Platform Limits

This page describes how to think about limits without exposing internal implementation details.

Resource planning limits

Each cluster is created with explicit CPU, memory, and storage allocations.

Guidance:

• Start with realistic requests/limits for initial workloads.
• Keep headroom for rollout spikes and incident handling.
• Use dashboards and usage trends before resizing.

Lifecycle limits

• Cluster upgrades are controlled operations and should be planned.
• Prefer one minor Kubernetes version step at a time.
• Delete actions are destructive and should follow a confirmation checklist.

Feature availability

• Console and CLI kubeconfig workflows are available now.
• Terraform provider and public API are planned (coming soon).
• Dedicated node placement is planned; shared-node workflow is current default.

Operational limits to account for

• DNS propagation latency for domain changes.
• Certificate issuance and renewal timing.
• Rollout and reconciliation timing for GitOps changes.