etap

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
Next revision Both sides next revision
etap [2015/06/16 16:53]
meesters
etap [2015/06/17 14:30]
meesters
Line 3: Line 3:
 When using etapfs you should acess these filesystem only within your job and not indicate paths on the command line with LSF commands (e.g. bsub). Especially ''stdout'' should not be directed to ''/etapfs'' as this can prevent a job to be ended (even if the wall clock time limit is hit), if one such filesystem "hangs". When using etapfs you should acess these filesystem only within your job and not indicate paths on the command line with LSF commands (e.g. bsub). Especially ''stdout'' should not be directed to ''/etapfs'' as this can prevent a job to be ended (even if the wall clock time limit is hit), if one such filesystem "hangs".
  
-This means that the '-o', '-e' and '-cwd' parameters of bsub should not point to etapfs. However, copying within your script to etapfs is ok, e.g.:+This means that the ''-o'', ''-e'' and ''-cwd'' parameters of bsub should not point to etapfs. However, copying within your script to etapfs is ok, e.g.:
  
 <code bash> <code bash>
Line 15: Line 15:
 Likewise, pointing the output of subshells to etapfs is an alternate solution. Likewise, pointing the output of subshells to etapfs is an alternate solution.
  
-**Also, a jobscript should not be started within a directory belonging to etapfs.**+**Also, you should not submit jobs when your actual working directory at the time of submit is on an etapfs.**
  
 The reason is that a "hanging" etapfs (e.g. lost mount) will cause the scheduler to be unable to clean up the job, while a job unable to write to etapfs can still be cleaned up, if the scheduler's daemons do not have file handles there. The reason is that a "hanging" etapfs (e.g. lost mount) will cause the scheduler to be unable to clean up the job, while a job unable to write to etapfs can still be cleaned up, if the scheduler's daemons do not have file handles there.