Propageted SIGTERM signals results in kill after 30 seconds although terminationGracePeriodSeconds is 120

Question

I have a pod with processes orchestrated from shell script (ubuntu). When SIGTERM is sent to the pod, SIGTERM is propageted to these processes. I find that these processes are killed after 30 seconds, although terminationGracePeriodSeconds is 120. The POD itself is killed after 120 seconds

I was expecting nothing to be done until the 120 seconds were up