During the upgrade procedure on a Node the Onload Operator evicts all pods that have requested an amd.com/onload resource on the current node. This is done so
that these application pods do not encounter unexpected errors during runtime and so
that the upgrade completes as expected.
- If your pods are created by a controller (for example a Deployment), they get
re-created once the upgrade has completed and
amd.com/onloadresources are available again. - If your Pod was created manually you must re-create it manually.
The Operator assumes that all users of either the sfc or onload kernel modules are in pods
that have an amd.com/onload resource. If any pods are
using the sfc interface but do not have a amd.com/onload
resource registered through the device plugin, you must shut them down before starting
the upgrade.