Web Apps, through the Lens of SRE (Site Reliability Engineering)
summary
Porter runs web SRE for a platform with ~500,000 driver partners across 19 Indian cities and two abroad, building web apps, webviews and Lambda invoices and standing up SRE in six months. Workflow is collect-monitor-fix with Sentry/Datadog for client errors, OpenSearch and Google Analytics for logs, and CloudWatch+SNS+OpsGenie for alerts. An example SLI uses (fatal+nonfatal)/total sessions = 4.83; mitigation includes hotfix pipelines and 1% canary blue‑green rollbacks, plus rate limiting, caching and postmortems.
AI-written summary · about
chapters (8)
- 0:00Introduction to SRE
- 2:58SRE principles and practices
- 5:27Porter's web systems overview
- 6:31Error and log collection
- 8:01Monitoring and alerting setup
- 10:25Mitigation and prevention strategies
- 12:55SRE mindset and postmortems
- 13:57Closing and key takeaways
related talks
40mHow Google SRE and developers work together
27mHow Google SRE and developers work together
30mThe Path to SRE
21mSRE: An incomplete guide to cultural Narnia
38mSRE for Mobile Applications
21mSRE, what is it all about?!
26mTechnology and Cultural Transformation using Site Reliability Engineering
47mSRE, Myths vs Realities