Microservices patterns in Serverless
Download as PPTX, PDF0 likes293 views
For a lot of people serverless = lambda! That's a good start, compute is where everything started for the cloud. But serverless goes beyond just using a bunch of Lambdas in the backend. I & my friend Adithya used to get frustrated at answering repetitive questions from other friends challenging serverless. So we thought, instead of talking about low hanging fruits which can be done in serverless, let's challenge ourselves into trying to implement what 'we think' would be hard to do on serverless.
More Related Content
What's hot (18)
Similar to Microservices patterns in Serverless (19)
Recently uploaded (20)
Microservices patterns in Serverless
- 1. Refactoring to Serverless Translating common application architecture patterns to serverless on AWS Adithya Reddy & Prashanth HN
- 2. Adithya Reddy Who Are We? @TheTallpants github.com/tallpants SDE, Branch Insurance Prashanth HN @prashanth github.com/hnprashanth CTO, WheelsBox
- 3. Source: AWS
- 4. External / Centralized Configuration Circuit Breakers & Retries Real time / web-sockets Asynchronous background tasks Just Another Serverless talk? Hell No!
- 6. Whats a Circuit Breaker? Circuit breaker is a design pattern used in modern software development. It is used to detect failures and encapsulates the logic of preventing a failure from constantly recurring, during maintenance, temporary external system failure or unexpected system difficulties
- 7. Circuit Breakers Allow systems to "open" the circuit when constant failures happen repeatedly giving immediate failure responses instead of loading the system. It can later try "half open" and see if the system is functioning fine now and "close" in case things are back to normal, making it self-healing. Allow us to contain failures and stop them from creating a cascading effect of other parts.
- 8. Circuit Breaker Using Lambda Source: https://www.jeremydaly.com/serverless-microservice-patterns-for-aws/
- 9. Circuit Breaker Using Step Functions Source: https://tech.trivago.com/2019/04/09/circuit-breaker-with-aws-step-functions/
- 12. Configuration is immutable for each published version 1. 2.
- 13. Config Servers
- 14. Source: gocd.org
- 15. Refresh config values while your app is still running. Your app can choose when to refetch configuration Per request Time-based refetch Etc.
- 18. Secure hierarchical storage for configuration and secrets Store as plain text / KMS encrypted strings Reference by unique name Version history, access control at key / hierarchy level Configurable on-change triggers via CloudWatch Events CloudWatch + CloudTrail Reference-able asTerraform data source (and CloudFormation but only specific versions of keys) Generous free limits (1000TPS free, $0.05 per 10,000TPS)
- 20. Access Control Standard ABAC /Tag based
- 23. github.com/tallpants/ssm-parameter-store Safer, nicer abstraction over SSM parameter store with built-in caching, idempotent preloading, TypeScript compile time checks, and handy autocompletion
- 26. Limitations of Lambda Resource limitations of Lambda: storage etc Lambdas will timeout after 15 minutes Limitations in tech stack & stack customisation
- 27. Amazon ECS
- 28. AWS Fargate
- 29. Example Architecture Resize image to multiple sizes
- 30. Advantages SQS can be integrated with almost any kind of service allowing wide set of triggers Auto-scale without hassle No infrastructure to manage Per second billing (min 1 minute) for the resources used, vCPU & memory Easy to handle failures and retries
- 33. Source: AWS
- 34. API Gateway Websocket Endpoint URL wss:// connectionId connectionId
- 35. API Gateway Websocket Endpoint URL 4. Body sent to client 3. HTTP POST to /connectionId 2. Get active connectionIds 1. Score update event wss://
- 38. Questions?
- 39. Adithya Reddy Who Are We? @TheTallpants github.com/tallpants SDE, Branch Insurance Prashanth HN @prashanth github.com/hnprashanth CTO, WheelsBox