Metrics on One-off-jobs?

Hi,
it seems there is no way to monitor the cpu and memory usage of one-off-jobs in the web interface.

It would be nice to see the utilization while running a job, and I can’t quite get top output to add up.

I have a job running on an 8GB instance (plan-srv-011), and according to top, more than 24GiB is being used…??

MiB Mem : 31386.8 total, 605.1 free, 24310.7 used, ...

Am I misunderstanding the top output?
What do others do?

Cheers,
Henrik