Paas

What I do

• Deploy applications to PaaS platforms
• Configure managed services and databases
• Implement auto-scaling and high availability
• Set up CI/CD for PaaS deployments
• Manage application performance
• Optimize PaaS costs

When to use me

• When focusing on code, not infrastructure
• When building web applications quickly
• When needing managed databases
• When implementing auto-scaling
• When reducing operational overhead
• When building cloud-native applications

Key Concepts

Heroku Deployment

# app.json - Heroku Review Apps
{
  "name": "myapp",
  "buildpacks": [
    {
      "url": "heroku/python"
    }
  ],
  "environments": {
    "review": {
      "addons": ["heroku-postgresql:hobby-dev"],
      "environment": {
        "LOG_LEVEL": "DEBUG"
      },
      "formation": {
        "web": {
          "quantity": 1,
          "size": "hobby"
        }
      }
    },
    "production": {
      "addons": ["heroku-postgresql:standard-0"],
      "formation": {
        "web": {
          "quantity": 2,
          "size": "performance-m"
        }
      }
    }
  }
}

AWS Elastic Beanstalk

# .ebextensions/python.config
option_settings:
  aws:elasticbeanstalk:container:python:
    WSGIPath: application:application
  aws:elasticbeanstalk:environment:proxy:staticfiles:
    /static: static
  aws:autoscaling:asg:
    MinSize: 2
    MaxSize: 10
  aws:autoscaling:trigger:
    MeasureName: CPUUtilization
    UpperThreshold: 80
    LowerThreshold: 20

Azure App Service

# azure-pipelines.yml
trigger: main

pool:
  vmImage: 'ubuntu-latest'

variables:
  azureSubscription: 'Azure-Service-Connection'
  appName: 'myapp'
  runtime: 'python'
  version: '3.11'

stages:
  - stage: Build
    jobs:
      - job: BuildJob
        steps:
          - task: UsePythonVersion@0
          - script: |
              pip install -r requirements.txt
              pip install pytest
              pytest tests/
          - task: ArchiveFiles@2
            inputs:
              rootFolder: '$(System.DefaultWorkingDirectory)'
              archiveFile: '$(Build.ArtifactStagingDirectory)/$(Build.BuildId).zip'
          - publish: $(Build.ArtifactStagingDirectory)/$(Build.BuildId).zip
            artifact: drop

  - stage: Deploy
    jobs:
      - deployment: DeployJob
        environment: 'production'
        strategy:
          runOnce:
            deploy:
              steps:
                - task: AzureWebApp@1
                  inputs:
                    azureSubscription: $(azureSubscription)
                    appType: 'webApp'
                    appName: $(appName)
                    package: '$(Pipeline.Workspace)/drop/**/*.zip'

GCP App Engine

# app.yaml
runtime: python311
env: standard

instance_class: F2
automatic_scaling:
  min_instances: 1
  max_instances: 10
  target_cpu_utilization: 0.65
  min_latency: '500ms'

inbound_services:
  - warmup

liveness_check:
  path: /_ah/health
  check_interval_seconds: 30
  timeout_seconds: 4
  failure_threshold: 2
  success_threshold: 2

readiness_check:
  path: /_ah/ready
  check_interval_seconds: 5
  timeout_seconds: 4
  failure_threshold: 2
  success_threshold: 1

env_variables:
  LOG_LEVEL: 'INFO'
  DATABASE_URL: 'postgres://...'

beta_settings:
  cloud_sql_instances: 'project:region:instance'

Cloud Run Deployment

# cloudrun.yaml
apiVersion: serving.knative.dev/v1
kind: Service
metadata:
  name: myapp
  annotations:
    run.googleapis.com/launch-stage: BETA
spec:
  template:
    metadata:
      annotations:
        autoscaling.knative.dev/minScale: "1"
        autoscaling.knative.dev/maxScale: "10"
    spec:
      containers:
        - image: gcr.io/project/myapp:latest
          ports:
            - containerPort: 8080
          resources:
            limits:
              cpu: "1000m"
              memory: "512Mi"
          env:
            - name: PORT
              value: "8080"
          startupProbe:
            httpGet:
              path: /_ah/health
            initialDelaySeconds: 0
            periodSeconds: 10
            timeoutSeconds: 1
            failureThreshold: 3

PaaS Comparison

Key Benefits

• Managed Runtime: Don't manage OS or middleware
• Auto-scaling: Handle traffic spikes automatically
• Managed Services: Databases, caches, queues
• CI/CD: Built-in deployment pipelines
• Monitoring: Integrated logging and metrics
• Security: Patched and secured platforms