University of Washington School of Public Health
Online/Offsite Data Backup
There are a great number of options and ways to backup data and not all backup/recovery scenarios satisfy the same goal. Individual faculty and researchers may need to find a reasonable way to protect their valued files/data from computer failures and recovery from local/regional disasters. Under a best case scenario the online/offsite backup systems usually would be your second or third line of backups. Relying only on the online/offsite backups can be problematic, but any backup is better than none). The online/offsite backup solutions are a way to satisfy a degree of recovery for random hardware failure as well as the catastrophic disaster recovery (fire, etc.) of critical data and computer files.
Always check first with your local IT support for backup policy and centrally supported methods. You may have a backup solution available and supported by your department or research group. Even if there is no central supported solution before implementing your solution review it with your local IT support because you may be calling on them for help sooner or later.
It’s a good time to consider a backup checklist,
- Review/know what should be backed up:
- Critical to daily operations
- Valuable/irreplaceable data
- HIPAA or FERPA ramifications
- NIH or other grant may require that research data related to grant if fully protected
- Amount/size (megabytes or gigabytes: MB or GB)
- What type of computer systems are involved in storing the data locally (PC or Mac, servers, mainframes)
- What backup/recovery scenario is desired/being satisfied?
- Random file deletion/error restore
- System restore from failed hardware
- Critical data restore from failed hardware
- Critical data restore from local disaster
- Check with your local IT support first for backup policy and solutions. Your department or research group may have a supported a backup solution. Use it if available.
- Always consider an offsite backup system that provides some automation. Unless the data/files are static (not changing) any manual backup system will fail fairly quickly and/or not provide the level of protection expected. Manual systems could be burning a CD/DVD of computer files or manually copying computer files to network storage locations. These methods force the user to deal with versioning, scheduling, and security tasks that automated systems do for you.
Here are two basic options for offsite backups:
- Tapes or other physical media transported and stored at a location out of the region.
The UW’s official offsite provider for physical media is Iron Mountain. With their service your physical media (tapes) are picked up in a locked container on a scheduled basis (weekly-monthly) and transported to one of their storage facilities out of the region. This service is best if you are already backing up to tapes or other transportable physical media and/or have large amounts of data to backup up. A single tape cartridge can hold 2TB+ of data and the container can hold multiple tapes so that several terabytes of data to be rotated and stored offsite. Refer to UWT website (http://depts.washington.edu/uwtscat/node/151) for additional information. Cost for a monthly container pickup will be approximately $225.00.
- Network or online backups
There are a plethora of 3rd party online services most with a low cost entry point designed for the single user Windows or Mac computer. The UW does not officially support or endorse any of these. Its officially supported/offered solution is the IBM Tivoli Storage Manager (TSM) (http://depts.washington.edu/uwtscat/node/151) which is an enterprise mission-critical per server backup offering. If a large amount of the department’s research data was stored on a single server the TSM service would be worth consideration. The consumer or small business oriented online backup services may seem like a perfect solution for the individual faculty/researcher but there are several points to be aware of:
- Single computer Most online backup solutions are consumer oriented single computer solutions supporting Windows and Mac. If you need to backup multiple computers you must install and duplicate procedures on each computer. Windows or Mac servers may or may not be supported.
- Speed Usual connection speeds will be approximately DSL (250Kb-1Mb/sec) quality. Providers must throttle the connection speeds to satisfy all client connections. Multi-gigabyte backups will take a long long time to complete. A one-time backup of 50GB can take 2-4 days to complete. If you are attempting to backup more than 20GB on a regular basis online solutions may be problematic.
- Automation Most provide very good scheduling or automatic backup of selected files. Where some lack are at the user interaction for selecting what to backup and the inability to run scheduled backups when the user is not logged on. Suggested providers should be easier to use and provide flexible scheduling options and unattended scheduled backups.
- Security Bear in mind how data/files are transmitted to the backup site and that break-ins or compromises at the provider can expose your sensitive data. Review the provider’s encryption policy. Know if your data has HIPAA/FERPA sensitivity. Suggested services provide encryption to satisfy sensitive data requirements.
- Retention Online services mirror your files/folders as they currently exist on your computer. Most services either automatically backup files when changed or via a set schedule. Changes or deletions on purpose or not will be reflected in the backup set! The suggested services do provide a 30-day retention policy so that such problems discovered within 30-days can be recovered.
Online backup for the single computer at the consumer/small business area is beginning to mature but is still changing. There are a significant number of providers such that all have not been reviewed nor should the suggested ones be thought of as the only ones to consider.