Invalid API Objects Causing Defunct Shell Processes (Fork Bomb Essentially)

[ekristen]

What happened: Defunct Shell Processes are being Created

What you expected to happen: No Defunct Shell Process Creation

How to reproduce it (as minimally and precisely as possible):

Simple provide a bad config to the plugin like trying to monitor a CRD

<source>
  @type kubernetes_objects
  tag kube.objects.*
  api_version "customcrdgroup/v1"
  insecure_ssl false
  <pull>
    resource_name mycrd
  </pull>
</source>

Anything else we need to know?:

Because the kubeclient doesn't have any knowledge of the CRD and because the kubeclient tries to use built in helpers to get the data it errors out. Unfortunately this causes the plugin on each try to create a defunc shell process and eventually eats up all system processes.

Environment:

• Kubernetes version (use kubectl version):
• Ruby version (use ruby --version):
• OS (e.g: cat /etc/os-release):
• Splunk version:
• Others: