RAL Tier-1

RAL Tier-1
Name	RAL Tier-1
Type	Computational facility
Location	Harwell Campus, Oxfordshire
Established	2000s
Operator	Science and Technology Facilities Council
Country	United Kingdom

RAL Tier-1

RAL Tier-1 is a major high-throughput computing facility serving scientific collaborations and experiments. It supports large-scale data processing for projects connected to facilities such as Euclid (spacecraft), Large Hadron Collider, Diamond Light Source, ISIS Neutron and Muon Source and national research programmes tied to CERN, European Space Agency, STFC Rutherford Appleton Laboratory and university consortia including University of Oxford, University of Cambridge, Imperial College London, University of Manchester. The centre integrates hardware, middleware and operations to provide reliable batch, grid and cloud resources to collaborations like ATLAS experiment, CMS experiment, LHCb experiment and international science partnerships such as Worldwide LHC Computing Grid and European Grid Infrastructure.

Overview

The centre functions as a regional computing hub connecting experiments at CERN with storage and compute resources used by teams at Oxford, Cambridge, Edinburgh, UCL and other institutions. It interoperates with infrastructures developed by OpenStack Foundation, HTCondor and GridPP while aligning with policies from Science and Technology Facilities Council and reporting into governance bodies such as STFC Council. Historically its provisioning cycles paralleled procurement efforts by suppliers like Dell Technologies, Hewlett Packard Enterprise, IBM and standards work from Open Grid Forum.

Infrastructure and Architecture

The hardware layer includes petabyte-scale object and block storage arrays from vendors like NetApp, EMC Corporation and distributed filesystem deployments influenced by Ceph and Lustre. Compute clusters run x86_64 processors from Intel Corporation and AMD with accelerator nodes using NVIDIA GPUs and FPGA racks akin to deployments by Xilinx. Network fabric employs 10/40/100 Gigabit Ethernet and InfiniBand technology supplied by Mellanox Technologies and connects to national research networks such as JANET (UK) and GEANT. Virtualisation and container orchestration involve Kubernetes, OpenStack and bespoke grid middleware from GOC Grid projects to serve workflows originating from ATLAS experiment and LSST-class surveys.

Services and Operations

Operational services provide workload management, data transfer, archival, and user support. Workload scheduling integrates systems like HTCondor, SLURM and bespoke queuing for collaborations including CMS experiment and ATLAS experiment. Data movement utilises tools and protocols pioneered by FTS (File Transfer Service), GridFTP, and rsync-style replication used across sites such as CERN and DESY. Archive services mirror custodial responsibilities exercised by consortia like WLCG while user support is coordinated with helpdesk models from GridPP and training partnerships with University of Oxford and University of Manchester.

Governance and Collaboration

Governance spans technical boards, user committees and funding relationships with bodies such as Science and Technology Facilities Council, UK Research and Innovation, European Commission projects, and international collaborations including CERN, ESA and ITER. Collaboration agreements reflect models used by GridPP, EIROforum, and memorandum frameworks similar to those between Diamond Light Source and partner universities. External partnerships include cloud credits and service-level agreements with commercial providers like Amazon Web Services and Google Cloud for burst capacity, coordinated through procurement and compliance offices associated with STFC and host institutions like Harwell Campus.

Performance and Metrics

Key performance indicators track throughput, job success rates, storage utilisation and network latency, referencing benchmarking suites and metrics frameworks used at CERN and ESnet. Capacity planning aligns with projections for experiments such as ATLAS experiment, CMS experiment, Square Kilometre Array data models, and space missions like Euclid (spacecraft), with metrics compared against operational baselines maintained by WLCG and EGI. Regular audits and performance reviews involve stakeholders from University of Cambridge, Imperial College London, and funding reviews by UK Research and Innovation.

Security and Compliance

Security policies adhere to standards influenced by national guidance from NCSC (United Kingdom) and European regulations such as directives administered within European Commission frameworks. Incident response and log analysis practices draw on capabilities used at CERN and national computer security incident response teams like CERT-UK. Data governance follows principles required by collaborations such as ATLAS experiment and CMS experiment, with compliance reporting coordinated with funders including Science and Technology Facilities Council and oversight by institutional IT governance at University of Oxford and STFC Rutherford Appleton Laboratory.

Category:High-performance computing Category:Research infrastructures