$BIGWORK – Work file system for very large files
The BIGWORK directory is intended for large files as well as simulation and scratch file output. Please start your batch jobs either from within your BIGWORK directory or within your batch script change your working directory to your BIGWORK directory before your simulation starts. For example:
cd $BIGWORK
./my_program
| $BIGWORK | /bigwork/username |
| Hardware | Infinite Storage |
| Capacity | 69 TB |
| Quota | 1TB soft quota; 10TB hard quota |
| Backup | no backup |
| file lifetime | unrestricted |
| Locality | entire cluster system |
| Availability | no guarantee of file space |
| Performance | high performance due to the global file system |
| Usage | large as well as checkpoint or restart files |
| Type | Lustre |
Top
Print