it happens pretty much always if once triggered but I believe to have noticed it is happening on different domains. I did set the parallel chunk nodes to one chunk and the issue was not triggered. Though, the scan for almost 1k websites ran over night so I might have missed it but I did not spot anything at first glance and the try-catch nodes did not log anything.
If I am able to reproduce it, I will share the workflow with you via Email as it contains sensitive information.