Workflow logs are showing that the workflow context/workflow activities are being canceled due to an insufficient amount of system memory.
Release or Environment
Endless loop in workflow script.
Check the workflow in question and ensure there are no loops such as for loops or while loops that never end. This can be taxing on an instance as the loop will continue running and eating up all the system memory until the instance crashes or cancels that job. Once the workflow activity job is canceled, eventually, the workflow activity will be picked up again by a node worker which will eat up all of the instance's memory again. In these cases, it is best to cancel the entire workflow, work on removing all endless loops, and then reattaching the new version of the workflow to the affected record.