Standardize node affinity strategy logs

[damemi]

These helpers are called in the node affinity strategy:

descheduler/pkg/descheduler/strategies/node_affinity.go

Line 47 in d7e9305

if !nodeutil.PodFitsCurrentNode(pod, node) && nodeutil.PodFitsAnyNode(pod, nodes) {

However at their current loglevels this leads to confusing logs, because if a pod doesn't fit on the current node that message is always logged. But if it doesn't fit on any node, that may not be logged unless log level is >=2 (see

descheduler/pkg/descheduler/node/node.go

Line 115 in d7e9305

klog.V(2).Infof("Pod %v can possibly be scheduled on %v", pod.Name, node.Name)

). This leads to the confusing implication that eviction was intended, but not possible for some other reason.

Since these are used in the same check, they should have the same log levels for their relevant messages.

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: damemi

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [damemi]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[ingvagabund]

I am not debating what's the right log level to use here. However, I agree to have the same log level here. Compared to other klog.V(?) lines log level 2 seems the most suitable here as log level 1 produces the key logs, log level 2 gives fills missing bits in log level 1 logs which are not so important. Log level 3 is used for strategy specific log which might be important for debugging.

[ingvagabund]

/lgtm