Several pods in Pending status and kube-scheduler pod crashing due to runtime error: integer divide by zero in OpenShift 4

Issue

• There are several pods in Pending status with no clear reason.

• The kube-scheduler Pods are repeatedly crashing with the error:

  Raw

  Observed a panic: "integer divide by zero" (runtime error: integer divide by zero)
  

• The panic references the findNodesThatFitPod function:

  Raw

  panic({0x2031cc0, 0x38e93e0})
   runtime/panic.go:884 +0x213
  k8s.io/kubernetes/pkg/scheduler.(*Scheduler).findNodesThatFitPod(0xc0008b1a70, {0x271a0f0, 0xc001ec09b0}, {0x273cbd0, 0xc0004e25a0}, 0x0?, 0xc002e91358)