You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm consistently seeing warnings about the amount of available memory on the controllers (masters), for example:
System memory usage of 1.266G on Node (N) ocp4-master1.aio.example.com exceeds 95% of the reservation.
The warning goes away when I increase the VM memory to 20G, so it seems like the reserved memory amount is auto-calculated from the node's total memory (but I can find no documentation of the calculation anywhere). Should the memory of the masters be bumped up to 20G?
The text was updated successfully, but these errors were encountered:
Upon further research, it looks like OpenShift is setting a default value of 1GB in /etc/node-sizing.env, which is included as an environment file by the kubelet.service unit file. So the warning probably went away simply because I rebooted the node.
I'm consistently seeing warnings about the amount of available memory on the controllers (masters), for example:
The warning goes away when I increase the VM memory to 20G, so it seems like the reserved memory amount is auto-calculated from the node's total memory (but I can find no documentation of the calculation anywhere). Should the memory of the masters be bumped up to 20G?
The text was updated successfully, but these errors were encountered: