AP A M I ITA AWSCd36cz9buwru1tt.cloudfront.net/Migrating_Internal_ IT_Apps_Hunter... · Internal...
Transcript of AP A M I ITA AWSCd36cz9buwru1tt.cloudfront.net/Migrating_Internal_ IT_Apps_Hunter... · Internal...
A PRACTICAL APPROACH TO MIGRATING INTERNAL IT APPS TO THE AWS CLOUD Jerry Hunter, VP Amazon IT
• Now: VP of Corporate ApplicaHons (1.75 yrs) – Includes: All Finance and HR systems, developer tools, knowledge management, SOX, personal producHvity(laptops, phones, email, calendar, printers, etc)
• Before: • VP IT OperaHons at SUNW (1.5 yrs)
• VP IT Infrastructure design SUNW(1 yr) • Sr Dir, Java Enterprise System, SUNW(2 yrs)
MY ROLE AT AMAZON
• Make Amazon more producHve – Remove waste from our value stream
• Improve hardware uHlizaHon rates – Hold just the capacity needed
• Drive so]ware efficiency – Deliver uHlity bills directly applicaHon owners
• Reduce our Total Cost of Ownership
MOTIVATING FACTORS
OUR LEGACY DEPLOYMENT MODEL
Internal Users Internal Applications Running on Internal Servers
Internal Network
OUR AMAZON VPC DEPLOYMENT MODEL
We extend our internal network into the cloud, securely hosting internal applications on EC2 within a VPC.
Employees
• Do you have or want clear visibility into your infrastructure usage (network, CPU, memory, storage)? Can you predict future demand?
• Are you considering infrastructure expansion? • Are you building new applicaHons? • Does maintaining business advantage have anything to do with Hme to market?
• Are you using SOA today? Would you like control over your hardware usage from within your so]ware?
WHEN IS CLOUD COMPUTING A FIT?
• Amazon is an enterprise customer of AWS
• Avoid custom, Amazon-‐specific soluHons • Meet or exceed exisHng metrics for availability, latency, etc.
• Earn customer trust with strict adherence to enterprise security requirements
• Emphasize frugality
CORE TENETS OF AMAZON IT MIGRATION
• Create Program Infrastructure
• System Assessment – Data ClassificaHon – Understand our Systems Beeer
– ApplicaHon CriHcality (Availability, SLAs) – Dependencies – Compliance Requirements
– HW Component Usage (Disk, I/O, Memory) – Current TCO
AMAZON IT PRE-‐MIGRATION CHECKLIST
• Move your encrypted data to Amazon S3 now
• Pilot applicaHons in Amazon EC2 – Evaluate latency and availability – Understand operaHonal procedures (provisioning, etc)
• Build phased migraHon plan • Work with third-‐party vendors for AWS cerHficaHon and licensing
RECOMMENDED IT APPROACH
CASE STUDY 1: BROADCAST.AMAZON.COM
CASE STUDY 1: BROADCAST (ARCHITECTURE)
Users
Dynamically Scaled Video Rendering
Unlimited Video Storage
Web Front-End
MySql
Encryption
Rendering Job Manager
Amazon Internal Network
• 900 hrs of video encoded in 5 days (would have taken 112 days)
• 2 software engineers, 3 weeks
• Videos stream within a Flash-‐based embedded player
• AutomaHc encoding pipeline to re-‐render legacy and new video hosted within Amazon EC2 – Over 900 hours of video re-‐encoded – Ordinarily, 900 hours * 3 hours to encode per video = 112 days – With Amazon EC2, we were able to parallelize encoding and finish
within one week
• Storing and serving “unlimited” video using Amazon S3 • Massive producHvity increase
– 2 so]ware engineers, 3 weeks, 1 applicaHon – Engineers empowered to build the soluHon on their own, no
requisiHon process involved
BROADCAST, THE AWS VERSION
CASE STUDY 2: BMC REMEDY MID-‐TIER
Amazon EC2 instances hosting part of Remedy mid-tier server fleet spread across three data centers
Employees
DC1
DC2
DC3
Load Balancer
Amazon Internal Network
• Mission CriHcal Apps on Amazon EC2 – BMC Remedy AR System on AWS
– So]ware deployment system integraHon with AWS Auto Scaling – ElasHc build fleet
• Business CriHcal Apps on Amazon EC2 – Test environments on-‐demand
• Broadened use of Amazon S3 – File archiving – logs, invoice images, reports, etc
– Desktop backups – Key -‐ blob storage
SOME CURRENT & FUTURE PROJECTS
VENDOR COLLABORATION
eBusiness Suite
Remedy AR System
• AdapHng license models
• Expanding AWS support of vendor requirements • TesHng performance against vendors’ benchmarks
• Scalability – Rapid, on-‐demand elasHcity versus standard old process for adding
new servers to your data center
• Development Lifecycle – Provisioning development capacity in minutes versus weeks
– Exposing applicaHon operaHng costs drives greater ownership and beeer decision making by so]ware developers
– Empowering engineers to innovate by removing barriers
• Security – Data centers hardened by Amazon experts versus spending to harden
your own
– Abstract the hardware from you
EVALUATING CLOUD SUCCESS
• Consider the total return on investment – Asset uHlizaHon – Hardware costs – Power efficiency – Enabling redundancy – Supply chain management – Personnel resources
EVALUATING CLOUD SUCCESS
• Assess which projects are a fit for cloud compuHng
• Pilot a single applicaHon, assign a great developer to it, cut him/her loose
• Evaluate your results
• Get started now! – aws.amazon.com
CALL TO ACTION