Option to parse number of events for profiling jobs
Add option to avoid failing profiling jobs in the PR tests.
The problem is that we run VTune on a running command, and assume that the command runs during the 60 seconds of profiling. Many profiling jobs were failing, since the event loop finished during profiling or even before (VTune throws very cryptic messages when this happens)
This should also help mitigating issues with profiling that offline users are facing (such that you don't have to think about the -n flag during profiling any longer)
The parsing and the way I wrote the code can be (much) improved.