View source for Generic resource requirements

==Generic resource requirements==

The sbatch command takes several options to specify the requirements, of which we list the most commonly used ones below.
  $ sbatch --time=2:30:00

For the simplest cases, only the amount of maximum estimated execution time (called“walltime”) is really important. Here, the job requests 2 hours, 30 minutes. As soon as the job exceeds the requested walltime, it will be “killed” (terminated) by the job scheduler. There is no harm if you slightly overestimate the maximum execution time. If you omit this option, the queue manager will not complain but use a default value (one hour on most clusters). If you want to run some ﬁnal steps (for example to copy ﬁles back) before the walltime kills your main process, you have to kill the main command yourself before the walltime runs out and then copy the ﬁle back. See section [[Running a command with a maximum time limit]]  for how to do this.  
  $ sbatch --mem=4gb

The job requests 4 GB of RAM memory. As soon as the job tries to use more memory, it will be “killed” (terminated) by the job scheduler. There is no harm if you slightly overestimate the requested memory.
  $ sbatch --ntasks=5 --ntasks-per-node=2

The job requests 5 compute nodes with two cores on each node ( “processors per node”, where "processors" here actually means "CPU cores").

  $ sbatch --ntasks=1 -–constraint=westmere
The job requests just one node, but it should have an Intel Westmere processor. A list with site-speciﬁc properties can be found [[here]]. These options can either be speciﬁed on the command line, e.g. 

  $ sbatch --ntasks=1 --mem==2gb --job-name=fibo.sh
or in the job script itself using the #PBS-directive, so “ﬁbo.sh” could be modiﬁed to:

  #!/bin/bash -l  
  #sbatch --ntasks=1  
  #sbatch --mem=2gb  
  cd $SLURM_SUBMIT_DIR 
  ./fibo.sh

Note that the resources requested on the command line will override those speciﬁed in the Slurm ﬁle.