TMCnet News

Importance of Back Up and Effective Ways of Doing It
[December 08, 2009]

Importance of Back Up and Effective Ways of Doing It

TMCnet Contributor
Organizations the world over are posed with the problem of maintaining back ups of all the data they have collected over a period of time. Data backup is important because it ensures that they are not lost completely and they can be used when the need arises. If data needs to be stored only for the sake of not losing it then archiving on magnetic tape is sufficient. But data needs to be available readily so that it can restored to usability.

The e-book on Back Up Philosophy published by Realtime Publishers talks about the hassles of backing up large volumes of data and how ineffective it is to restore them and make them usable by using Backup 1.0 techniques. The necessity of back ups, where to store the data, whether it should be tested and continuously monitored are some of the critical points covered in the paper. Further various approaches to back up, file based backups and image backups are also discussed in detail.

The older technologies involved punch cards and UNIVAC I computing machines followed by magnetic tape, floppy disks and recordable compact disks. These methods are outdated now and there is a compelling need to think and use Backup 2.0 technologies, says the white paper. Simply copying data does not make restoring it easy. As the volume of backed up data increases it becomes even more difficult. For example Microsoft’s (News - Alert) TerraServer stores satellite photos with data over 1 terabyte. It takes 8 hours to back it all up. Even SQL Server’s native backup capabilities are not enough.

Some organizations try to save time in backing up by taking images of their servers which aids in disaster recovery. The software takes a copy of entire hard drive, compresses it and stores it. This helps in backing up entire server but traditional methods will be needed to get recent data back. This is a limitation as it provides archiving and not restoration.

The paper discusses in detail the need for data back up as in disaster recovery of an entire server or entire data center which needs to be restored on site or at a different location. A good back up is one which restores data without losing any data or work and with minimum downtime. The back up should restore all the data in less time and fewer steps and the system should have a physically protected storage.

The discussion deals with the content that needs to be backed up, when should backup be attempted, and the advantages and disadvantages of RAID 1 disk mirroring. Backups can be made as a full backup, differential backup, incremental backup or agent-based back up, according to the paper. It also recommends off site back up of data as in the eventuality of floods, fire or natural disasters all data will perish without a trace if stored in one area. Back ups are a form of insurance.

The discussion leads to Backup 2.0 technologies such as Virtualization and the creation of an event log entry alert in SCOM which is a set of preconfigured rules for monitoring and alerting systems managers on a specific application. The desirable features of a Backup 2.0 technology are clearly stated such as real time availability, minimum interference and ability to move snapshots of backups to tape among others. For more information, click here.

Shamila Janakiraman is a contributing editor for TMCnet. To read more of Shamila’s articles, please visit her columnist page.

Edited by Patrick Barnard

[ Back To's Homepage ]