Log deletions at INFO level instead of DEBUG

[vdemeester]

Problem

When the pruner deletes resources (PipelineRuns/TaskRuns), it logs at DEBUG level (Debugw):

• pkg/config/history_limiter.go:389 — logger.Debugw("deleting resource", ...)
• pkg/config/history_limiter.go:87 — logger.Debugw("resource is in deletion state", ...)

With the default INFO log level, operators cannot tell whether the pruner is actually deleting anything without manually counting resources. The only logs visible are Reconcile succeeded from the knative controller framework — which appear whether resources were deleted or not.

Impact

On the Tekton dogfooding cluster, we had 220 PVCs accumulating (215 orphaned) and could not tell from pruner logs whether it was working. We had to manually count PipelineRuns per namespace to verify.

Proposed fix

Change deletion-related log lines from Debugw to Infow. Resource deletions are important operational events and should be visible at the default log level.

At minimum, line 389 in history_limiter.go should be Infow:

logger.Infow("deleting resource",
    "resource", hl.resourceFn.Type(),
    "namespace", resource.GetNamespace(),
    "name", resource.GetName(),
)

/kind bug