You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
For now, the docker kills the judge when it exceeds the set memory limit. This should be handled by the judge itself (if possible). Oddly enough, the judge reports the error as a Timeout, whereas it should be an indication that the docker was killed because the memory limit has been exceeded.
@ninewise Can you check why Dodona does not report this error as a memory limit issue. Clearly, when I run the script, it is killed before the time limit is exceeded.
When the last script in the docker exits with status code 143 (which is will when it's killed by the docker), this is reported as a timeout. I can fix this (change it to an internal error, which is more accurate).
However, I can't make the distinction between this memory-error and, say, another exit 143 (like the one from an actual timeout). I left it as is, because the pythia judge has quite some submissions hitting this memory-killing/timeout backup catch.
For now, the docker kills the judge when it exceeds the set memory limit. This should be handled by the judge itself (if possible). Oddly enough, the judge reports the error as a
Timeout
, whereas it should be an indication that the docker was killed because the memory limit has been exceeded.Example: submission 817578
@ninewise Can you check why Dodona does not report this error as a memory limit issue. Clearly, when I run the script, it is killed before the time limit is exceeded.
Original issue by @pdawyndt on Sun Apr 30 2017 at 18:31.
The text was updated successfully, but these errors were encountered: