The Kubernetes cluster which runs transformations had hit its memory resource limit as a result of our operations team installing new monitoring software.
12/03/2020 ~2:00 PM PDT - Monitoring software was installed and memory started to climb.
12/03/2020 ~4:00 PM PDT - Memory on the Kubernetes cluster was increased and we saw transformations being to complete successfully.
We setup our clusters to run nodes with 8GB of memory. This not only gives plenty of room for the transformation running and the other agents that run, but also lets the transformation running actually hit its 4GB limit.