Reconsider the default of the TerminateAgentOnImagePullBackOff Reaper being installed, or at least make it easier to disable.

In my Jenkins, every build gets a new pod dynamically launched when that run starts. When a storm of new runs occurs, a lot of pods are scheduled at once, many times resulting in new nodes created (this is running in AWS). This creates many incidents of "Back-off pulling image" showing up in the container state waiting message, resulting in the termination of the pod. This situation is not an error in this case, and the pod will eventually become running.

The good news is TerminateAgentOnImagePullBackOff is installed as a stand alone Jenkins Extension, so I am able to disable it using the Extension Filter plugin.

However, I feel that this is a lurking problem in that the "Back-off pulling image" message is not always a fatal pod condition and is likely causing unexplained pod startups.