Sabalcore’s Dynamic Storage Aggregation (DSA) uses cutting-edge technology to aggregate the performance and capacity of all the local SSDs of the nodes used in a job. This dramatically increases I/O performance. DSA can be used to provide an independent high-performance shared parallel file system for the duration of the job. This provides additional performance and a very efficient way of burst buffering.
DSA can be used to significantly boost the performance without requiring code changes by the user. DSA uses the performance of RDMA over Infiniband and aggregates the storage capacity of all the local drives and SSDs of the nodes of the job, creating a powerful, parallel file system in a single namespace. When the job completes, DSA wipes the drives before exiting. Best of all, there is no additional cost to use DSA – it is free!
- Aggregates all the local drives into one high-performance file system (similar to RAID 0 or RAID10)
- Can increase I/O
- Used by the head-node of the job
- Independent and secure
- Data can be copied to your home directory simultaneously
- Is created at the beginning of the job, then is completely torn down after job completes
How to Use DSA
To use DSA, simply add “-l other=scratch:shared” to the pbs directive and use $SCRATCH_SHARED for the directory. For example:
Using a .pbs script
#PBS -l nodes=4:blue:ppn=12 -l other=scratch:shared cd ~/work # copy the input.dat file cp input.dat $SCRATCH_SHARED
Interactive Job
[user@sci ~]$ qsub -I -l nodes=4:ppn=12:copper -l other=scratch:shared qsub: waiting for job 566450.jman to start qsub: job 566450.jman ready [user@n724002 ~]$ echo $SCRATCH_SHARED /mnt/scratch/566450.jman