hi eric, thats because 16 jobs are running very long time (those jobs are taking 57 hours!! may be it is an error?
Le 14/09/2012 17:18, Tina ODAKA a écrit : those jobs are blocked because of user disk quotat exceded, they doesn't output any activity.
if you are sure that each job runs like in 5 minutes, you can add this option -l walltime=00:05:00 if you use 1G of memory, you can put this, then you have higher chance that your job goes before other big memory job -l mem=1G
if you put walltime=5min, incase there is pb and it take long time, it will kill that job array indecies job, and goes to next array indecies job.
This case never happen (only due to disk usage exceded).
tina
3954951[3031].s lgasche sequenti simulation 20400 1 1 2800mb 96:00 R 57:08 3954951[3032].s lgasche sequenti simulation 24897 1 1 2800mb 96:00 R 57:09 3954951[3033].s lgasche sequenti simulation 20528 1 1 2800mb 96:00 R 57:07 3954951[3034].s lgasche sequenti simulation 24994 1 1 2800mb 96:00 R 57:08 3954951[3035].s lgasche sequenti simulation 29774 1 1 2800mb 96:00 R 57:07 3954951[3036].s lgasche sequenti simulation 25054 1 1 2800mb 96:00 R 57:07 3954951[3038].s lgasche sequenti simulation 30101 1 1 2800mb 96:00 R 57:05 3954951[3039].s lgasche sequenti simulation 25099 1 1 2800mb 96:00 R 57:06 3954951[3040].s lgasche sequenti simulation 29517 1 1 2800mb 96:00 R 57:05 3954951[3041].s lgasche sequenti simulation 29532 1 1 2800mb 96:00 R 57:05 3954951[3045].s lgasche sequenti simulation 29628 1 1 2800mb 96:00 R 57:05 3954951[3046].s lgasche sequenti simulation 29807 1 1 2800mb 96:00 R 57:03 3954951[3048].s lgasche sequenti simulation 20729 1 1 2800mb 96:00 R 57:02 3954951[3049].s lgasche sequenti simulation 20744 1 1 2800mb 96:00 R 57:02 3954951[3050].s lgasche sequenti simulation 32626 1 1 2800mb 96:00 R 57:02 3954951[3051].s lgasche sequenti simulation 25288 1 1 2800mb 96:00 R 57:01 service0>989%
What is the command line for that output ? -- Éric Chatellier <chatellier@codelutin.com> Tel: 02.40.50.29.28 http://www.codelutin.com