Features

• auto load balancing & auto scaling

• managed platform updates

• application health monitoring

• application versioning

• traffic splitting

Compute Engine vs App Engine

compute E = IAAS

app E = Paas, serverless

Environments

• standard - language specific sandbox

• • v1 - java python php go (old versions)

  • • python & php -> restricted network access + restricted libs

  • v2 - java, python, php, node, ruby, go (newer versions)

• flexible - docker containers

• • uses compute engine VMs

  • access to background processes

  • attach local disks

Component Hierarchy

Application: 1 per project

Service: 1 per micro-service

• (eg. backend & frontend)

• each service has settings

• aka "modules"

Version: code + config

Instance:

• run multiple versions

• split traffic

• rollback

Scaling

A) Automatic - based on load

• recommended: continuous workloads

• triggers

• • CPU threshold

  • Throughput threshold

  • max concurrent requests

• config: max & min instances

B) Basic - scales down to 0

• recommended: ad hoc workloads

• cold start possible

• only app engine env standard (not flexible env aka docker)

• config

• • max instances

  • idle timeout

C) Manual

• configure how many instances

• manually change how many instances