Use the quota command to see a list of all disks on which you have been allocated disk space and how much disk space you may use.
| Available Disks | |||
|---|---|---|---|
| Filesystem | NFS Host | Size (GB) | Also Mounted On | /h1 /hpc3 /h2 /k1 |
Alcor Alcor Tempus Tempus |
2 2 5 2 |
Plato, Alioth, Alphaid, Mizar Plato, Alioth, Alphaid, Mizar Plato, Kronos Kronos |
In summary, Tempus users can use either /h2 or /k1 while users of the Alpha cluster can use either /h1 or /hpc3.
If you decide to use the Alpha cluster and your job is I/O intensive, you will probably want to ensure that the job is accessing local disk on the same system on which the job is running. This will help to minimize the network traffic involved in disk access which can be a limiting factor in job performance. For example, if your I/O intensive job is running on alioth and writing to the shared disk /h1 hosted on alcor, you will notice that the job will run much slower. For this reason, it is advised that you copy your files to /scratch, a separate local filesystem on each computer, before running your job. An example of this is found in the LSF FAQ.