Skip to main content

Research Data Management: Store

This guide provides information and links to resources for data management, including data management plans, file organization and storage, and data sharing in repositories.
Last Updated: Oct 16, 2019 3:28 PM

Making a Backup Strategy

Follow the 3-2-1 rule for data storage:

To ensure that your data is sufficiently backed up, you should keep...
• at least 3 copies of a file
• on at least 2 different media
• with at least 1 offsite

For example: your research data is stored on the internal hard drive of your office desktop computer, on an encrypted flash drive in your office, and on UB Box. There are 3 separate copies of your files, including the original copies and at least 2 backups; stored on at least 2 different media - internal hard drive and external flash media; with at least 1 offsite - in this case, in the cloud (distributed across one or more of the cloud service Box's data centers).

If one storage medium fails - your desktop computer crashes or you lose your flash drive - you should hopefully still have two copies of your data. Keeping one copy offsite or in the cloud helps prevent total data loss if there is a local disaster, such as fire or flood, that affects UB.

What details should a backup strategy include?

  • How will you back up your data?
  • How regularly will backups be made?
  • Will all data be backed up or only those that have been changed? (incremental vs. full backups)
  • How often will full and incremental backups be made?
  • How long will backups be stored?
  • How much storage space is required to maintain backups?
  • How will you keep track of different data versions, across multiple devices?
  • What backup services or tools will you use?

File Organization Best Practices

File Structure & Naming Conventions

When naming your research folders and files, take the following into consideration.

1. Organization - Your folder structure is important for future access and retrieval, and folder naming conventions should take into account the file naming constraints of the system where the file is located. For example, be specific when naming folders so that sub-folders in two different directories do not have identical names, or include symbols or characters that will cause errors if the directories are moved to a different system.

2. Context - This could include content-specific or descriptive information, independent of where the data are stored. Consider including some contextual information in your file names in case a file is moved to a new directory or shared without its containing file structure.

3. Consistency - Choose a naming convention and ensure that the rules are followed systematically by always including the same information in a file name (such as date and time) in the same order (e.g. YYYY-MM-DD).

Best Practices

  • Don't use spaces in file names.
  • All file extensions should be lowercase:     .jpg, .mp4, .mov   —not—   .JPG, .MP4, .MOV
  • Limit the file name to 32 characters (preferably less).
  • When using sequential numbering, use leading zeros:
    • For a sequence where you anticipate 10+ items:     project01.csv, project12.csv
    • For a sequence where you anticipate 100+ items:     sample001.txt, sample099.txt
  • Don't use special characters (! @ # $ % ^ & * {} [] ? <>).
  • Use only one period, before the file extension:     name_date.doc
  • Avoid using generic filenames that may conflict when moved from one location to another.
  • Use ordinal numbers to track major versions and dashes for minor versions:     data_v1.csv, data_v2-02.csv

SUNY Policies

The SUNY Research Foundation's Records Management Policy may be used as a guideline when developing a data management plan and assigning responsibilities to project members for organizing and maintaining project records and information.

The related Record Retention for Project Administration Records document recommends retaining project records, such as administrative award records as well as project data, for 6 years after award-closed dates.