Disk Hardware

The RDF consists of 23PB of usable storage, with an additional 48 PB of backup tape capacity for disaster recovery purposes. Data recovery for accidental deletion is not supported.

The disk storage (which was completed in March 2015) is based on seven DDN 12K couplets. The Metadata storage is on NetApp EF550/EF540 arrays populated with SSD drives.


The disaster recovery backup capability is managed by Tivoli Storage Manager, and an IBM TS3500 tape library with 36 drives.

Data Transfer Nodes (DTNs)

The DTNs are used for transferring data onto and off the RDF disks. As well as high-bandwidth connections to the JANET backbone, they support direct connections to the JASMIN facilty and to the PRACE pan-european HPC infrastructure.

Each DTN is equipped with:

  • One Intel Westmere 8-core processor, 11 GB memory

Data Analytic Cluster (DAC)

The DAC is a flexible compute resource based on Linux containers. This means that the facility can provide compute resource as "standard" HPC images available to all RDF users though a batch system and can also support specialist images for particular communities.

  • 1 login node: Two Intel Ivy Bridge 10-core processors, 128 GB memory
  • 12 standard compute nodes each with two Intel Ivy Bridge 10-core processors, 128 GB memory
  • 2 high-memory compute nodes each with four Intel Westmere 8-core processors, 2 TB memory

HyperThreads are enabled on all nodes meaning that standard compute nodes each have 40 CPUs available and the high-memory compute nodes each have 64 CPUs available.

All DAC nodes have high-bandwidth, direct Infiniband connections to the UK-RDF disks.