Argo CD Application health status lag during deployment failures

Argo CD monitors the health of Kubernetes resources by comparing the live state against the desired state. When a deployment fails—such as a pod entering a CrashLoopBackOff or ImagePullBackOff state—the Application controller updates the health status to Degraded.

There is a known operational gap regarding the latency between the actual Kubernetes event and the status reflection in the Argo CD UI. Because the Application controller relies on a polling interval to synchronize state, the reported health may not align in real-time with the resource logs.

Given the default polling behavior in recent versions, what are the specific configuration parameters available to reduce this synchronization lag? Additionally, does adjusting these intervals impact the performance of the Application controller when managing a high volume of resources?