Running a Real-World, Mission-Critical System on Azure
summary
Concrete runbook: a production system runs on Azure Kubernetes Service with Azure SQL/Active Geo‑Replicas, Service Bus in two regions, blob storage with GRS and soft‑delete, and Application Insights for full telemetry. Deployments and infra use ARM templates in Azure DevOps pipelines; canary rollouts, pod security policies and non‑root containers reduce risk. Design choices include the outbox pattern, circuit breaker plus retries, leader election for ordered processing, CQRS/event sourcing for auditability, and an MFA‑exempt admin plus regular fire drills.
AI-written summary · about
chapters (14)
- 0:00Introduction to mission-critical systems
- 4:41Azure platform and architecture
- 7:00Service level objectives and redundancy
- 12:18Kubernetes for resilience
- 13:42Security in DevOps
- 17:21Threat modeling strategies
- 19:35Monitoring with Application Insights
- 22:55Automation and team confidence
- 25:24Code strategies for reliability
- 30:30CQRS and event sourcing
- 33:05Azure security features
- 38:27Availability zones and redundancy
- 40:15Costly lessons from outages
- 44:14Final takeaways and Q&A
related talks
54mBuilding Azure Apps
46mReal World DevOps
37mModern Architecture in the Cloud of 2026
22mAzure Infrastructure - What Not to Do
44mBuilding and scaling cloud-native intelligent applications on Azure
47mBuild scalable cloud-native apps with AKS and Azure Cosmos DB
01h 04mZero Downtime Application Deployments on Azure PaaS
50mA Deep Dive into DevOps Practices with Azure SQL