Enter a container's CPU and memory requests and limits to see them normalised, validated, and classified. Accepts every form Kubernetes does — 250m , 0.5 , 512Mi , 1Gi , 500M . Seedable from the URL, e.g. ?cpuRequest=250m&memLimit=1Gi .
Container resources
Leave a field empty to omit it from the spec.
Presets
Manifest snippet
Goes under a container in spec.containers[].
QoS class
Burstable
Requests are set but below limits, or the spec is partial. Can use spare capacity, and is evicted before Guaranteed pods.
Normalised
| Request | Limit | |
|---|---|---|
| CPU (millicores) | 250m | 1000m |
| CPU (cores) | 0.250 | 1.000 |
| Memory (MiB) | 512.00 | 1,024.00 |
| Memory (GiB) | 0.500 | 1.000 |
Replicas per node
| Node | Capacity | Replicas | Bound by |
|---|---|---|---|
| t3.micro | 2 vCPU / 1 GiB | 1 | memory |
| t3.small | 2 vCPU / 2 GiB | 2 | cpu |
| t3.medium | 2 vCPU / 4 GiB | 2 | cpu |
| t3.large | 2 vCPU / 8 GiB | 2 | cpu |
| m5.large | 2 vCPU / 8 GiB | 2 | cpu |
| m5.xlarge | 4 vCPU / 16 GiB | 4 | cpu |
| m5.2xlarge | 8 vCPU / 32 GiB | 8 | cpu |
| m5.4xlarge | 16 vCPU / 64 GiB | 16 | cpu |
Raw capacity by limit. Real schedulable capacity is lower — the kubelet, the CNI and system reservations take a slice of every node.