Alleviating Disk Pressure in ROKS

Disk Pressure warnings are typically due to temporary space being held by container images or other data. This is usually caused by the application usage on the cluster.

Steps to help fix this:

1. You will need to restart worker nodes showing disk pressure:
   a. Log in to cluster console
   b. Go to "Compute/Nodes"
   c. Select a node with disk pressure
   d. Click the Terminal Tab
   e. Enter 'chroot /root'
   f. Enter 'shutdown -r now' - this will reboot the worker.
   Note: this can also be done from the Worker Nodes section of the IBM Cloud console. Please make sure that you are in the correct cluster and Cloud account: https://cloud.ibm.com/containers/cluster-management/clusters
2. If the above does assist, please reach out to ITZ Support for additional nodes. Please know that step 1 is required to have been completed first and the request for additional nodes are subject to further reviewal and approval.
3. A further review may be needed to review application usage of local data storage (specifically those Pods that are using temp/local storage instead of volumes). This is application specific and further inquiries regarding this should be directed towards the product's support team.