Overview
Today's scientific research involves increasingly large volumes of increasingly complex data. At the same time, CPU performance has improved well beyond storage media, which can create bottlenecks in HPC environments. Read this section to learn how to optimize your storage usage and improve I/O performance on Palmetto.
Palmetto provides three data spaces: home, scratch, and group project storage.
Every user has a home directory and access to the scratch file systems. Group project storage is allocated for faculty to house their data and can only be accessed by users having the owner’s approval. Each data space is accessible from anywhere in the cluster.
Storage Hardware Grid
The table below describes each file system on Palmetto and the hardware for the storage medium and connection to the nodes.
Name | Size | Disk Type | File System | Connection Type | Backup |
---|---|---|---|---|---|
/home | 250GB per user | SSD | Indigo | Network (IB, ethernet) | ✅ yes |
/scratch | 5 TB per user | SSD | Indigo | Network (IB, ethernet) | ❌ no |
/local_scratch | 99GB - 2.7TB per node | HDD, SSD, NVMe | ext4 | Internal (SAS, SATA, PCIe) | ❌ no |
/project/{path} | Varies by owner | SSD | Indigo | Network (IB, ethernet) | ✅ yes |
What do the abbreviations in the storage hardware grid mean?
The legend below can be used to understand the abbreviations in the storage hardware grid table.
Term | Explanation |
---|---|
HDD | Hard Disk Drive |
SSD | Solid State Drive |
NVMe | Non-Volatile Memory Express |
PCIe | PCI Express |
PCI | Peripheral Component Interconnect |
SAS | Serial Attached SCSI |
SCSI | Small Computer System Interface |
SATA | Serial ATA |
ATA | Advanced Technology Attachment |
IB | InfiniBand |
GB | Gigabyte |
TB | Terabyte |
Performance Guidelines
Generally speaking, /local_scratch
will always be the fastest file system to
use because there is no network involved. However, this space cannot be shared
between a group of nodes participating in a job, and the data must be moved to
permanent storage before the job completes.
/scratch
is our newest parallel file system that runs atop flash storage. It
performs well across most access patterns.
The RCDE team encourages you to test your workflows against all three file systems to see which one works best for you.