HMS IT Prepares for Data Center Relocation in January 2025

MyHMS speaks with Spencer Pruitt and Jeffrey Farese

HMS OFFICE OF COMMUNICATIONS & EXTERNAL RELATIONS HMS Community
Front view of Gordon Hall surrounded by vibrant fall foliage.
Image: Jessica Kisluk

As announced earlier this year by HMS IT, the entirety of HMS and HSDM’s data storage systems and enterprise and research technology hardware will be relocated in January 2025 to a new facility within Boston’s Markley Data Center.

This move is part of a comprehensive initiative to build a new state-of-the-art facility with upgraded power, cooling systems, and structured cabling. The project represents a significant investment in modernizing HMS IT’s infrastructure, ensuring the stability and scalability needed to support the groundbreaking work being done across HMS and HSDM. It also represents significant cost savings — approximately $2 million annually — on power, cooling, and space usage.

Promotion for Surgical Leadership Program's One Year Certificate from Harvard Medical School, learn more

MyHMS recently spoke with Chief Information Officer Spencer Pruitt and Assistant Chief Technology Officer for Architecture and Operations Jeffrey Farese about the project and what members of the HMS and HSDM communities should know when they return from winter recess.

MyHMS: What is a data center, and how does it support our work and research at HMS and HSDM?

Spencer Pruitt: The Markley Data Center in Boston is the backbone of HMS and HSDM’s technology infrastructure. It supports every facet of the Schools’ operations, from routing network traffic to providing cloud services, storing research data, and offering computational resources. Essentially, all HMS and HSDM data flows through this facility.

Jeffrey Farese: Additionally, the work HMS IT did during our research data migration efforts, which concluded in September, allowed us to establish a second data center at the Massachusetts Green High Performance Computing Center (MGHPCC) in Holyoke, Massachusetts. This is an intercollegiate high-performance computing (HPC) facility and a collaboration among research universities, including Harvard, Boston University, MIT, Northeastern, the University of Massachusetts system, and Yale.

Now, all research data is instantly replicated across both the Markley and MGHPCC sites, which enhances data protection and computational resiliency.

MyHMS: Why is the data center located off campus?

Pruitt: The availability of power to support the infrastructure of HMS, HSDM, and the many Longwood Medical Area hospitals creates some restrictions on what can and should be housed on or near the Quad. The Markley Data Center provides world-class facilities with robust power, cooling, and monitoring services — away from campus. This level of resilience is critical to ensuring business continuity, safeguarding research data, and supporting the innovative work happening at HMS and HSDM now and in the future.

MyHMS: Why is this move happening now?

Farese: Server technology has evolved significantly since HMS and HSDM first started using the Markley Data Center, which has made our current space too constrained to support the growing power and cooling demands of newer infrastructure.

Pruitt: As part of a multiyear technology infrastructure revitalization, we’ve renegotiated our lease with Markley, which will result in significant savings on power, cooling, and space usage. Consequently, we need to move the Schools’ data center to a new, state-of-the-art space within the facility.

MyHMS: What does the move entail, and what disruptions should we expect?

Pruitt: The move will involve relocating network connectivity, as well as primary storage and computing resources that support critical research, teaching, and administrative systems. There will be significant service disruptions from Jan. 3 to Jan. 10, 2025. While this brief period of inconvenience is necessary, it will ultimately lead to long-term benefits, such as improved performance, enhanced infrastructure, and reduced energy consumption.

Farese: The build-out of the new facility is already underway. The first major migration is scheduled for the evening of Jan. 3, during which most of our systems and services will be relocated. This phase will be highly disruptive, with an anticipated 18-hour outage from 6 p.m. Friday, Jan. 3, until noon on Saturday, Jan. 4. HMS and HSDM building access, basic internet, and HarvardKey-based services will not be affected. From Jan. 4 to Jan. 10, we will conduct a second major move, focused primarily on the HPC cluster O2 and its supporting infrastructure, which will impact campus researchers only.

After these two major migrations, several smaller, coordinated moves will occur, with timing tailored to minimize disruption for affected groups. The entire relocation process is expected to be completed by Jan. 31.

MyHMS: Once the new data center is operational, will we notice any changes in our day-to-day work? Will we need to do this again?

Pruitt: Our primary goal is to create a robust, reliable, and resilient infrastructure that will provide seamless service for years to come. Most day-to-day work will remain unaffected, though users may notice improvements in system performance due to significant upgrades and facility enhancements.

Farese: As for whether we’ll need to move again — yes, but not for at least another 10 years. The upgrades we’re implementing are designed to meet both current and future needs, while allowing room for growth and technological advancements.

For more information about the data center relocation, including frequently asked questions, visit HMS IT’s data center relocation webpage. For additional questions, contact HMS IT at 617-432-2000 or via email at itservicedesk@hms.harvard.edu.