Best practice to prevent robot crashes such as out of memory

This is a question in the Mock exam. The middle two choices appear incorrect. Choice “A-Add Clear Collection…” appears to be reasonable as a best practice, but I think that answer was marked wrong. That leaves choice “D-After every transaction…” Yes, it will work to prevent crashes, but it appears to be a very expensive way to achieve this, because of the added overhead for each transaction. What is the best practice? Thanks in advance for clarifying.

image

I was also going through the practice recently.

Since the question is about “repetitive process and out of memory” , best way to do it is isolate the work flows. so the answer is B.

Hi Kayobot,

Can you elaborate on your answer?
The screenshot of this “Isolated” option doesn’t indicate anything relates to robot crash or out of memory. Sounds more like a good way to do “Error handling”.
image

Thank you!