Since two commits my submission fail on the 40 cores machine.
Here is the message I have :
error on a 40-cores HT machine :
couldn't unzip submission -> timeout of command execution, >2000 ms.
Does anyone have the same error?
I happened once a few days ago. The reason probably was that we submitted a new solution before the previous ran on that machine.
Ahh... Ok, this must effectively be the reason. Thank you very much, I will wait before submitting anew.
This is strange but we are having the same problem right now.
I have another question : can we assume that our program will not be tested on small inputs (less diagonals than the max number of threads)? It would avoid me to handle special cases.
I wouldn't rely on this assumption. Your program should always give the same results regardlessly of the selected number of threads. Why shouldn't someone try to run your program with 1000 threads?
Well, because there are no machines with 1000 CPUs... But OK, to be safe I will correct this. Do you know when I will be allowed to submit anew?
yeah, i get this message since today too
It seems good for us now
I can also test again.