The perils of disk imaging

I’ve been having some very good discussions lately with MSPs who offer backup from across the industry representing many different fields and specialties. One thing I have noticed is that some people have an all or nothing approach when it comes to backup and disaster solutions. Personally I feel it’s a bad idea to put all your eggs in one basket. As such the next couple of posts are going to look into perils of using a one size fits all strategy for protecting your business.

Disk imaging, as defined by techterms: “a disk image is a software copy of a physical disk. It saves the entire data from the disk, including the file structure and all files and folders from the disk, in a single file. Because disk images are exact copies, or ‘clones,’ of original disks, they can be used to duplicate disks or serve as full backups in case a system restore must be done.” While useful for disaster recovery scenarios, as a daily backup solution there are a number a drawbacks that users should be aware of.

Large Data Footprints: because the image is a byte-for-byte copy of a physical drive, the initial backup of the target machine will be exactly as large as said machine. Advanced imaging software will compress the data as it gets stored but the rate of compression is heavily reliant on the compression algorithm and the type of data being backed up. Second, because the backup will contain the entire structure of a drive, many unnecessary files can get backed up in the process, bloating the footprint of every target machine in an environment.

For example if you had a small office with a server containing SQL, Exchange and the A/D for the environment that had 300 GB of data total, the initial backup at 2:1 compression would sit at 150 GB on the backup server. If a new full backup was taken every week and retention was set for 1 month, every month that server would use 600 GB of storage. Add incremental changes into the mix and it would be easy for a 300 GB server to take up nearly a terabyte of space in your storage pool. Not an efficient means of saving storage space. If that same office also backed up 10 workstations at 100 GB apiece, with daily growth and incremental changes it’s easy to see another terabyte of storage used every month.

Reliability: As imaging software copies each sector of the hard drive and makes a perfect image of precisely what it sees, this raises questions about the reliability of the image created. If you needed to restore, you would be banking on the usefulness of the hard drives and hardware involved. If anything is corrupt, the restore will fail. This would require keeping multiple versions of the image as a failsafe and would cause a loss of any data since the last non-corrupted image was taken.