hardScenarioDesignMidSeniorStaff
Your autoscaling is configured exactly as designed and the service still browns out every Monday at nine. Where is the time going?
6 min readUpdated 2026-08-15Target archetype: Big Tech, Enterprise Captive, Product Startup
Answer
Short answer
Reactive scaling has a latency of its own — metric window, evaluation, provisioning, warm-up, health checks — and a surge that arrives faster than that latency is served entirely by the capacity you already had.
Was this answer helpful?
Likely follow-ups
- The app tier scaled out in ninety seconds and the database connection count hit its ceiling immediately. What do you change first?
- How would you choose the scaling signal for a service whose work arrives on a queue rather than as requests?
- How much traffic should a freshly started instance receive, and what stops the balancer giving it a full share at once?
- How do you tell a genuine surge apart from a retry storm your own clients are producing?
Related questions
- How would you autoscale a GPU inference service?hardAlso on autoscaling and capacity-planning6 min
- How do you decide whether to use a managed service or self-host a component, and which cloud costs catch teams out?hardAlso on capacity-planning6 min
- You move an internal service to gRPC behind the same load balancer your REST services use, and one pod ends up serving most of the traffic. What is happening, and what do you change?hardAlso on load-balancing6 min
- Before we build this, how do you estimate what the LLM feature will cost to run?mediumAlso on capacity-planning6 min