K8s Topology spread constraint在Pod缩容时未被遵守的解决方法问询

Hey there, I totally get how frustrating this must be—especially for apps that depend on strict high availability, an imbalanced pod spread across availability zones (AZs) after scaling down is a huge red flag.

You’ve nailed the core issue here: this is a known limitation with Kubernetes topology spread constraints. When pods are scaled down, the system doesn’t guarantee that your desired balanced distribution across AZs stays intact, which can leave you with a lopsided setup that puts your app’s resilience at risk.

As noted in the official documentation:

“There’s no guarantee that the constraints remain satisfied when Pods are removed. For example, scaling down a Deployment may result in imbalanced Pods distribution. You can use Descheduler to rebalance the Pods distribution.”

The go-to solution here is using the Descheduler tool. Here’s how it works: Descheduler runs on a periodic schedule in your cluster, scans for pods that are violating your topology spread constraints, and evicts those misaligned pods. Once evicted, the Kubernetes scheduler will re-schedule those pods following your original topology spread rules, effectively rebalancing the distribution across all your target AZs.

Just make sure to configure Descheduler with the right profile to focus specifically on topology spread violations. You can tweak its run frequency to match your workload’s needs—this way, you keep your pod distribution balanced without causing unnecessary disruptions to your app.

备注：内容来源于stack exchange，提问作者user10869670