From System TAYE7 to High Availability: What, Why, How?
summary
Defines concrete failure modes (power, cooling, disk, network, misconfiguration) and maps fixes: multi-region active‑active with a global load balancer, Kubernetes/OpenShift for container-level redundancy, and F5 Big‑IP or cloud global LB for traffic steering. Cites the Microsoft outage on July 19 as a real failure case, recommends planned chaos (Chaos Monkey) and controlled failovers, describes server clustering with automatic VM migration and rolling upgrades, and shows that these measures cut unplanned downtime from hours to minutes in practice.
AI-written summary · about
chapters (6)
- 0:00Introduction and welcome
- 3:18What is system downtime
- 8:02Causes of system downtime
- 11:21Ensuring high availability
- 13:33Multi-cloud and hybrid cloud strategies
- 15:33Conclusion and thanks
related talks
22mFailing Over without Falling Over
16mEngineering Failure-Resilient Systems
45mDeveloping a Chaos Architecture Mindset
28mBuilding Resilient DevOps Pipelines: Strategies for Handling Unexpected Failures &...
48mFault Tolerance on the Cheap Making Systems That Probably Won't Fall Over
51mDesigning for Failure: Build Resilient Services
36mDesigning for failure - architecting fault-tolerant system
44mChaos Engineering for SQL Server