Managing Namespace Capacity with Resource Quotas

Learners will implement ResourceQuotas to restrict the aggregate consumption of compute and storage resources across an entire Kubernetes namespace.

In this lesson, you will learn how to enforce global resource boundaries on a per-namespace basis to prevent a single team or workload from consuming all available cluster capacity.

Why Resource Quotas?

While LimitRanges (which you’ve already mastered) govern the resources of individual pods, ResourceQuotas act as a "hard ceiling" for the entire namespace. Think of it like a budget for a department: even if an individual employee (a Pod) is thrifty, the department (the Namespace) has a maximum total spend for the year. This prevents a "noisy neighbor" from accidentally—or intentionally—consuming all the CPU or Memory nodes have to offer, ensuring there is always headroom for other namespaces.

Implementing a Quota

When you define a ResourceQuota, Kubernetes begins tracking the aggregate requests and limits of every pod created in that namespace. If a new pod would push the namespace over its specified limit, the Kubernetes API server will reject the creation of that pod entirely.

Here is an example of a quota configuration:

apiVersion: v1
kind: ResourceQuota
metadata:
  name: compute-quota
  namespace: production-team
spec:
  hard:
    requests.cpu: "4"
    requests.memory: 8Gi
    limits.cpu: "10"
    limits.memory: 20Gi
    pods: "10"

In this setup, the production-team namespace is restricted to a total request of 4 CPUs and 8Gi of memory. Additionally, by setting a limit on pods: "10", you prevent the deployment of an excessive number of microservices that could exhaust the cluster's IP address range or overhead.

Managing Storage

You can also control storage consumption to ensure your Persistent Volumes don't balloon beyond expectations. By adding requests.storage to your quota, you limit the total amount of persistent data that can be provisioned within the namespace. If a developer attempts to claim a Persistent Volume that exceeds the remaining quota, the request will fail, alerting them to the namespace's storage constraints.

By combining ResourceQuotas with the scheduling constraints and affinity rules you’ve studied previously, you ensure your cluster remains stable, predictable, and fair for all stakeholders.

Key Takeaway: ResourceQuotas provide an aggregate safety net that prevents a single namespace from over-consuming cluster resources, effectively turning a shared cluster into a reliable multi-tenant environment.