Scaling on AWS for the First 10 Million Users at Websummit Dublin
3 likes1,239 views
In this talk from the Dublin Websummit 2014 AWS Technical Evangelist Ian Massingham discusses the techniques that AWS customers can use to create highly scalable infrastructure to support the operation of large scale applications on the AWS cloud. Includes a walk-through of how you can evolve your architecture as your application becomes more popular and you need to scale up your infrastructure to support increased demand.
More Related Content
More from Ian Massingham (20)
Recently uploaded (20)
Scaling on AWS for the First 10 Million Users at Websummit Dublin
- 1. WEB SUMMIT DUBLIN 2014 #AWS #WEBSUMMIT Scaling on AWS for the First 10 Million Users Ian Massingham — Technical Evangelist [email protected] @IanMmmm STARTING AT 11:00
- 2. An Introduction to Scaling on AWS
- 3. An awful lot of things to read Not where we want to start EC2 Auto Scaling is one tool, not the single thing that ‘fixes’ everything or enables you to scale
- 4. SO WHAT DO WE NEED FIRST?
- 5. SOME BASICS
- 6. 2004 2014 $7B+ retail business 8,000+ employees A whole lot of servers Every day, AWS adds enough server capacity to power that entire $7B enterprise
- 7. Foundation Services Infrastructure Collaboration and Sharing Storage (Object, Block and Archive) Security & Networking Regions Availability Zones Access Control Platform Services Databases Relational NoSQL Caching Analytics Hadoop Real-time Data warehouse App Services Queuing Orchestration App streaming Transcoding Email Search Deployment & Management Containers Dev/ops Tools Resource Templates Mobile Services Identity Sync Mobile Analytics Notifications Enterprise Applications Virtual Desktops Data Workflows Usage Tracking Monitoring and Logs Compute (VMs, Auto-scaling and Load Balancing) CDN and Points of Presence
- 8. Startups on AWS http://aws.amazon.com/solutions/case-studies/
- 9. DAY ONE, USER ONE. PROBABLY YOU! • A single EC2 instance • With a full stack on this host • Web app • Database • Management • A single Elastic IP address • Amazon Route 53 for DNS Elastic IP address EC2 instance Amazon Route 53 User
- 10. WE’RE GOING TO NEED A BIGGER BOX! • Simplest approach to scaling • You can now leverage PIOPs • High I/O instances • High memory instances • High CPU instances • High storage instances • Easy to change instance sizes • Will hit an endpoint eventually
- 11. WE’RE GOING TO NEED A BIGGER BOX! • We could potentially get to a few hundred to a few thousand depending on application complexity and traffic • No failover • No redundancy • Too many eggs in one basket Elastic IP address EC2 instance Amazon Route 53 User
- 12. DAY TWO, USER >1 Web instance Database instance Elastic IP address Amazon Route 53 User • Separate out the functions on our single host onto more than one: • Web • Database • Make use of a database service
- 13. Microsoft SQL, Oracle, MySQL or PostgresSQL managed service Flexible licensing: BYOL or Included Self-Managed Fully-Managed Massively parallel, petabyte scale, data warehouse service Fast, powerful & easy to scale Managed NoSQL database service using SSD storage Seamless scalability & zero administration Database server on Amazon EC2 Amazon RDS Amazon DynamoDB Amazon Redshift Your chosen database running on Amazon EC2 Database Options
- 14. Building Blocks for Scaling on AWS
- 15. HOW DO I DETERMINE WHICH DB TECHNOLOGY I NEED?
- 16. Why start with SQL? • Established and well-worn technology • Lots of existing code, communities, books & tools • You aren’t going to break SQL DBs in your first 10 million users. No really, you won’t* • Clear patterns to scalability * Unless you are manipulating data at MASSIVE scale; even then, SQL will have a place in your stack
- 17. Why might you need to use a NoSQL DB? • Super low latency applications • Metadata driven datasets • Highly non-relational data • Need schema-less data constructs* • Massive amounts of data (again, in the TB range) • Rapid ingest of data ( thousands of records/sec ) *Need != “it is easier to do dev without schemas”
- 18. IF YOU DO NEED NOSQL, CONSIDER AMAZON DYNAMODB • Managed, provisioned throughput NoSQL database • Fast, predictable performance • Fully distributed, fault tolerant architecture
- 19. Elastic IP address Web instance Amazon Route 53 RDS DB instance User USER >100 • First, let’s separate out our single host into more than one: • Web • Database • Use Amazon RDS to make your life easier
- 20. USER >1000 Web instance Amazon RDS DB instance Web instance Amazon RDS DB instance Active (Multi-AZ) Availability Zone Availability Zone Standby (Multi-AZ) Elastic Load Balancing Amazon Route 53 User • Next, let’s address the lack of failover and redundancy issues: • Elastic Load Balancing • Another web instance • In another Availability Zone • Enable Amazon RDS Multi-AZ
- 21. USER 10,000 to 100,000s RDS DB Instance Active (Multi-AZ) RDS DB Instance Standby (Multi-AZ) Elastic Load Balancing RDS DB Instance Read Replica Availability Zone Availability Zone RDS DB Instance Read Replica RDS DB Instance Read Replica RDS DB Instance Read Replica Web instance Web instance Web instance Web instance Web instance Web instance Web instance Web instance Amazon Route 53 User
- 22. Shift Some Load Around Elastic Load Balancing Web instance RDS DB Instance Active (Multi-AZ) Availability Zone Amazon CloudFront Amazon S3 Amazon Route 53 User ElastiCache Amazon Dynamo DB • Think components and services: • Move static content from the web instance to Amazon S3 and Amazon CloudFront • Move session/state and DB caching to Amazon ElastiCache or Amazon Dynamo DB
- 23. NOW THAT OUR WEB TIER IS MUCH MORE LIGHTWEIGHT: AUTOSCALING!
- 24. AMAZON EC2 AUTO SCALING Automatic resizing of compute clusters based on demand metrics Trigger auto-scaling policy Amazon CloudWatch Feature Details Control Define minimum and maximum instance pool sizes and when scaling and cool down occurs Integration Use metrics gathered by CloudWatch to drive scaling (including custom metrics) Instance Types Use Auto Scaling with On-Demand and Spot Instances
- 25. NOVEMBER TRAFFIC TO AMAZON.COM AUTO SCALING LETS YOU DO THIS 76% of traditionally provisioned capacity Traditional provisioning level 24% of traditionally provisioned capacity
- 26. Web instance Availability Zone Amazon Route 53 User Amazon S3 Amazon CloudFront Web instance Availability Zone Elastic Load Balancing Amazon RDS DB Instance DynamoDB Read Replica Web instance Web instance ElastiCache RDS DB Instance Read Replica Web instance Web instance RDS DB Instance ElastiCache Standby (Multi-AZ) RDS DB Instance Active (Multi-AZ) USER 500,000+
- 27. GOING BEYOND 500,000 USERS
- 28. GOING BEYOND 500,000 USERS • Monitoring, metrics and logging • Listen and respond to customer feedback as well as your metrics • Optimise each service/component • Implement a Service Orientated Architecture with loose coupling • Load balance between tiers
- 29. RDS DB Instance Active (Multi-AZ) Availability Zone Elastic Load Balancing RDS DB Instance Read Replica RDS DB Instance Read Replica Web instance Web instance Web instance Web instance Amazon Route 53 User Amazon S3 Amazon CloudFront Amazon SQS Amazon DynamoDB ElastiCache Worker instance Worker instance Amazon CloudWatch Internal app instance Internal app instance Amazon SES USER 1 MILLION+
- 30. WANT TO LEARN MORE? COME TO OUR ARCHITECTURE ASK US ANYTHING TODAY AT 16:00
- 31. WEB SUMMIT DUBLIN 2014 #AWS #WEBSUMMIT DON’T FORGET YOUR CREDITS Ian Massingham — Technical Evangelist @IanMmmm AWS Global News and Service Launches @AWScloud AWS UK & Ireland News, Events & Webinars @AWS_UKI