OpenShift - Resolve "DiskPressure"

by
Jeremy Canfield |
Updated: November 04 2022
| OpenShift articles
Let's say you get a DiskPressure alert. The status of the node with the DiskPressure alert should have something like this.
~]# oc get node my-node --output yaml
status:
conditions:
- lastHeartbeatTime: "2022-11-04T03:26:09Z"
lastTransitionTime: "2022-11-03T15:36:51Z"
message: kubelet has no disk pressure
reason: KubeletHasNoDiskPressure
status: "True"
type: DiskPressure
You can also get the capacity and allocatable memory, storage and CPU resources of the node.
~]# oc get node worker-x86lt --output yaml
status:
allocatable:
attachable-volumes-aws-ebs: "25"
cpu: 7910m
ephemeral-storage: "289416708846"
hugepages-1Gi: "0"
hugepages-2Mi: "0"
memory: 65536m
pods: "250"
capacity:
attachable-volumes-aws-ebs: "25"
cpu: "6"
ephemeral-storage: 314037228Ki
hugepages-1Gi: "0"
hugepages-2Mi: "0"
memory: 324
pods: "250"
The machine should also have similar memory and CPU values.
~]$ oc get machine worker-x86lt --namespace openshift-machine-api --output yaml
spec:
providerSpec:
value:
memoryMiB: 65536
numCPUs: 6
If you increase the amount of CPU or memory being used by the machineset, you will need to delete each worker node so that a new worker node gets created with the update CPU/memory values.
Did you find this article helpful?
If so, consider buying me a coffee over at