Adopt sustainable data practices to safeguard storage capacity and integrity for future research.

To support existing experiments, ensure space for future research data, minimize security risks, and maintain compliance with federal regulations, we must work together to identify and delete non-essential research data that are not required to be retained. This will help promote sustainable storage practices and protect our essential data, in line with University requirements and the expectations of sponsors and contractors.

To address this, HMS IT has launched a community-wide data cleanup initiative. In early 2026, we will contact principal investigators, lab data managers, and researchers to provide more information about how to improve data lifecycle management. Opportunities to learn will include workshops, one-on-one consultations, and office hours; data management workflow optimization tools will be made available.

Who is affected

This initiative applies to all users of HMS centralized storage systems.
Users Scope

All users of HMS centralized storage are affected, including:

  • Principal Investigators (PIs)
  • Department administrators
  • Lab data managers
  • Lab members
  • IT staff managing research data

Timeline and rollout

Work continues through 2026 with workshops and lab engagements scheduled for the first half of the year.
Schedule Phases

The initiative will continue in phases through 2026. During the first half of the year, you can expect planned workshops, targeted communications, and scheduled individual lab engagements. During this time, we will assist with data cleanup and introduce improvements, including data tiering, lifecycle management, and opportunities for automation.

Required actions

When contacted, PIs and lab members must review usage reports and evaluate for retention research data that are not in active use.
Clean up Compliance

The PI responsible for the data should be consulted to confirm whether the data are eligible for deletion and to ensure there are no other relevant retention requirements (for example, as mandated by an IRB or DUA). To learn more about essential vs. non-essential research data and Harvard data retention requirements, review the Office of the Vice Provost for Research (OVPR) Retention and Maintenance of Research Records and Data Frequently Asked Questions (PDF).

If you are a principal investigator (PI) or lab data manager, you will be contacted to participate. When notified:

  • Learn the guidelines – Familiarize yourself with guidance on data retention and data storage offerings. Follow instructions provided in targeted communications.
  • Review usage – Analyze data usage reports and dashboards for your group and share the findings during internal lab meetings.
  • Evaluate inactive data:
    • Delete redundant, unused, or irrelevant data that are not required to be retained.
    • Organize data that must be retained but is not in active use; move it to Standby and Cold storage systems.

Risky behaviors to avoid

Do not store research data on personal devices, and do not delete files without confirming with the PI responsible for the data.
Security Risks

To ensure data security and compliance, follow these guidelines:

  • Personal local storageDon't store HMS research data on local hard drives (like external drives or ad hoc servers) or personal devices. This risks data integrity and security, and removes the data from proper lifecycle management and HMS IT support.
  • Personal cloud accountsDon't move or store HMS research data on personal cloud accounts (like Google Drive, Dropbox, or iCloud). This risks non-compliance, data security issues, and loss of data access when researchers depart HMS.
  • Data deletionDon't delete data that may still be needed, or that are required to be retained, without consulting the PI or designated data manager. This risks the irrecoverable loss of essential research data.
  • Storage usage reportsDon't ignore storage usage reports, dashboards, or requests for data clean-up. This risks capacity constraints and possible restrictions on storage space for new data.
  • Sensitive dataDon't move or store regulated, sensitive, or human subjects data to locations not approved for that data type. This risks compliance violations and privacy breaches.

Policy and infrastructure updates

Recent changes to HMS research data storage systems include reduced snapshot retention, usage-based quota reviews, and infrastructure migration.
Snapshots Quotas
  • Snapshot retention – As of November 7, 2025, daily Standby storage snapshot retention has been reduced from eight weeks to 21 days, and weekly snapshot retention is no longer available.
  • Storage quotas – Lab and group allocations are reviewed periodically. Adjustments may occur based on usage, data type, and institutional priorities.
  • Cost containment – Primary research storage for the O2 cluster was migrated to cost-effective infrastructure in FY24. While we are committed to maintaining centralized storage as a shared resource, persistent capacity pressure may require usage-based limits or cost-sharing models in the future.

Support

Access training, dashboards, and direct assistance from the Research Data Management team.
Training Contact

We appreciate your efforts to improve data stewardship. To assist you, support options will be available throughout the first half of 2026:

  • Tools – Dashboards for managing your data.
  • Guidance – Regular communications regarding expectations and policy updates (sent in advance of enforcement).
  • Education – Training sessions and office hours.

For questions, feedback, or assistance, contact the HMS IT Research Data Management team.

Learn the criteria for retaining or cleaning research data securely.

Detailed guidance on data retention requirements for different scenarios.

Compare Active, Standby, and Cold storage tiers for your research needs.

Email the RDM team for assistance with dashboards, training, or cleanup.