2. Today’s Section
• Contrast the AWS Cloud to traditional on-premises model.
• Serverless: The New Cloud Trend
• Q&A
3. 14 YEARS
70%
66%
New Trends On The S&P 500
The average tenure of companies is forecast to
shrink to 14 years by 2026 from 20 years by 1990.
IT Budgets : 70% for Run Business, 20% for Grow
Business, only 10% for Transform Business.
66% planning to increase IT security budgets as a
direct response to this risk.
4. What The Meaning of Number?
The traditional model of enterprises lacks agility
and cannot keep pace with new companies.
Lack of the resources to meet new business
needs to ensure competitive advantage.
Safety, compliance, and lack of usability have
reduced the competitiveness and put into risk.
9. The Facing Problems
• Hiring a group of people or buying equipment
unrelated to core business development.
• It is difficult to estimate the IT budget.
• Lack of flexibility in adjusting resources.
• Build own wheels that belong to the company.
The company loses competitiveness
10. How Companies Succeed
Innovation or adopting new technologies to
"incremental" improvement over time.
Minimum viable product and faster of delivery
speed to reached the target market.
Scalability of IT infrastructure for expansion and
reduction for response to the business cost.
11. Cloud Computing
A simple way to access servers, storage, databases and a
broad set of application services over the Internet.
20. The Traditional VS Cloud Computing
Traditional
• Massive Up-front cost, Big “Steps”
• IT Staff focus on the data center
• In-house knowledge limits
• Recreate the wheel
• It’s yours
• “Monster server” capability
Cloud
• Pay as you go, Pricing Models
• Elastic Computing, Grow as needed
• Economy of scale
• Immediate security accreditation
• Globalize Infrastructure
• Collaborative Innovation
• Horizontal Scaling
21. Six Advantages of AWS
VARIABLE
EXPENSE SCALABILITY
SPEED
&
AGILITY
STOP WASTE
ON DATA CENTER
STOP
GUESSING
GO GLOBAL
IN MINUTES
23. Transferring of Enterprise IT
Your Datacenter Public CloudHybrid Cloud
1. Development,
Testing and New
workloads
2. Analysis existing workloads
and Migrating Applications
3. Data Center
Migration
4. All-in
Build & Maintain Create More Value
24. • Stop guessing capacity needs
• Test system at production scale
• Automate to ease architectural experimentation
• Allow for evolutionary architectures
• Drive your architecture using data
• Improve through Game Days
The Cloud: What’s Different?
27. Recap : How Companies Succeed
Innovation or adopting new technologies to
"incremental" improvement over time.
Minimum viable product and faster of delivery
speed to reached the target market.
Scalability of IT infrastructure for expansion and
reduction for response to the business cost.
31. Elastic
React to Load
Responsive
Experiments
Resilient
React to Failures
Message Driven
Module / Components Interaction
GOAL
PRINCIPLES
METHOD
Reactive
Continuous Delivery with
workable product to delivery
Minimum Viable Product to
customer. And it response to
the business with High-
availability and Scalability
automatically.
32. Server to Serverless
Application and
Operation teams design,
acquire, and maintain
servers to meet the
specific needs of
application.
Hardware abstracted
away to allow for
specific CPU, Memory,
Storage, and Network to
meet the needs of
application on a per
server basis.
Hardware and Operating
System is abstracted
away. Can packaged to
run in resource-isolated
processes.
Applications are built
using a combination of
services. Break
application into small
bite sized pieces.
Server to Serverless
On-premise Virtual Machines Containers Serverless
Deploy in months
Live for years
Deploy in hours
Live for weeks
Deploy in minutes
Live for hours
Deploy in seconds
Live for seconds
33. From One Big To Many Small
Monolithic
Application
Microservice
Microservice
Microservice
Function
Function
Function
Function
Function
35. Let’s Talk About Lambda
Serverless compute platform for stateless code execution
in response to events
36. AWS Lambda Architecture
Trigger Resources
Permissions
Amazon Kinesis
Amazon DynamoDBAmazon Dynamo DB Amazon SNS
AWS CodeCommitAmazon API Gateway
Alexa skillAlexa Skill
Amazon S3Amazon S3
AWS IoTAWS IoTAlexa smart home skillAlexa Smart Home
Amazon CloudWatch
Amazon CognitoAmazon Cognito
Code Permissions CloudFormation
37. Programming Language Support
• JavaScript with Node.js
• Java
• Python
• C# with .NET Core
https://docs.aws.amazon.com/lambda/latest/dg/current-supported-versions.html
38. Pay per request
• Buy compute time in 100ms increments.
• Low request charge.
• No hourly, daily or monthly minimums.
• Only pay for running, never pay for idle.
Free tier: 1 million requests, and 400,000 GBs of compute
every month for every customer.
39. Case : Real-time File Processing
https://aws.amazon.com/lambda/ https://aws.amazon.com/solutions/case-studies/the-seattle-times/
40. Case : Real-time Stream Processing
https://aws.amazon.com/lambda/ https://aws.amazon.com/solutions/case-studies/localytics/
41. Case : IoT Backends
https://aws.amazon.com/lambda/
42. Case : Web Applications
https://aws.amazon.com/lambda/
44. Background
As a Customer
I want a New Service that can auto-generate daily summary report
according to the display from current back-end system(web), and
send this report to related people
So that we can do the Marketing Analysis more accurately.
45. Todo List
• Make a Crawler to gathering information from system display.
• Design a mechanism to convert the information and generate
daily report.
• Integrate File System for report stored.
• Integrate Mail Service to send the report.
• Design the system high-availability and scalability.
46. Traditional Architecture
Buying Server and setup environment for 2 weeks.
Programing development
- 1 weeks for service setup(file server, email, web
server)
- 1 weeks for core functions
- 2 weeks for high availability and scalability.
Difficult to estimate Cost and Capacity. And Lost the
Focus on development.
49. Results
• Delivery from 6 weeks reduce to 1 week.
• Save 90% cost in the front of the project and has
predictable billing report for evaluating.
• Enable fully Auto-scaling and High-availability.
• Report file been Protecting from AWS S3.
• 1 command to deploy and delivery in seconds.
50. Benefits of Going Serverless
6.7Cost-effective6.7
Time to
Market
6.7
Focus on
Logic
51. When to Use Serverless Architecture
Cron Job Service Lightweight API
Data Statistics Analysis
Chatbot & WebHookInternet Crawler
Trigger