Operating Rails: What About After You Deploy?
summary
Framework for deploying Rails centers on five operational risks: errors, data, speed, security, lead time. Recommends concrete tools and practices: use Honeybadger or Sentry for exceptions and uptime, Honeycomb or Prometheus/DataDog for traces and metrics, Fastly for CDN, Strong Migrations for safe schema changes, Dependabot/Renovate for deps, GitHub Actions + artifacts for deploys and backups. Emphasizes testing backups, profiling percentiles not averages, and shrinking lead time so fixes deploy in minutes.
AI-written summary · about
chapters (8)
- 0:00Operating Rails apps
- 3:20Deployment tutorials' limits
- 7:31Staging environments for confidence
- 12:47Monitoring errors and uptime
- 18:53Rails app speed optimization
- 22:47Securing user data and servers
- 25:55Reducing deployment lead time
- 29:01Final advice and Q&A
related talks
32mOperating Rails: what about after you deploy?
34mGuide to Continuous Deployment with Rails
36mContinuous Deployment with Rails
34mIndustrial Rails
27mScaling Ruby Applications: Challenges, Solutions and Best Practices
39mZero downtime deploys for Rails apps
35mTen Years of Rails Deployment
35mCI/CD and devops with Ruby and Rails