I have a script that scrapes a website by making parallel HTTP requests.
I find that when I run the script, the cpu time usage goes up by different amounts seemingly not related to how long the script took to run.
For example, it ran for 50 seconds and only used 40 seconds of cpu time. Then I ran it again and it took 65 seconds and the cpu time usage increased by 135 seconds.
The script was doing the exact same requests both times. Is there a reason the cpu usage increase is so variable and is there anything I can do to make it more predictable?