RunAs Radio Azure Resiliency with Chris Ayers
4 snips
Nov 12, 2025 In this engaging discussion, Chris Ayers, a senior software engineer at Microsoft’s Azure Reliability team, dives into the complexities of Azure resiliency. He highlights the Well-Architected Framework and its role in aligning reliability with business needs. Chris outlines types of outages, the importance of availability zones, and the necessity of balancing costs with service reliability. With practical advice on monitoring, operational excellence, and data protection, he emphasizes the value of automated tools while stressing the importance of human oversight in decision-making.
AI Snips
Chapters
Transcript
Episode notes
Architecture Defines Who Handles Failures
- Moving from single VMs to platform services changes responsibility and failure modes.
- Architectural choices determine which failures you must handle in software versus the cloud.
Design For Graceful Degradation
- Design graceful degradation paths so optional features can be turned off under load.
- Keep core functions online while deferring nonessential services during incidents.
Pre-Scale For Predictable Load
- Pre-scale for known traffic patterns instead of relying solely on autoscale.
- Use scheduled scaling for predictable spikes and reserve autoscale for unexpected surges.