|
|
## The situation
|
|
|
When a job need to do read the same input files again and again, or that the read mode is not as efficient as it could be, and that the data does not need to be shared between the allocated nodes, it can be more efficient to first copy the nodes `/tmp` at the begining of the job and read that data.
|
|
|
|
|
|
Slurm provides the [sbcat](https://slurm.schedmd.com/sbcast.html) to that purpose.
|
|
|
|
|
|
## The example
|
|
|
|
|
|
|