Guides at Other Places
This guide drew heavily from the guides listed below. In particular, Managing and Sharing Data from the UK Data Archive was a major source for information.
Access to Data
Security should be loose enough that who is working on the project and needs to access the data can what they need, but it should be tight enough to minimize malicious or accidental damage or destruction.
Security considerations include
- Physical Access
such as locked doors and file cabinets
- Network Security
such as whether the data files are accessible to an external network and the virus protection on the computers used to access the data
- File security
such as password protection to access data files
Encryption & Compression
Storing your files unencripted and uncompressed reduces the chances that a lost password or a computer encription or compression error will corrupt your files. On the other hand, security considerations for confidential data or the cost of storage space for large files may outweigh this consideration.
Systems that automatically create a backup with every change to a file are good for always having an up-to-date backup. Periodic backups that are not synchronized with the active files are useful, say, to retrieve a copy of the data when a file is accidentally deleted. Backups should be made regularly, and automated systems can make backups for you, so they don't get skipped when you are too busy.
On-site or off-site?
MIT's recommendation is to have three copies: the active data file, an on-site backup, and an off-site backup, but the UK Data Archive warns that for confidential data, no more copies than needed should be made, so data files are more easily kept secure.
On-site backups can be made and accessed more quickly than off-site backups -- assuming you store them in a convenient place keep your backups well-organized and well-labeled.
Off-site backup provides protection in case of incidents such as fire and flood that could destroy both an original and an on-site backup. Ideally, off-site backup is geographically far from you to provide protection from major natural disasters that could affect an entire geographic region. Considerations such as cost, data confidentiality, and ability to back up large files may limit your ability to have off-site backup.
Regardless of whether your backup systems are automated or manual, and whether they are on-site or off-site, you should periodically check that the backup files can be opened. Use information such as date, file size and checksum comparisons to verify that the backups match the originals.