Nodes transition to NotReady / processes stuck in D-state with Palo Alto Cortex XDR

Issue

• One or more worker nodes abruptly transition to a NotReady state.
• Application pods on the affected nodes fail, remain in ContainerCreating, or report CNI errors.
• The container runtime (crio) and kubelet stop logging or updating their status.
• Multiple processes (such as crio, ovnkube, or bash) are stuck in uninterruptible sleep (D-state):

      ps -eo state,pid,cmd | grep "^D"

• Stack traces of blocked processes show them waiting indefinitely on fanotify_get_response.
• System journals or coredumpctl show the Cortex XDR pmd agent process crashing (e.g., SIGABRT, SIGSEGV, or kernel traps) shortly before the node hangs:

      coredumpctl list /opt/traps/bin/pmd