Your team deploys a payment API to a production stage in Azure Pipelines using a deployment job. After each deployment, transient error rates spike for about 5 minutes while caches warm up before stabilizing. You configure a post-deployment release gate that queries an Azure Monitor alert rule to verify the service is healthy before the pipeline reports success. During the first few minutes the alert is active due to the expected warm-up spike, causing the gate to evaluate as failed prematurely. You must ensure the gate reflects the steady-state health without disabling the health check. What should you configure on the gate?