Panel
summary
Virtual event workload spikes to 60 talks/year for one role and teams double headcount, causing multiple deploys per day; fixes are concrete: add telemetry and logging, use feature flags/modules to turn off Slack notifications and critical paths, prefer wide, shallow health checks to deep ones that can queue and crash services, instrument dependencies so outages are attributable, design graceful degradation and roll-forward strategies, require new hires to edit onboarding docs, and schedule recovery (PTO, enforced boundaries, offline hours) to prevent human-system collapse.
AI-written summary · about
chapters (6)
- 0:00Start
- 0:21Kat Cosgrove
- 0:33Katie Farmer
- 31:30Wide Shallow Testing
- 34:41Graceful Degradation
- 46:20Final Thoughts
