If the cluster has came into a state in which live migrations that have been triggered by kubevirt-workload-update are constantly failing, it produce a situation in which a target virt-launcher pod is created every 5 minutes for every VM in the cluster.
This results in thousands of virt-launcher pods with "Error" state lying around the cluster, overloading etcd and might cause the cluster to become less responsive in short time.

all versions

if VMIMs are failing, 100%

1. reproduce bug https://issues.redhat.com/browse/RHEL-131697 (for example)
2. observe that a new virt-launcher target pod is created every 5 minutes and then got failed, over and over again for every VM in the cluster.
3.

thousands of Errored virt-launcher pods reside on the cluster in 1 day

there should be an exponential backoff mechanism in such case.

If there is a permanent issue, retries shouldn't be executed in constant intervals of 5 minutes.
Instead the next retry should be twice as long as the previous one.
Until an upper limit of, for example, 4 hours.