Field notes on incidents, recovery, cost, and the operating practices that keep production work safer.
We moved workloads to a new Alibaba Cloud ACK node pool and removed the old ones. Kubernetes looked fine. That was not enough to close the prepaid ECS commitments behind the old pools.
Human error describes what a person did before an incident. A useful investigation explains why the action made sense at the time and which controls allowed it to cause harm.
Reduce MTTR by fixing ownership, incident context, rollback, runbooks and recovery paths instead of adding another noisy alert after each outage.
Cloud cost work belongs in the operating model. Cut waste with ownership and clear trade-offs, not by quietly removing the capacity and observability that keep incidents small.
AIOps can correlate signals and reduce noise. It does not replace ownership, runbooks, or judgment about customer impact. Set expectations before you buy a platform story.