Infrastructure // Scaling

Kubernetes Capacity and Scaling

Nova manages cluster capacity and per-site resource controls on Kubernetes. WordPress sites use one tenant pod by default; per-site multi-pod scaling is an operator-configured exception, not an automatic feature on every plan.

Cluster capacity

The GKE node pool can add or remove nodes within configured bounds so the platform has room to schedule workloads.

  • + Managed node-pool capacity
  • + Kubernetes scheduling
  • + Health checks

Per-site controls

Each site receives explicit CPU and memory requests and limits. These controls isolate workloads and make resource use predictable.

  • + Namespace isolation
  • + CPU and memory limits
  • + Plan-specific capacity

Default topology

Every production plan defaults to one WordPress replica. Nova does not claim that tenant pods automatically multiply during a traffic spike.

Validated exceptions

Nova can configure multiple replicas only after reviewing storage, cache, session, and database behavior for the workload. That configuration is operated case by case.

FAQ

Does my WordPress pod automatically scale with traffic?

No. Each plan defaults to one WordPress replica. Nova can configure multiple replicas only after validating the site’s storage, cache, session, and database behavior.

What does scale automatically?

The Kubernetes cluster node pool can adjust capacity within configured bounds. That is cluster capacity scaling, not automatic multiplication of each tenant WordPress pod.

How does Nova prepare for heavier workloads?

Nova uses per-site resource limits, cluster capacity controls, monitoring, and operator-reviewed scaling configuration where a workload supports it.