↑↓ move · ↵ open · esc close
3Back to the workbench
Everything filed under capacity — essays, notes, playbooks and signals.
Systems do not break at scale. They break in a predictable order, and knowing the order is most of the job.
Replicas fix read saturation and nothing else. Here is the exact point where adding another one stops paying for itself.
When read load is the bottleneck, the cheapest correct answer is almost never sharding.