Closed johanvos closed 4 years ago
Hi @johanvos! How are you running this? Is it the stress-test you mentioned in an email?
correct
@johanvos The current version of SimulaQron and CQC is unfortunately not so robust against heavy load or stress-test. This will hopefully improve in the next iteration. Is this an issue for you at the moment or something that you stumbled upon while testing?
Thanks, It's not a blocking issue for me. I noticed that the issue doesn't occur when adding small sleeps in between invocations, so I can do that. The more important question is probably if the CQC protocol should (ultimately) deal with concurrency, or if that is the responsibility of a higher layer? I can live with both.
In my view, CQC should deal with concurrency, since the higher layer has little control and information about the timing and execution-time of operations. I think it's therefore useful if the higher layer does not have to care about such things.
I'd be happy to hear if you have different thoughts :)
Occasionally, I get failures running the CQC-Python teleport sample. The expected outcome for Bob to measure should be
0
, but sometimes1
is measured.The log (with debug level) for a failure (not the final measured outcome
1
) is shown below: