Kubernetes Capacity Calculator

Pod Workload

CPU Request per Pod (millicores)

m

Memory Request per Pod (MiB)

MiB

Pod Replicas

Node Specifications

CPU Cores per Node

cores

Memory per Node (MB)

MB

System Overhead (per node)

System Reserved CPU (millicores)

m

System reserved includes OS daemons, Kubernetes components (kubelet, container runtime), and eviction thresholds. Typical values: 500-1000m CPU, 500-1000 MiB memory.

System Reserved Memory (MiB)

MiB

System reserved includes OS daemons, Kubernetes components (kubelet, container runtime), and eviction thresholds. Typical values: 500-1000m CPU, 500-1000 MiB memory.

DaemonSet CPU (millicores)

m

DaemonSets run on every node (monitoring agents, log collectors, network plugins). Typical values: 200-500m CPU, 256-512 MiB memory per node.

DaemonSet Memory (MiB)

MiB

DaemonSets run on every node (monitoring agents, log collectors, network plugins). Typical values: 200-500m CPU, 256-512 MiB memory per node.

Target Utilization

Target CPU Utilization (%)

%

Target utilization leaves headroom for autoscaling and traffic spikes. Industry standard: 70% CPU, 80% memory. 100% leaves no buffer.

Target Memory Utilization (%)

%

Target utilization leaves headroom for autoscaling and traffic spikes. Industry standard: 70% CPU, 80% memory. 100% leaves no buffer.