Have you been able to give some traceability to this error?
I have a similar situation in which my kong-ingress-controller pod is in CrashLoopBackOff status.
This happens with many frequencies, and after to some minutes (usually more that 1 minute), my pod is running again. I’ve posted details here in this question