Among HT’s six core facilities, there is a high performing Data Centre to store, manage and analyse huge amounts of data and information and support researchers’ activities.
Once Human Technopole is fully operational the Data Centre will be in a new technical area which complements the South Building and will support the computational needs of the entire institute.
In the meantime, we’ve had to come up with alternative solutions to get the Data Centre up and running. We spoke with Daniel Anderson, IT Manager at HT to better understand how his team is working to provide data storage and analysis to a growing and changing institute.
What is a Data Centre and why is it important for a research institute?
A Data Centre is a vital part of any organisation, big or small. Physically, it is the room which hosts the servers and IT equipment allowing to store, archive and process data and information from various departments.
Nowadays research centres generate a huge amount of data. For example, the daily data output of cryo-EM microscopes amounts to around 5 terabytes, while a full genomic characterisation of around 200-300.000 people requires around 4/5 petabytes. This high volume of scientific data output demands a Data Centre as close as possible to the instrumentation that produces it, with enough supporting IT infrastructure to manage the throughput. Having an on-premise Data Centre allows HT researchers the shortest possible time to process their data.
What is the current setup of HT’s Data Centre?
Up until today, the Data Centre was housed in a prefabricated shelter just outside Palazzo Italia. It hosted the first fundamental infrastructures necessary to start computational research activities and provide an initial home for core services.
However, over the past two years, the demands of the IT infrastructure have increased, and we are now reaching the limits of the shelter in terms of power consumption, cooling and physical space.
What solution did you come up with?
It was a real team effort: together with the ICT & Digitalisation team and the colleagues from Procurement, Administration and Campus Development & Facility Management, we have identified a mid-term strategy for the coming years as we wait for the South Building to be completed.
Three rooms in the basement of Palazzo Italia are currently under renovation and will be the new home for HT’s High Performance Computing (HPC) cluster and for a soon to be acquired tape library to be used for long term backup.
This solution provides enough space to comfortably host the HPC plus it’s planned CPU expansion arriving later in 2021 and the data backup systems, leaving some extra room for additional services while at the same time allowing to free up precious space in the shelter for further expansion.
A lot of care and attention has gone into the design to ensure the correct cooling and power are provided and that both can remain operational in the event of a failure.
In addition to this new space on the HT campus, we are currently setting up an additional offsite Data Centre which will host new HPC and data storage infrastructures, further extending our compute and storage capacity while adding redundancy to our services. As a final pillar to our Data Centre Strategy, we will leverage cloud services to ensure continuity of HT’s core IT Services.
When will the resources be ready?
The work is expected to finish in November 2021, at which point we will start distributing the various computing services across the infrastructures.