Browse knowledge base
InitialFirst draft, may evolve based on feedback

Capacity Planning

Size the application from measured workload and recovery needs

Size the application from measured workload and recovery needs

Measure peak concurrency, record counts, request mix, batch sizes, query latency, memory, CPU, connections, and storage growth. Include failure and recovery scenarios, background work, and dependent service limits.

Load-test a representative environment and compare results against your service objectives. Adjust one constraint at a time and repeat the workload. QQQ 4.0 does not provide universal throughput numbers, scaling thresholds, or an automatic managed scaling service.

Was this helpful?

Help us improve this article. Feedback is used to prioritize content updates.

Powered by qqq