A Cloud Run service queries Cloud SQL and daily morning traffic spikes cause Cloud SQL API quota errors; project hit API quota. Which configuration change mitigates this?

Setting a higher minimum number of Cloud Run instances ensures that a pre-warmed pool of instances is always available to handle incoming requests. This prevents cold starts and the sudden surge of new instance creation requests during traffic spikes, which can overwhelm the Cloud SQL API and lead to quota errors. By having instances ready, the service can absorb the initial load more smoothly, reducing the API call rate to Cloud SQL during the critical warm-up phase. Using traffic splitting is for A/B testing or gradual rollouts, not for mitigating quota errors. Modifying the maximum number of Cloud Run instances controls the upper limit of scaling but doesn't address the initial burst of requests causing quota issues. Setting a minimum concurrent requests (container concurrency) environment variable affects how many requests a single instance can handle, not the number of instances available to handle traffic spikes.