kubearmor/kubearmor-relay stopping abruptly

[nyrahul]

Bug Report

It is found that the logs/telemetry from kubearmor-relay are abruptly stopped once in while.
The root-cause of the problem is not known yet.

Following are the observations (attached below is a screenshot):

1. Observed failure to send a log in kubearmor GRPC server
2. Observed in one case that the NodeIP (of kubearmor node/pod) is empty on kubearmor-relay
3. WatchLogs() no more working but WatchAlerts() keeps working fine.

General Information

• Environment description (GKE, VM-Kubeadm, vagrant-dev-env, minikube, microk8s, ...)
  GKE-COS

• Kernel version (run uname -a)
  NA

• Orchestration system version in use (e.g. kubectl version, ...)
  k8s 1.22

To Reproduce
We do not have exact steps to reproduce the issue. However, the issue springs up when karmor log is started-stopped while connecting to kubearmor-relay.

Expected behavior
It is expected that log/telemetry keeps working regardless of failures on karmor, kubearmor-relay or kubearmor itself.

Screenshots

[nyrahul]

Another instantiation of the same problem (reproduced in the same env):

CC: @nam-jaehyun