Easy, scalable backends for Android developers
summary
Explains concrete tradeoffs: use Cloud Functions (Firebase) for the easiest path with callable functions (automatic auth, JSON serialization, Node/TypeScript), or choose Cloud Run with Docker to run Kotlin/Java servers (Gradle, Ktor, build a fat JAR, gcloud/cloud build). Notes limits: 10 MB payloads, no persistent sockets, temp is in-memory, CPU/memory/time caps; scales to zero and (practically) hundreds or more instances, billing per use. Provides working samples and an Android client repo.
AI-written summary · about
chapters (24)
- 0:00Start
- 5:36Google App Engine
- 7:05Cloud Functions runtimes
- 13:16HTTPS simple function sample
- 14:58"Callable" functions - a convenience wrapper for HTTP
- 16:15Callable function example
- 18:08Cloud Functions is event-driven
- 18:33Writing a background function
- 19:25(Some) Cloud Functions event providers
- 19:58Firestore trigger example
- 22:27What you have to know to use Cloud Functions for Firebase
- 23:24Cloud Functions for Firebase Resources
- 24:09Kotlin on Cloud Functions
- 24:45Partial build.gradle for a Cloud Function running Kotlin
- 25:31Hello world HTTP endpoint
- 25:58Deployment
- 30:03Partial build.gradle for a Kotlin/ktor backend
- 31:56Sample Dockerfile for a JVM backend
- 35:54Build and deploy a docker container to Cloud Run
- 37:25Cloud Run: What exactly do you want to run?
- 38:13What you have to know to use Cloud Run
- 41:16CF vs. CR - Capabilities and limits
- 42:28Serverless products scale up and down with load
- 43:44What serverless backends are NOT good for
related talks
40mEasy, Scalable Backends for Android Developers
43mEasy, scalable backends for Android developers
58mEasy, scalable serverless backend for mobile/web app
18mGo full-stack with Kotlin or Dart on Google Cloud
50mGo Serverless with Google Cloud
35mGoogle Cloud Functions and Firebase
19mFull Stack in a Weekend
28mBackend Engineering for Android Developers