Serverless Rails: Understanding the pros and cons
summary
Argues that “serverless” is a misleading label and drills into concrete trade-offs: CPU and disk are not part of the function abstraction, so Lambdas/Cloud Functions can reassign CPU and provide no durable local disk; functions bill per execution while containers (Cloud Run, Fargate, Kubernetes) preserve more server-like guarantees. Gives integration notes: Google Cloud Functions uses Rack and Lambdi helps with AWS Lambda. Recommends fast lazy initialization, avoid background threads and local file persistence, log to stdout, and prefer Ruby stdlib APIs.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 0:28Overview
- 1:40Questions
- 2:11Serverless
- 7:38Cloud abstractions
- 9:05What is a function
- 15:20Saving money
- 16:09Containers
- 20:12What about Rails
- 22:30Functions and Rails
- 26:06Practical Thoughts
- 28:26Rails
related talks
58mServerless Under the Hood
49mYou Don't Want Serverless
53mServerless is the Abstraction We Deserve
35mServerless Experiences -- On Your Own Servers?
01h 00mServerless - reality or BS - notes from the trenches
01h 00mMicroservices and Serverless
51mConfusion In The Land Of The Serverless
52mConfusion In The Land Of The Serverless