Problem solve Get help with specific problems with your technologies, process and projects.

The guidelines of data retention

Bits & Bytes: Have you ever wondered how long you need to keep your data? Bill Oliver goes over the rules of data retention.

Ed note: A reader posed this question to Bill Oliver about data retention policy. Bill is currently the storage practices manager for PDC Solutions, Inc. Pose your own backup and recovery question to our resident experts.

Question: What are the average industry standards for the length of time data/tape are stored off site? Do we keep a quarter, a month, etc. I understand it's different based on your business but just as a general guideline...

Bill Oliver's answer:

Data is archived for several different reasons. One reason is to comply with state and federal regulations. Another is to provide the ability to recover business critical data in the event of a site-wide data loss, such as a fire or flood. Another reason is to provide a secure repository for point-in-time, snapshot data, for baseline reference in programming, design, custom manufacturing, etc.

The minimum records retention requirements regulations vary by state and by data type, but typically they range from three years to permanent. Legally, most businesses need to keep records long term. Broadly speaking, state and federal regulation require:


  • Business records: 7 years to permanent
  • Contracts: 7 years to permanent
  • Employee records: 3 years
  • Payroll records: 3 to 7 years

    Most high-end backup/recovery software will generate weekly, monthly, or quarterly archive tapes. Generally, a set of archive tapes is generated to coincide with month-end, quarter-end and year-end financial activities. For disaster recovery purposes, data retention is calculated as a trade-off between the costs of recovering lost data vs. the costs of maintaining the data. Like an auto insurance policy, the larger burden the company is willing to shoulder in rebuilding lost data, the lower the costs of insuring that all data isn't lost.

    Many large companies keep weekly archives off-site for a period of one month, and monthly archives for one year. Yearly archives are retained for the required state or federal term.

    A typical full-service backup policy would be: daily incremental backups, retained in the tape library and cycled weekly. Weekly full backups are shipped off-site, typically on Monday and cycled every four weeks. Every fifth week, weekly archive set is retained for one year off-site -- recycled or not, depending on the nature of the data. A full set of archives are generated at a specific time, usually coinciding with the fiscal year, to be retained for the mandated retention period.

    In the event of a site-wide disaster, such as a fire, the maximum amount of data lost would be one week. For point-in-time snapshot data, baselines would need to be established based on the data in question, then full system archives generated. Retention times would depend on the data, and could vary from a few months to permanent.

    For more information:

    The benefits of WORM disk

    Don't give in to e-mail retention hype

    Compliance and digital-content growth drive content-storage market

Dig Deeper on Long-term archiving