use The Leading Platform Largest customer & partner community Most functionality Longest experience
  48. 48. Jack Sen VP of Technology jacksen.pang@123rf.com linkedin.com/in/jacksen
  49. 49. ●  40 million stock images, vectors, footage and audio file ●  Top 5 global microstock site ●  No. 1 microstock site from Asia ●  Ranked top 900 websites *according to Alexa.com ●  15.4M visits per month *according to similarWeb.com
  50. 50. Datacenter Web, Search Assets Cache Database Data Warehouse DNS Cache and Assets Storage Amazon EC2! Amazon S3! Amazon CloudFront! Amazon RDS! Amazon Redshift! Amazon Route 53!
  51. 51. 80 to 120 Amazon EC2 Instances ●  batch image resizing ●  real time image processing and watermark stamping ●  image fingerprinting and recognition ●  video and audio transcoding ●  search engine 120TB of Amazon S3 Storage ●  Images, Footage and Audio Files 106TB Amazon Cloudfront Traffic Per Month How are we using AWS
  52. 52. Benefits of Amazon Cloudfront Traditional CDN ●  CDN only for the big players. ●  Big commitment and annual contract needs to be signed. ●  We set up our own CDN by buying servers in various regions. Amazon Cloudfront CDN ●  AWS Pay-as-you-go pricing ●  Deploy an Amazon Cloudfront distribution in minutes ●  No long term contract needed, pay for what you use.
  53. 53. Benefits of Amazon EC2 Flexibility ●  Scale your platform elastically as your business grows by provisioning infrastructure on demand. ●  dealing with ‘hardware’ failure is easier. Just spin up a new instance and restore from backup. Even better if auto deploy and failover infra. ●  Auto scaling in response to load event ●  Makes pre launch capacity planning easy Cost Effectiveness ●  Flexibility Many different choices of instance type Only pay for what you use Choices of on-demand, spot or dedicate instances ●  Operations Efficiency No need for engineers to maintain hardware Instead they concentrate on planning,scaling, auto failover, etc
  54. 54. Advice Aim for Flexibility Make your infra horizontally scalable (ability to work on multiple ‘smaller’ nodes). Go Hybrid It doesn’t have to be an all or nothing decision. You can go hybrid first. Use AWS Direct Connect or VPN to connect your datacenter and your virtual stack. Amazon Cloudfront Ensure that you have adequate origin capacity to satisfy your edge demand, or better still, use Amazon S3 as your origin.
  55. 55. Future with AWS ●  Amazon Elastic File System ●  Amazon Cloudsearch ●  Amazon Machine Learning ●  Amazon Kinesis
  56. 56. The Old Shackles Are Loosening 4
  57. 57. Old World Databases Lock inProprietary Punitive licensing with limited flexibility Contract Very expensive S
  58. 58. It’s Why Customers Have Been Moving To More Open, Customer Friendly, Less Expensive Database Engines
  59. 59. …But, to get comparable performance to proprietary databases is hard
  60. 60. MySQL compatible Available, durable, and fault tolerant 5X better performance of high-end MySQL database Highly scalable and secure Thousands of customers in the preview 1/10th the cost of the leading commercial database solutions Amazon Aurora
  61. 61. Reinvention Is Continuous 5
  62. 62. For Example, Consider Compute… m1.small General Purpose (M3) Compute Optimized (C4) Memory Optimized (R3) GPU Optimized (G2) Storage Optimized (D2) IO Optimized (I2) Low cost, burst-able performance (T2)
  63. 63. Building With Smaller Blocks Quicker to build Lower costEasier to adapt and update
  64. 64. Amazon EC2 Container Service Launch and terminate Docker containers Across a cluster of EC2 instances Mount persistent volumes at launch Private Docker repositories
  65. 65. EC2 Container Service Is Now Generally Available Available to all customers Management console Geographic expansion US East, US West, EU West, Sydney and Japan CloudTrail integration aws.amazon.com/ecs
  66. 66. A New Service Scheduler For Long-running Applications Scale containers up and down Distribute traffic via ELB Automatically recover unhealthy containers Deploy updated containers & definitions
  67. 67. Shrinking Compute To Atomic Scale With AWS Lambda
  68. 68. AWS Lambda: An Event Driven Computing Service Events from AWS services Cloud Functions in Node.js Automatic execution with no servers to provision Cloud functions in Java
  69. 69. How Are Customers Using AWS Lambda? Data triggers Stream processing Indexing & synchronization IoT Server-free back-end
  70. 70. Moving To The Cloud Is Not A Binary Decision 6
  71. 71. On-premises Cloud? What Role Does Hybrid IT Play?
  72. 72. Services For Hybrid IT On-premises AWS Virtual Private Cloud Direct Connect AWS Config Storage Gateway Integrated networking Directory Service Identity Federation Integrated identity Integrated management AWS CloudTrail Backups OpsWorks Deployment CodeDeploy vCenter & System Center plugins
  73. 73. Hybrid IT Is Part Of The Journey, Not The Destination
  74. 74. Customers & Partners Are Becoming Cloud-First 7
  75. 75. Bold Move In 2009
  76. 76. Cloud-First with AWS: Now Common in 2015 Enterprise customers ISV partners
  77. 77. “Friends don’t let friends build data centers” Charles Philips, CEO
  78. 78. Fighting Gravity
  79. 79. Thank You
  80. 80. Glenn Gore Head of Architecture, Asia Pacific Amazon Web Services
  81. 81. Companies Will Use Data More Expansively Than At Any Other Point In History
  82. 82. It’s never been easier and less expensive to collect, store, analyze & share data
  83. 83. COLLECT | STORE | ORGANIZE | ANALYZE | SHARE
  84. 84. COLLECT | STORE | ORGANIZE | ANALYZE | SHARE
  85. 85. Introducing The Amazon Elastic File System A fully managed file system for EC2 Highly available and durable SSD-based Automatic data replication across AZs (like S3) Very high throughput and low latency Grows to petabyte scale, elastically
  86. 86. Simple: An Easy To Use, Fully Managed File System Create a scalable file system in seconds Seamless integration with existing apps Supports NFSv4 Deploy No upfront or minimum costs
  87. 87. Elastic: A File System That Grows And Shrinks, Automatically Dynamically and automatically resizes as data is added or removed No need to provision storage capacity or performance Pay only for the storage space you use
  88. 88. Scalable: A File System For Virtually Any Workload Grow to petabyte scale Throughput and IOPS scale automatically Support for thousands of concurrent NFS connections Consistent low latency access
  89. 89. Amazon EFS Preview Summer 2015 Register for the preview today: aws.amazon.com/efs
  90. 90. COLLECT | STORE | ORGANIZE | ANALYZE | SHARE
  91. 91. More And More Customers Are Adopting Machine Learning Email targeting Recommendations Social news Digital health Language processing Auto-scaling
  92. 92. A Legacy Of Machine Learning At Amazon “Customers who bought this also bought…” Natural language processing Search Vision systems in Fulfillment Centers
  93. 93. How do we provide all internal teams with access to machine learning?
  94. 94. The Spark For Hundreds Of New Machine Learning Applications Counterfeit goods detection Customer adoption models Item classification Sales lead ranking Search intent Demand estimation Customer support Display ads
  95. 95. Announcing Amazon Machine Learning A Fully Managed Machine Learning Service for Developers
  96. 96. Prediction Data-driven Development Real time & dashboards Kinesis EC2 & Lambda Analysis & reporting Redshift, S3 & EMR Mobile Analytics
  97. 97. Automatically find patterns in existing data & make confident predictions on new data Predictive Models And Machine Learning
  98. 98. Machine Learning By Example Based on what you know about a customer: Will they use your product?
  99. 99. Machine Learning By Example Based on what you know about a customer: Will they use your product? Based on what you know about an order: Is that order fraudulent?
  100. 100. Based on what you know about a news article: What other articles are interesting? Machine Learning By Example Based on what you know about a customer: Will they use your product? Based on what you know about an order: Is that order fraudulent?
  101. 101. Statistics Model building Cross-validation Algorithms Transformation Machine Learning Challenges for Developers
  102. 102. Statistics Model building Cross-validation Algorithms Transformation In production Machine Learning Challenges for Developers
  103. 103. Machine Learning Challenges for Developers Statistics Model building Cross-validation Algorithms Transformation In production At scale
  104. 104. Easily create machine learning models Visualize and optimize models Put models into production in seconds Battle-hardened technology Introducing Amazon Machine Learning
  105. 105. Train and optimize models on GBs of data Batch process predictions Real-time prediction API in one-click No servers to provision or manage Easy to Use, High Performance
  106. 106. Amazon Machine Learning is integrated across AWS data stores S3, Redshift and RDS Wealth of data now available to ML Data & Amazon Machine Learning
  107. 107. Grouping Customers: An Experiment 45 days 2 Developers 92% accuracy
  108. 108. 20 minutes 1 Developer 92% accuracy Grouping Customers: An Experiment 45 days 2 Developers 92% accuracy
  109. 109. A Fully Managed Machine Learning Service for Developers Amazon Machine Learning
  110. 110. Fighting Gravity
  111. 111. Thank You

×