SRE or Devops - I am an Architect - Why Do I Care?
summary
Claim: availability matters as much as features and SRE is the engineering function that enforces it. Defines SLA/SLO/SLI — example SLA 99.99% equals 4.4 minutes allowed downtime and SLO should sit above the SLA. Recommends concrete tooling: synthetic checks and /health check API for blackbox monitoring, end-to-end scripts for APNS/FCM, APM agents, sysdig/ELK/Grafana/Kibana/LogDNA, PagerDuty for callouts, plus playbooks, circuit breakers, bulkheads, feature toggles, multi-AZ and rate limiting (HTTP 429). Monitoring adds small overhead but is essential.
AI-written summary · about
chapters (13)
- 0:00Introduction and setup
- 1:18Why architects should care about SRE
- 3:04Availability as a business requirement
- 5:53SLA, SLO and SLI explained
- 9:09Measuring service level indicators
- 11:18Blackbox monitoring techniques
- 15:36Infrastructure and APM metrics
- 18:51Automation in site reliability
- 22:00Disaster recovery and business continuity
- 24:10Team structures for SRE
- 26:40Patterns for resilient cloud apps
- 30:01Q&A on cloud costs and monitoring tools
- 35:20Q&A on SRE and stability trade-offs
